diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,15180 +1,60500 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.0,
+  "epoch": 4.0,
   "eval_steps": 500,
-  "global_step": 10806,
+  "global_step": 43168,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.0,
-      "grad_norm": 0.33984375,
-      "learning_rate": 3.7002775208140615e-07,
-      "loss": 2.5996,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.63284688441047e-08,
+      "loss": 2.5207,
       "step": 1
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.8501387604070305e-06,
-      "loss": 2.6552,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.3164234422052353e-07,
+      "loss": 2.6419,
       "step": 5
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.3984375,
-      "learning_rate": 3.700277520814061e-06,
-      "loss": 2.6426,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.6328468844104705e-07,
+      "loss": 2.6615,
       "step": 10
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.373046875,
-      "learning_rate": 5.550416281221091e-06,
-      "loss": 2.6381,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.949270326615706e-07,
+      "loss": 2.6051,
       "step": 15
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.353515625,
-      "learning_rate": 7.400555041628122e-06,
-      "loss": 2.6589,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.265693768820941e-07,
+      "loss": 2.6728,
       "step": 20
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.43359375,
-      "learning_rate": 9.250693802035152e-06,
-      "loss": 2.6584,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.1582117211026177e-06,
+      "loss": 2.6555,
       "step": 25
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.3984375,
-      "learning_rate": 1.1100832562442182e-05,
-      "loss": 2.6289,
+      "epoch": 0.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3898540653231411e-06,
+      "loss": 2.6214,
       "step": 30
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.4765625,
-      "learning_rate": 1.2950971322849214e-05,
-      "loss": 2.6038,
+      "epoch": 0.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6214964095436648e-06,
+      "loss": 2.6596,
       "step": 35
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.447265625,
-      "learning_rate": 1.4801110083256244e-05,
-      "loss": 2.5957,
+      "epoch": 0.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8531387537641882e-06,
+      "loss": 2.6326,
       "step": 40
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.431640625,
-      "learning_rate": 1.6651248843663276e-05,
-      "loss": 2.63,
+      "epoch": 0.0,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.0847810979847117e-06,
+      "loss": 2.6636,
       "step": 45
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.291015625,
-      "learning_rate": 1.8501387604070305e-05,
-      "loss": 2.6358,
+      "epoch": 0.0,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.3164234422052353e-06,
+      "loss": 2.6103,
       "step": 50
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.28125,
-      "learning_rate": 2.0351526364477336e-05,
-      "loss": 2.5969,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.5480657864257586e-06,
+      "loss": 2.6451,
       "step": 55
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.2201665124884365e-05,
-      "loss": 2.5457,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.7797081306462822e-06,
+      "loss": 2.6152,
       "step": 60
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.40518038852914e-05,
-      "loss": 2.5137,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.011350474866806e-06,
+      "loss": 2.6408,
       "step": 65
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.590194264569843e-05,
-      "loss": 2.5039,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.2429928190873295e-06,
+      "loss": 2.6617,
       "step": 70
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.2578125,
-      "learning_rate": 2.7752081406105457e-05,
-      "loss": 2.4904,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.474635163307853e-06,
+      "loss": 2.6408,
       "step": 75
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.960222016651249e-05,
-      "loss": 2.4536,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.7062775075283764e-06,
+      "loss": 2.6493,
       "step": 80
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.228515625,
-      "learning_rate": 3.145235892691952e-05,
-      "loss": 2.4519,
+      "epoch": 0.01,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.9379198517489e-06,
+      "loss": 2.6687,
       "step": 85
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 3.330249768732655e-05,
-      "loss": 2.4403,
+      "epoch": 0.01,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.169562195969423e-06,
+      "loss": 2.6142,
       "step": 90
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.2578125,
-      "learning_rate": 3.5152636447733584e-05,
-      "loss": 2.4031,
+      "epoch": 0.01,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.4012045401899466e-06,
+      "loss": 2.6185,
       "step": 95
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 3.700277520814061e-05,
-      "loss": 2.3909,
+      "epoch": 0.01,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.632846884410471e-06,
+      "loss": 2.6026,
       "step": 100
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.224609375,
-      "learning_rate": 3.885291396854764e-05,
-      "loss": 2.3601,
+      "epoch": 0.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.864489228630994e-06,
+      "loss": 2.6018,
       "step": 105
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.220703125,
-      "learning_rate": 4.070305272895467e-05,
-      "loss": 2.3702,
+      "epoch": 0.01,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.096131572851517e-06,
+      "loss": 2.5747,
       "step": 110
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 4.2553191489361704e-05,
-      "loss": 2.3449,
+      "epoch": 0.01,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.327773917072041e-06,
+      "loss": 2.6114,
       "step": 115
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 4.440333024976873e-05,
-      "loss": 2.3317,
+      "epoch": 0.01,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.5594162612925644e-06,
+      "loss": 2.6337,
       "step": 120
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 4.625346901017577e-05,
-      "loss": 2.3406,
+      "epoch": 0.01,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.7910586055130885e-06,
+      "loss": 2.6179,
       "step": 125
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.181640625,
-      "learning_rate": 4.81036077705828e-05,
-      "loss": 2.3098,
+      "epoch": 0.01,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.022700949733612e-06,
+      "loss": 2.5657,
       "step": 130
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.19140625,
-      "learning_rate": 4.9953746530989825e-05,
-      "loss": 2.319,
+      "epoch": 0.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.254343293954135e-06,
+      "loss": 2.5616,
       "step": 135
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1953125,
-      "learning_rate": 5.180388529139686e-05,
-      "loss": 2.2914,
+      "epoch": 0.01,
+      "grad_norm": 0.375,
+      "learning_rate": 6.485985638174659e-06,
+      "loss": 2.6128,
       "step": 140
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1953125,
-      "learning_rate": 5.365402405180389e-05,
-      "loss": 2.315,
+      "epoch": 0.01,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.7176279823951815e-06,
+      "loss": 2.5304,
       "step": 145
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 5.5504162812210914e-05,
-      "loss": 2.3091,
+      "epoch": 0.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.949270326615706e-06,
+      "loss": 2.5924,
       "step": 150
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.181640625,
-      "learning_rate": 5.735430157261795e-05,
-      "loss": 2.2459,
+      "epoch": 0.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.180912670836229e-06,
+      "loss": 2.5377,
       "step": 155
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 5.920444033302498e-05,
-      "loss": 2.3152,
+      "epoch": 0.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 7.412555015056753e-06,
+      "loss": 2.5353,
       "step": 160
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.19140625,
-      "learning_rate": 6.105457909343201e-05,
-      "loss": 2.2955,
+      "epoch": 0.02,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.644197359277277e-06,
+      "loss": 2.5549,
       "step": 165
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 6.290471785383904e-05,
-      "loss": 2.2946,
+      "epoch": 0.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.8758397034978e-06,
+      "loss": 2.5021,
       "step": 170
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 6.475485661424607e-05,
-      "loss": 2.262,
+      "epoch": 0.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.107482047718323e-06,
+      "loss": 2.5607,
       "step": 175
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 6.66049953746531e-05,
-      "loss": 2.2851,
+      "epoch": 0.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.339124391938847e-06,
+      "loss": 2.4964,
       "step": 180
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 6.845513413506012e-05,
-      "loss": 2.235,
+      "epoch": 0.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.57076673615937e-06,
+      "loss": 2.5277,
       "step": 185
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 7.030527289546717e-05,
-      "loss": 2.2502,
+      "epoch": 0.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.802409080379893e-06,
+      "loss": 2.523,
       "step": 190
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 7.21554116558742e-05,
-      "loss": 2.2395,
+      "epoch": 0.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.034051424600418e-06,
+      "loss": 2.5444,
       "step": 195
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 7.400555041628122e-05,
-      "loss": 2.2683,
+      "epoch": 0.02,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.265693768820941e-06,
+      "loss": 2.5215,
       "step": 200
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 7.585568917668826e-05,
-      "loss": 2.2791,
+      "epoch": 0.02,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.497336113041465e-06,
+      "loss": 2.5169,
       "step": 205
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 7.770582793709528e-05,
-      "loss": 2.2839,
+      "epoch": 0.02,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.728978457261988e-06,
+      "loss": 2.4616,
       "step": 210
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 7.955596669750231e-05,
-      "loss": 2.282,
+      "epoch": 0.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.960620801482513e-06,
+      "loss": 2.5017,
       "step": 215
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1953125,
-      "learning_rate": 8.140610545790935e-05,
-      "loss": 2.2102,
+      "epoch": 0.02,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.0192263145703034e-05,
+      "loss": 2.4929,
       "step": 220
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.201171875,
-      "learning_rate": 8.325624421831638e-05,
-      "loss": 2.2292,
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.0423905489923559e-05,
+      "loss": 2.4526,
       "step": 225
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.189453125,
-      "learning_rate": 8.510638297872341e-05,
-      "loss": 2.2687,
+      "epoch": 0.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.0655547834144082e-05,
+      "loss": 2.4752,
       "step": 230
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 8.695652173913044e-05,
-      "loss": 2.2431,
+      "epoch": 0.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.0887190178364606e-05,
+      "loss": 2.447,
       "step": 235
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.203125,
-      "learning_rate": 8.880666049953746e-05,
-      "loss": 2.2333,
+      "epoch": 0.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1118832522585129e-05,
+      "loss": 2.4641,
       "step": 240
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 9.06567992599445e-05,
-      "loss": 2.2432,
+      "epoch": 0.02,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.1350474866805652e-05,
+      "loss": 2.4148,
       "step": 245
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 9.250693802035154e-05,
-      "loss": 2.2293,
+      "epoch": 0.02,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.1582117211026177e-05,
+      "loss": 2.4417,
       "step": 250
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.205078125,
-      "learning_rate": 9.435707678075855e-05,
-      "loss": 2.28,
+      "epoch": 0.02,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1813759555246699e-05,
+      "loss": 2.4551,
       "step": 255
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 9.62072155411656e-05,
-      "loss": 2.2034,
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.2045401899467223e-05,
+      "loss": 2.4278,
       "step": 260
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 9.805735430157262e-05,
-      "loss": 2.2733,
+      "epoch": 0.02,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.2277044243687747e-05,
+      "loss": 2.4301,
       "step": 265
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 9.990749306197965e-05,
-      "loss": 2.223,
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.250868658790827e-05,
+      "loss": 2.4305,
       "step": 270
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0001017576318223867,
-      "loss": 2.2366,
+      "epoch": 0.03,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.2740328932128795e-05,
+      "loss": 2.3581,
       "step": 275
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00010360777058279371,
-      "loss": 2.2828,
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.2971971276349318e-05,
+      "loss": 2.371,
       "step": 280
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00010545790934320075,
-      "loss": 2.2162,
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.3203613620569841e-05,
+      "loss": 2.4669,
       "step": 285
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00010730804810360778,
-      "loss": 2.229,
+      "epoch": 0.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.3435255964790363e-05,
+      "loss": 2.3959,
       "step": 290
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0001091581868640148,
-      "loss": 2.2279,
+      "epoch": 0.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.3666898309010886e-05,
+      "loss": 2.4037,
       "step": 295
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00011100832562442183,
-      "loss": 2.2605,
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.3898540653231413e-05,
+      "loss": 2.3724,
       "step": 300
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00011285846438482887,
-      "loss": 2.2645,
+      "epoch": 0.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.4130182997451934e-05,
+      "loss": 2.4055,
       "step": 305
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0001147086031452359,
-      "loss": 2.2575,
+      "epoch": 0.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4361825341672458e-05,
+      "loss": 2.3735,
       "step": 310
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00011655874190564294,
-      "loss": 2.2618,
+      "epoch": 0.03,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.459346768589298e-05,
+      "loss": 2.3289,
       "step": 315
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00011840888066604995,
-      "loss": 2.2295,
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.4825110030113506e-05,
+      "loss": 2.3823,
       "step": 320
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00012025901942645699,
-      "loss": 2.219,
+      "epoch": 0.03,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.5056752374334029e-05,
+      "loss": 2.3309,
       "step": 325
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00012210915818686402,
-      "loss": 2.2416,
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.5288394718554554e-05,
+      "loss": 2.4147,
       "step": 330
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00012395929694727104,
-      "loss": 2.1879,
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.5520037062775074e-05,
+      "loss": 2.3459,
       "step": 335
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00012580943570767808,
-      "loss": 2.2485,
+      "epoch": 0.03,
+      "grad_norm": 0.25,
+      "learning_rate": 1.57516794069956e-05,
+      "loss": 2.2881,
       "step": 340
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00012765957446808513,
-      "loss": 2.2153,
+      "epoch": 0.03,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.5983321751216124e-05,
+      "loss": 2.3361,
       "step": 345
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00012950971322849215,
-      "loss": 2.2566,
+      "epoch": 0.03,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.6214964095436647e-05,
+      "loss": 2.3493,
       "step": 350
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00013135985198889916,
-      "loss": 2.2282,
+      "epoch": 0.03,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.644660643965717e-05,
+      "loss": 2.3583,
       "step": 355
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0001332099907493062,
-      "loss": 2.2299,
+      "epoch": 0.03,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.6678248783877693e-05,
+      "loss": 2.3455,
       "step": 360
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00013506012950971323,
-      "loss": 2.2566,
+      "epoch": 0.03,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.6909891128098217e-05,
+      "loss": 2.3814,
       "step": 365
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00013691026827012025,
-      "loss": 2.2137,
+      "epoch": 0.03,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.714153347231874e-05,
+      "loss": 2.4082,
       "step": 370
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00013876040703052732,
-      "loss": 2.2106,
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.7373175816539263e-05,
+      "loss": 2.2791,
       "step": 375
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00014061054579093434,
-      "loss": 2.2085,
+      "epoch": 0.04,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.7604818160759786e-05,
+      "loss": 2.2882,
       "step": 380
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00014246068455134135,
-      "loss": 2.195,
+      "epoch": 0.04,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.7836460504980313e-05,
+      "loss": 2.3111,
       "step": 385
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0001443108233117484,
-      "loss": 2.2238,
+      "epoch": 0.04,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.8068102849200836e-05,
+      "loss": 2.3629,
       "step": 390
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00014616096207215542,
-      "loss": 2.2245,
+      "epoch": 0.04,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.829974519342136e-05,
+      "loss": 2.3043,
       "step": 395
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00014801110083256244,
-      "loss": 2.2254,
+      "epoch": 0.04,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.8531387537641883e-05,
+      "loss": 2.3207,
       "step": 400
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00014986123959296948,
-      "loss": 2.252,
+      "epoch": 0.04,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.8763029881862406e-05,
+      "loss": 2.35,
       "step": 405
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00015171137835337653,
-      "loss": 2.2084,
+      "epoch": 0.04,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.899467222608293e-05,
+      "loss": 2.3065,
       "step": 410
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00015356151711378355,
-      "loss": 2.2621,
+      "epoch": 0.04,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9226314570303452e-05,
+      "loss": 2.3007,
       "step": 415
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00015541165587419056,
-      "loss": 2.2136,
+      "epoch": 0.04,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.9457956914523976e-05,
+      "loss": 2.3214,
       "step": 420
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0001572617946345976,
-      "loss": 2.2319,
+      "epoch": 0.04,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.96895992587445e-05,
+      "loss": 2.293,
       "step": 425
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00015911193339500463,
-      "loss": 2.2309,
+      "epoch": 0.04,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.9921241602965025e-05,
+      "loss": 2.3052,
       "step": 430
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00016096207215541165,
-      "loss": 2.2266,
+      "epoch": 0.04,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.0152883947185545e-05,
+      "loss": 2.3307,
       "step": 435
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0001628122109158187,
-      "loss": 2.2047,
+      "epoch": 0.04,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.038452629140607e-05,
+      "loss": 2.2789,
       "step": 440
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00016466234967622574,
-      "loss": 2.2206,
+      "epoch": 0.04,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.061616863562659e-05,
+      "loss": 2.307,
       "step": 445
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00016651248843663275,
-      "loss": 2.1969,
+      "epoch": 0.04,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.0847810979847118e-05,
+      "loss": 2.3108,
       "step": 450
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0001683626271970398,
-      "loss": 2.219,
+      "epoch": 0.04,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.107945332406764e-05,
+      "loss": 2.2605,
       "step": 455
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00017021276595744682,
-      "loss": 2.191,
+      "epoch": 0.04,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.1311095668288165e-05,
+      "loss": 2.2822,
       "step": 460
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00017206290471785384,
-      "loss": 2.2323,
+      "epoch": 0.04,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1542738012508688e-05,
+      "loss": 2.2924,
       "step": 465
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00017391304347826088,
-      "loss": 2.2011,
+      "epoch": 0.04,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.177438035672921e-05,
+      "loss": 2.324,
       "step": 470
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0001757631822386679,
-      "loss": 2.1978,
+      "epoch": 0.04,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.2006022700949734e-05,
+      "loss": 2.2709,
       "step": 475
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00017761332099907492,
-      "loss": 2.2146,
+      "epoch": 0.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.2237665045170258e-05,
+      "loss": 2.2676,
       "step": 480
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.000179463459759482,
-      "loss": 2.2108,
+      "epoch": 0.04,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.246930738939078e-05,
+      "loss": 2.3367,
       "step": 485
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000181313598519889,
-      "loss": 2.2392,
+      "epoch": 0.05,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.2700949733611304e-05,
+      "loss": 2.278,
       "step": 490
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00018316373728029603,
-      "loss": 2.231,
+      "epoch": 0.05,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.293259207783183e-05,
+      "loss": 2.3137,
       "step": 495
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00018501387604070307,
-      "loss": 2.19,
+      "epoch": 0.05,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.3164234422052354e-05,
+      "loss": 2.2971,
       "step": 500
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0001868640148011101,
-      "loss": 2.2252,
+      "epoch": 0.05,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.3395876766272877e-05,
+      "loss": 2.3141,
       "step": 505
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0001887141535615171,
-      "loss": 2.2244,
+      "epoch": 0.05,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.3627519110493397e-05,
+      "loss": 2.2874,
       "step": 510
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00019056429232192415,
-      "loss": 2.1966,
+      "epoch": 0.05,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.3859161454713924e-05,
+      "loss": 2.2287,
       "step": 515
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0001924144310823312,
-      "loss": 2.2371,
+      "epoch": 0.05,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4090803798934447e-05,
+      "loss": 2.3199,
       "step": 520
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00019426456984273822,
-      "loss": 2.1947,
+      "epoch": 0.05,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.432244614315497e-05,
+      "loss": 2.262,
       "step": 525
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00019611470860314524,
-      "loss": 2.2009,
+      "epoch": 0.05,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.4554088487375493e-05,
+      "loss": 2.3342,
       "step": 530
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00019796484736355228,
-      "loss": 2.1746,
+      "epoch": 0.05,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4785730831596017e-05,
+      "loss": 2.3089,
       "step": 535
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0001998149861239593,
-      "loss": 2.1965,
+      "epoch": 0.05,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.501737317581654e-05,
+      "loss": 2.2909,
       "step": 540
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00020166512488436634,
-      "loss": 2.1786,
+      "epoch": 0.05,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.5249015520037063e-05,
+      "loss": 2.2878,
       "step": 545
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0002035152636447734,
-      "loss": 2.2412,
+      "epoch": 0.05,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.548065786425759e-05,
+      "loss": 2.2444,
       "step": 550
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002053654024051804,
-      "loss": 2.2405,
+      "epoch": 0.05,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.5712300208478113e-05,
+      "loss": 2.3005,
       "step": 555
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00020721554116558743,
-      "loss": 2.2099,
+      "epoch": 0.05,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.5943942552698636e-05,
+      "loss": 2.2211,
       "step": 560
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00020906567992599447,
-      "loss": 2.2487,
+      "epoch": 0.05,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.617558489691916e-05,
+      "loss": 2.2703,
       "step": 565
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0002109158186864015,
-      "loss": 2.2261,
+      "epoch": 0.05,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.6407227241139683e-05,
+      "loss": 2.2859,
       "step": 570
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0002127659574468085,
-      "loss": 2.2253,
+      "epoch": 0.05,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.6638869585360206e-05,
+      "loss": 2.1875,
       "step": 575
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00021461609620721555,
-      "loss": 2.2131,
+      "epoch": 0.05,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.6870511929580726e-05,
+      "loss": 2.2319,
       "step": 580
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00021646623496762257,
-      "loss": 2.2409,
+      "epoch": 0.05,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.710215427380125e-05,
+      "loss": 2.306,
       "step": 585
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0002183163737280296,
-      "loss": 2.2165,
+      "epoch": 0.05,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.7333796618021772e-05,
+      "loss": 2.2547,
       "step": 590
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00022016651248843664,
-      "loss": 2.2068,
+      "epoch": 0.06,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.7565438962242302e-05,
+      "loss": 2.255,
       "step": 595
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00022201665124884365,
-      "loss": 2.2435,
+      "epoch": 0.06,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.7797081306462826e-05,
+      "loss": 2.2705,
       "step": 600
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00022386679000925073,
-      "loss": 2.1959,
+      "epoch": 0.06,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.802872365068335e-05,
+      "loss": 2.2698,
       "step": 605
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00022571692876965774,
-      "loss": 2.2377,
+      "epoch": 0.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.826036599490387e-05,
+      "loss": 2.267,
       "step": 610
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002275670675300648,
-      "loss": 2.2028,
+      "epoch": 0.06,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.8492008339124392e-05,
+      "loss": 2.255,
       "step": 615
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002294172062904718,
-      "loss": 2.1899,
+      "epoch": 0.06,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.8723650683344915e-05,
+      "loss": 2.277,
       "step": 620
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00023126734505087883,
-      "loss": 2.2237,
+      "epoch": 0.06,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.895529302756544e-05,
+      "loss": 2.2649,
       "step": 625
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00023311748381128587,
-      "loss": 2.2715,
+      "epoch": 0.06,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.918693537178596e-05,
+      "loss": 2.2691,
       "step": 630
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0002349676225716929,
-      "loss": 2.2158,
+      "epoch": 0.06,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.941857771600649e-05,
+      "loss": 2.3182,
       "step": 635
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002368177613320999,
-      "loss": 2.2045,
+      "epoch": 0.06,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.965022006022701e-05,
+      "loss": 2.2879,
       "step": 640
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00023866790009250695,
-      "loss": 2.19,
+      "epoch": 0.06,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.9881862404447535e-05,
+      "loss": 2.2671,
       "step": 645
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00024051803885291397,
-      "loss": 2.175,
+      "epoch": 0.06,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.0113504748668058e-05,
+      "loss": 2.2477,
       "step": 650
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.000242368177613321,
-      "loss": 2.2398,
+      "epoch": 0.06,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.034514709288858e-05,
+      "loss": 2.303,
       "step": 655
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00024421831637372804,
-      "loss": 2.2246,
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.057678943710911e-05,
+      "loss": 2.2719,
       "step": 660
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0002460684551341351,
-      "loss": 2.2097,
+      "epoch": 0.06,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.0808431781329624e-05,
+      "loss": 2.2825,
       "step": 665
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00024791859389454207,
-      "loss": 2.1865,
+      "epoch": 0.06,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.104007412555015e-05,
+      "loss": 2.2941,
       "step": 670
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0002497687326549491,
-      "loss": 2.1929,
+      "epoch": 0.06,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.127171646977067e-05,
+      "loss": 2.1921,
       "step": 675
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00025161887141535616,
-      "loss": 2.223,
+      "epoch": 0.06,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.15033588139912e-05,
+      "loss": 2.2714,
       "step": 680
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0002534690101757632,
-      "loss": 2.1875,
+      "epoch": 0.06,
+      "grad_norm": 0.244140625,
+      "learning_rate": 3.1735001158211724e-05,
+      "loss": 2.3,
       "step": 685
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00025531914893617025,
-      "loss": 2.2754,
+      "epoch": 0.06,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.196664350243225e-05,
+      "loss": 2.2673,
       "step": 690
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.00025716928769657724,
-      "loss": 2.2107,
+      "epoch": 0.06,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.219828584665277e-05,
+      "loss": 2.2974,
       "step": 695
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.0002590194264569843,
-      "loss": 2.2189,
+      "epoch": 0.06,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.2429928190873294e-05,
+      "loss": 2.2699,
       "step": 700
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00026086956521739134,
-      "loss": 2.2028,
+      "epoch": 0.07,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.266157053509382e-05,
+      "loss": 2.316,
       "step": 705
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0002627197039777983,
-      "loss": 2.2138,
+      "epoch": 0.07,
+      "grad_norm": 0.232421875,
+      "learning_rate": 3.289321287931434e-05,
+      "loss": 2.2344,
       "step": 710
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00026456984273820537,
-      "loss": 2.2098,
+      "epoch": 0.07,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.312485522353486e-05,
+      "loss": 2.2931,
       "step": 715
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0002664199814986124,
-      "loss": 2.1761,
+      "epoch": 0.07,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.3356497567755387e-05,
+      "loss": 2.2358,
       "step": 720
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002682701202590194,
-      "loss": 2.1912,
+      "epoch": 0.07,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.3588139911975917e-05,
+      "loss": 2.2585,
       "step": 725
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.00027012025901942645,
-      "loss": 2.1687,
+      "epoch": 0.07,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.381978225619643e-05,
+      "loss": 2.2558,
       "step": 730
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0002719703977798335,
-      "loss": 2.2049,
+      "epoch": 0.07,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.4051424600416956e-05,
+      "loss": 2.2907,
       "step": 735
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1689453125,
-      "learning_rate": 0.0002738205365402405,
-      "loss": 2.1749,
+      "epoch": 0.07,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.428306694463748e-05,
+      "loss": 2.2164,
       "step": 740
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00027567067530064754,
-      "loss": 2.1944,
+      "epoch": 0.07,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.4514709288858e-05,
+      "loss": 2.2689,
       "step": 745
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.16796875,
-      "learning_rate": 0.00027752081406105464,
-      "loss": 2.206,
+      "epoch": 0.07,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.4746351633078526e-05,
+      "loss": 2.3152,
       "step": 750
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0002793709528214616,
-      "loss": 2.1873,
+      "epoch": 0.07,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.497799397729905e-05,
+      "loss": 2.2954,
       "step": 755
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00028122109158186867,
-      "loss": 2.2215,
+      "epoch": 0.07,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.520963632151957e-05,
+      "loss": 2.2745,
       "step": 760
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0002830712303422757,
-      "loss": 2.2269,
+      "epoch": 0.07,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.5441278665740096e-05,
+      "loss": 2.2902,
       "step": 765
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0002849213691026827,
-      "loss": 2.2421,
+      "epoch": 0.07,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.5672921009960626e-05,
+      "loss": 2.2466,
       "step": 770
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00028677150786308975,
-      "loss": 2.2153,
+      "epoch": 0.07,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.590456335418115e-05,
+      "loss": 2.3003,
       "step": 775
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.1689453125,
-      "learning_rate": 0.0002886216466234968,
-      "loss": 2.2253,
+      "epoch": 0.07,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.613620569840167e-05,
+      "loss": 2.2727,
       "step": 780
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.0002904717853839038,
-      "loss": 2.1929,
+      "epoch": 0.07,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.6367848042622195e-05,
+      "loss": 2.272,
       "step": 785
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00029232192414431084,
-      "loss": 2.2065,
+      "epoch": 0.07,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.659949038684272e-05,
+      "loss": 2.2654,
       "step": 790
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.0002941720629047179,
-      "loss": 2.2065,
+      "epoch": 0.07,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.683113273106324e-05,
+      "loss": 2.2957,
       "step": 795
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00029602220166512487,
-      "loss": 2.1999,
+      "epoch": 0.07,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.7062775075283765e-05,
+      "loss": 2.2905,
       "step": 800
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0002978723404255319,
-      "loss": 2.2407,
+      "epoch": 0.07,
+      "grad_norm": 0.25,
+      "learning_rate": 3.729441741950429e-05,
+      "loss": 2.2757,
       "step": 805
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.00029972247918593896,
-      "loss": 2.2366,
+      "epoch": 0.08,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.752605976372481e-05,
+      "loss": 2.2713,
       "step": 810
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.1669921875,
-      "learning_rate": 0.00030157261794634595,
-      "loss": 2.2105,
+      "epoch": 0.08,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.7757702107945335e-05,
+      "loss": 2.3101,
       "step": 815
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00030342275670675305,
-      "loss": 2.2145,
+      "epoch": 0.08,
+      "grad_norm": 0.25,
+      "learning_rate": 3.798934445216586e-05,
+      "loss": 2.2944,
       "step": 820
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00030527289546716004,
-      "loss": 2.1843,
+      "epoch": 0.08,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.822098679638638e-05,
+      "loss": 2.2471,
       "step": 825
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.0003071230342275671,
-      "loss": 2.1932,
+      "epoch": 0.08,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.8452629140606905e-05,
+      "loss": 2.279,
       "step": 830
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.00030897317298797414,
-      "loss": 2.1702,
+      "epoch": 0.08,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.868427148482743e-05,
+      "loss": 2.2432,
       "step": 835
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.0003108233117483811,
-      "loss": 2.1664,
+      "epoch": 0.08,
+      "grad_norm": 0.248046875,
+      "learning_rate": 3.891591382904795e-05,
+      "loss": 2.2741,
       "step": 840
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.00031267345050878817,
-      "loss": 2.2214,
+      "epoch": 0.08,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.9147556173268474e-05,
+      "loss": 2.2537,
       "step": 845
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.0003145235892691952,
-      "loss": 2.2186,
+      "epoch": 0.08,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.9379198517489e-05,
+      "loss": 2.2916,
       "step": 850
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.0003163737280296022,
-      "loss": 2.2511,
+      "epoch": 0.08,
+      "grad_norm": 0.244140625,
+      "learning_rate": 3.961084086170953e-05,
+      "loss": 2.2481,
       "step": 855
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00031822386679000925,
-      "loss": 2.225,
+      "epoch": 0.08,
+      "grad_norm": 0.25,
+      "learning_rate": 3.984248320593005e-05,
+      "loss": 2.276,
       "step": 860
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1689453125,
-      "learning_rate": 0.0003200740055504163,
-      "loss": 2.1627,
+      "epoch": 0.08,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.0074125550150574e-05,
+      "loss": 2.2527,
       "step": 865
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0003219241443108233,
-      "loss": 2.1866,
+      "epoch": 0.08,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.030576789437109e-05,
+      "loss": 2.3016,
       "step": 870
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00032377428307123034,
-      "loss": 2.2332,
+      "epoch": 0.08,
+      "grad_norm": 0.24609375,
+      "learning_rate": 4.0537410238591614e-05,
+      "loss": 2.2292,
       "step": 875
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.16796875,
-      "learning_rate": 0.0003256244218316374,
-      "loss": 2.2043,
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 4.076905258281214e-05,
+      "loss": 2.237,
       "step": 880
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00032747456059204437,
-      "loss": 2.1818,
+      "epoch": 0.08,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.100069492703266e-05,
+      "loss": 2.1882,
       "step": 885
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00032932469935245147,
-      "loss": 2.1872,
+      "epoch": 0.08,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.123233727125318e-05,
+      "loss": 2.2809,
       "step": 890
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003311748381128585,
-      "loss": 2.1946,
+      "epoch": 0.08,
+      "grad_norm": 0.24609375,
+      "learning_rate": 4.1463979615473707e-05,
+      "loss": 2.2748,
       "step": 895
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003330249768732655,
-      "loss": 2.2237,
+      "epoch": 0.08,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.1695621959694237e-05,
+      "loss": 2.2359,
       "step": 900
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00033487511563367255,
-      "loss": 2.2096,
+      "epoch": 0.08,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.192726430391476e-05,
+      "loss": 2.2351,
       "step": 905
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0003367252543940796,
-      "loss": 2.1838,
+      "epoch": 0.08,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.215890664813528e-05,
+      "loss": 2.193,
       "step": 910
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0003385753931544866,
-      "loss": 2.2126,
+      "epoch": 0.08,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.2390548992355806e-05,
+      "loss": 2.1983,
       "step": 915
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00034042553191489364,
-      "loss": 2.1749,
+      "epoch": 0.09,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.262219133657633e-05,
+      "loss": 2.2548,
       "step": 920
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.166015625,
-      "learning_rate": 0.0003422756706753007,
-      "loss": 2.1816,
+      "epoch": 0.09,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.285383368079685e-05,
+      "loss": 2.3173,
       "step": 925
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00034412580943570767,
-      "loss": 2.2064,
+      "epoch": 0.09,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.3085476025017376e-05,
+      "loss": 2.2156,
       "step": 930
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.0003459759481961147,
-      "loss": 2.1977,
+      "epoch": 0.09,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.33171183692379e-05,
+      "loss": 2.2295,
       "step": 935
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00034782608695652176,
-      "loss": 2.2267,
+      "epoch": 0.09,
+      "grad_norm": 0.25,
+      "learning_rate": 4.354876071345842e-05,
+      "loss": 2.2386,
       "step": 940
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.00034967622571692875,
-      "loss": 2.1538,
+      "epoch": 0.09,
+      "grad_norm": 0.25,
+      "learning_rate": 4.3780403057678946e-05,
+      "loss": 2.2298,
       "step": 945
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003515263644773358,
-      "loss": 2.2294,
+      "epoch": 0.09,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.401204540189947e-05,
+      "loss": 2.2697,
       "step": 950
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00035337650323774284,
-      "loss": 2.2237,
+      "epoch": 0.09,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.424368774611999e-05,
+      "loss": 2.277,
       "step": 955
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00035522664199814984,
-      "loss": 2.2067,
+      "epoch": 0.09,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.4475330090340515e-05,
+      "loss": 2.2493,
       "step": 960
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00035707678075855694,
-      "loss": 2.18,
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.470697243456104e-05,
+      "loss": 2.2649,
       "step": 965
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.000358926919518964,
-      "loss": 2.2009,
+      "epoch": 0.09,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.493861477878156e-05,
+      "loss": 2.2728,
       "step": 970
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00036077705827937097,
-      "loss": 2.2513,
+      "epoch": 0.09,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.5170257123002085e-05,
+      "loss": 2.2355,
       "step": 975
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.000362627197039778,
-      "loss": 2.1711,
+      "epoch": 0.09,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.540189946722261e-05,
+      "loss": 2.2788,
       "step": 980
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00036447733580018506,
-      "loss": 2.2227,
+      "epoch": 0.09,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.563354181144313e-05,
+      "loss": 2.2246,
       "step": 985
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.00036632747456059205,
-      "loss": 2.181,
+      "epoch": 0.09,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.586518415566366e-05,
+      "loss": 2.2562,
       "step": 990
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0003681776133209991,
-      "loss": 2.2341,
+      "epoch": 0.09,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.6096826499884185e-05,
+      "loss": 2.2272,
       "step": 995
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1650390625,
-      "learning_rate": 0.00037002775208140614,
-      "loss": 2.1537,
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.632846884410471e-05,
+      "loss": 2.237,
       "step": 1000
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1689453125,
-      "learning_rate": 0.00037187789084181314,
-      "loss": 2.2413,
+      "epoch": 0.09,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.656011118832523e-05,
+      "loss": 2.2229,
       "step": 1005
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003737280296022202,
-      "loss": 2.2119,
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.6791753532545755e-05,
+      "loss": 2.2127,
       "step": 1010
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0003755781683626272,
-      "loss": 2.2427,
+      "epoch": 0.09,
+      "grad_norm": 0.240234375,
+      "learning_rate": 4.702339587676627e-05,
+      "loss": 2.2435,
       "step": 1015
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.169921875,
-      "learning_rate": 0.0003774283071230342,
-      "loss": 2.2112,
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 4.7255038220986794e-05,
+      "loss": 2.2293,
       "step": 1020
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00037927844588344126,
-      "loss": 2.1871,
+      "epoch": 0.09,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.748668056520732e-05,
+      "loss": 2.2203,
       "step": 1025
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003811285846438483,
-      "loss": 2.1889,
+      "epoch": 0.1,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.771832290942785e-05,
+      "loss": 2.2288,
       "step": 1030
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00038297872340425535,
-      "loss": 2.1728,
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.794996525364837e-05,
+      "loss": 2.205,
       "step": 1035
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0003848288621646624,
-      "loss": 2.1774,
+      "epoch": 0.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.8181607597868894e-05,
+      "loss": 2.2207,
       "step": 1040
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0003866790009250694,
-      "loss": 2.2112,
+      "epoch": 0.1,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.841324994208942e-05,
+      "loss": 2.2315,
       "step": 1045
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00038852913968547644,
-      "loss": 2.2181,
+      "epoch": 0.1,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.864489228630994e-05,
+      "loss": 2.2299,
       "step": 1050
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0003903792784458835,
-      "loss": 2.1747,
+      "epoch": 0.1,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.8876534630530464e-05,
+      "loss": 2.2181,
       "step": 1055
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.171875,
-      "learning_rate": 0.00039222941720629047,
-      "loss": 2.2193,
+      "epoch": 0.1,
+      "grad_norm": 0.25,
+      "learning_rate": 4.910817697475099e-05,
+      "loss": 2.3207,
       "step": 1060
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003940795559666975,
-      "loss": 2.1631,
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.933981931897151e-05,
+      "loss": 2.2336,
       "step": 1065
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.00039592969472710456,
-      "loss": 2.1777,
+      "epoch": 0.1,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.9571461663192033e-05,
+      "loss": 2.219,
       "step": 1070
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.00039777983348751155,
-      "loss": 2.2187,
+      "epoch": 0.1,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.980310400741256e-05,
+      "loss": 2.2644,
       "step": 1075
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003996299722479186,
-      "loss": 2.1994,
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.003474635163308e-05,
+      "loss": 2.2397,
       "step": 1080
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003999998330292309,
-      "loss": 2.226,
+      "epoch": 0.1,
+      "grad_norm": 0.2734375,
+      "learning_rate": 5.02663886958536e-05,
+      "loss": 2.217,
       "step": 1085
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039999915471095927,
-      "loss": 2.2093,
+      "epoch": 0.1,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.0498031040074126e-05,
+      "loss": 2.2862,
       "step": 1090
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.0003999979546112803,
-      "loss": 2.1598,
+      "epoch": 0.1,
+      "grad_norm": 0.25,
+      "learning_rate": 5.072967338429465e-05,
+      "loss": 2.1939,
       "step": 1095
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039999623273332497,
-      "loss": 2.1882,
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.096131572851518e-05,
+      "loss": 2.2207,
       "step": 1100
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039999398908158553,
-      "loss": 2.1787,
+      "epoch": 0.1,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.1192958072735696e-05,
+      "loss": 2.259,
       "step": 1105
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00039999122366191547,
-      "loss": 2.2214,
+      "epoch": 0.1,
+      "grad_norm": 0.24609375,
+      "learning_rate": 5.1424600416956226e-05,
+      "loss": 2.2305,
       "step": 1110
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039998793648152956,
-      "loss": 2.2272,
+      "epoch": 0.1,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.165624276117674e-05,
+      "loss": 2.2457,
       "step": 1115
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003999841275490038,
-      "loss": 2.1692,
+      "epoch": 0.1,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.188788510539727e-05,
+      "loss": 2.2631,
       "step": 1120
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039997979687427535,
-      "loss": 2.1905,
+      "epoch": 0.1,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.211952744961779e-05,
+      "loss": 2.2347,
       "step": 1125
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039997494446864255,
-      "loss": 2.2032,
+      "epoch": 0.1,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.235116979383832e-05,
+      "loss": 2.2145,
       "step": 1130
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00039996957034476503,
-      "loss": 2.203,
+      "epoch": 0.11,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.2582812138058835e-05,
+      "loss": 2.2654,
       "step": 1135
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039996367451666337,
-      "loss": 2.1985,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.2814454482279365e-05,
+      "loss": 2.2535,
       "step": 1140
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039995725699971935,
-      "loss": 2.1622,
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.304609682649989e-05,
+      "loss": 2.229,
       "step": 1145
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 0.00039995031781067566,
-      "loss": 2.1439,
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.327773917072041e-05,
+      "loss": 2.2414,
       "step": 1150
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039994285696763615,
-      "loss": 2.2177,
+      "epoch": 0.11,
+      "grad_norm": 0.265625,
+      "learning_rate": 5.3509381514940935e-05,
+      "loss": 2.1534,
       "step": 1155
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003999348744900655,
-      "loss": 2.183,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.374102385916145e-05,
+      "loss": 2.2552,
       "step": 1160
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003999263703987894,
-      "loss": 2.1907,
+      "epoch": 0.11,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 5.397266620338198e-05,
+      "loss": 2.1484,
       "step": 1165
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0003999173447159943,
-      "loss": 2.1811,
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.42043085476025e-05,
+      "loss": 2.2861,
       "step": 1170
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.00039990779746522744,
-      "loss": 2.2091,
+      "epoch": 0.11,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.443595089182303e-05,
+      "loss": 2.2669,
       "step": 1175
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00039989772867139685,
-      "loss": 2.1883,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.4667593236043545e-05,
+      "loss": 2.2294,
       "step": 1180
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039988713836077117,
-      "loss": 2.1967,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.4899235580264075e-05,
+      "loss": 2.2626,
       "step": 1185
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039987602656097967,
-      "loss": 2.2445,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.5130877924484605e-05,
+      "loss": 2.2662,
       "step": 1190
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039986439330101203,
-      "loss": 2.1954,
+      "epoch": 0.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 5.536252026870512e-05,
+      "loss": 2.2233,
       "step": 1195
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039985223861121864,
-      "loss": 2.1925,
+      "epoch": 0.11,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.559416261292565e-05,
+      "loss": 2.252,
       "step": 1200
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.00039983956252331,
+      "epoch": 0.11,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.582580495714617e-05,
       "loss": 2.1998,
       "step": 1205
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.0003998263650703569,
-      "loss": 2.1904,
+      "epoch": 0.11,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.60574473013667e-05,
+      "loss": 2.2492,
       "step": 1210
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003998126462867906,
-      "loss": 2.1962,
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.6289089645587214e-05,
+      "loss": 2.2195,
       "step": 1215
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039979840620840213,
-      "loss": 2.1577,
+      "epoch": 0.11,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.652073198980774e-05,
+      "loss": 2.2243,
       "step": 1220
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003997836448723427,
-      "loss": 2.1957,
+      "epoch": 0.11,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.675237433402826e-05,
+      "loss": 2.2604,
       "step": 1225
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0003997683623171236,
-      "loss": 2.2043,
+      "epoch": 0.11,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.6984016678248784e-05,
+      "loss": 2.2104,
       "step": 1230
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0003997525585826155,
-      "loss": 2.1525,
+      "epoch": 0.11,
+      "grad_norm": 0.25,
+      "learning_rate": 5.7215659022469314e-05,
+      "loss": 2.2232,
       "step": 1235
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039973623371004923,
-      "loss": 2.1551,
+      "epoch": 0.11,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.744730136668983e-05,
+      "loss": 2.2872,
       "step": 1240
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.000399719387742015,
-      "loss": 2.1806,
+      "epoch": 0.12,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.767894371091036e-05,
+      "loss": 2.2722,
       "step": 1245
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039970202072246263,
-      "loss": 2.1651,
+      "epoch": 0.12,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.791058605513088e-05,
+      "loss": 2.219,
       "step": 1250
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003996841326967011,
-      "loss": 2.1923,
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.814222839935141e-05,
+      "loss": 2.2631,
       "step": 1255
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039966572371139886,
-      "loss": 2.2101,
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.837387074357192e-05,
+      "loss": 2.2438,
       "step": 1260
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003996467938145835,
-      "loss": 2.1676,
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.860551308779245e-05,
+      "loss": 2.2906,
       "step": 1265
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.0003996273430556415,
-      "loss": 2.1938,
+      "epoch": 0.12,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.883715543201298e-05,
+      "loss": 2.2086,
       "step": 1270
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0003996073714853184,
-      "loss": 2.1765,
+      "epoch": 0.12,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.90687977762335e-05,
+      "loss": 2.2441,
       "step": 1275
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039958687915571823,
-      "loss": 2.1636,
+      "epoch": 0.12,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.930044012045402e-05,
+      "loss": 2.2506,
       "step": 1280
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.000399565866120304,
-      "loss": 2.208,
+      "epoch": 0.12,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.9532082464674546e-05,
+      "loss": 2.2248,
       "step": 1285
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039954433243389676,
-      "loss": 2.1786,
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.976372480889507e-05,
+      "loss": 2.2479,
       "step": 1290
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.0003995222781526764,
-      "loss": 2.1826,
+      "epoch": 0.12,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.999536715311559e-05,
+      "loss": 2.2061,
       "step": 1295
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003994997033341805,
-      "loss": 2.1526,
+      "epoch": 0.12,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.0227009497336116e-05,
+      "loss": 2.2389,
       "step": 1300
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.000399476608037305,
-      "loss": 2.1645,
+      "epoch": 0.12,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.045865184155663e-05,
+      "loss": 2.2206,
       "step": 1305
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003994529923223036,
-      "loss": 2.2188,
+      "epoch": 0.12,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.069029418577716e-05,
+      "loss": 2.2163,
       "step": 1310
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003994288562507879,
-      "loss": 2.201,
+      "epoch": 0.12,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.092193652999769e-05,
+      "loss": 2.2314,
       "step": 1315
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039940419988572685,
-      "loss": 2.2056,
+      "epoch": 0.12,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.115357887421822e-05,
+      "loss": 2.1981,
       "step": 1320
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003993790232914468,
-      "loss": 2.1827,
+      "epoch": 0.12,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.138522121843873e-05,
+      "loss": 2.2286,
       "step": 1325
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039935332653363154,
-      "loss": 2.1834,
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.161686356265925e-05,
+      "loss": 2.1782,
       "step": 1330
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039932710967932177,
-      "loss": 2.1927,
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.184850590687978e-05,
+      "loss": 2.2228,
       "step": 1335
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0003993003727969151,
-      "loss": 2.1535,
+      "epoch": 0.12,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.20801482511003e-05,
+      "loss": 2.1949,
       "step": 1340
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00039927311595616585,
-      "loss": 2.1653,
+      "epoch": 0.12,
+      "grad_norm": 0.25,
+      "learning_rate": 6.231179059532082e-05,
+      "loss": 2.1905,
       "step": 1345
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039924533922818493,
-      "loss": 2.2489,
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.254343293954134e-05,
+      "loss": 2.2335,
       "step": 1350
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039921704268543954,
-      "loss": 2.1557,
+      "epoch": 0.13,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.277507528376187e-05,
+      "loss": 2.2337,
       "step": 1355
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.000399188226401753,
-      "loss": 2.2094,
+      "epoch": 0.13,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.30067176279824e-05,
+      "loss": 2.2387,
       "step": 1360
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003991588904523047,
-      "loss": 2.1912,
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.323835997220292e-05,
+      "loss": 2.2701,
       "step": 1365
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0003991290349136297,
-      "loss": 2.1823,
+      "epoch": 0.13,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.347000231642345e-05,
+      "loss": 2.2506,
       "step": 1370
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003990986598636186,
-      "loss": 2.1842,
+      "epoch": 0.13,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.370164466064396e-05,
+      "loss": 2.2421,
       "step": 1375
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039906776538151746,
-      "loss": 2.1882,
+      "epoch": 0.13,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.39332870048645e-05,
+      "loss": 2.2232,
       "step": 1380
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039903635154792745,
-      "loss": 2.1994,
+      "epoch": 0.13,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.416492934908501e-05,
+      "loss": 2.2508,
       "step": 1385
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003990044184448047,
-      "loss": 2.1443,
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.439657169330554e-05,
+      "loss": 2.2692,
       "step": 1390
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003989719661554601,
-      "loss": 2.1881,
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.462821403752606e-05,
+      "loss": 2.2538,
       "step": 1395
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039893899476455886,
-      "loss": 2.2267,
+      "epoch": 0.13,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.485985638174659e-05,
+      "loss": 2.1792,
       "step": 1400
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.00039890550435812077,
-      "loss": 2.184,
+      "epoch": 0.13,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.509149872596712e-05,
+      "loss": 2.2675,
       "step": 1405
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039887149502351954,
-      "loss": 2.2021,
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.532314107018763e-05,
+      "loss": 2.2567,
       "step": 1410
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003988369668494827,
-      "loss": 2.189,
+      "epoch": 0.13,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.555478341440816e-05,
+      "loss": 2.2394,
       "step": 1415
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003988019199260915,
-      "loss": 2.1912,
+      "epoch": 0.13,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.578642575862868e-05,
+      "loss": 2.2586,
       "step": 1420
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039876635434478046,
-      "loss": 2.1916,
+      "epoch": 0.13,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.601806810284921e-05,
+      "loss": 2.2046,
       "step": 1425
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039873027019833725,
-      "loss": 2.177,
+      "epoch": 0.13,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.624971044706973e-05,
+      "loss": 2.1925,
       "step": 1430
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003986936675809024,
-      "loss": 2.2042,
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.648135279129026e-05,
+      "loss": 2.2581,
       "step": 1435
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003986565465879692,
-      "loss": 2.1839,
+      "epoch": 0.13,
+      "grad_norm": 0.2734375,
+      "learning_rate": 6.671299513551077e-05,
+      "loss": 2.2016,
       "step": 1440
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0003986189073163833,
-      "loss": 2.187,
+      "epoch": 0.13,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.69446374797313e-05,
+      "loss": 2.1903,
       "step": 1445
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003985807498643424,
-      "loss": 2.1903,
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.717627982395183e-05,
+      "loss": 2.1747,
       "step": 1450
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039854207433139623,
-      "loss": 2.2007,
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.740792216817235e-05,
+      "loss": 2.2051,
       "step": 1455
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039850288081844614,
-      "loss": 2.1817,
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.763956451239287e-05,
+      "loss": 2.2427,
       "step": 1460
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003984631694277446,
-      "loss": 2.1768,
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.78712068566134e-05,
+      "loss": 2.215,
       "step": 1465
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1767578125,
-      "learning_rate": 0.0003984229402628956,
-      "loss": 2.1591,
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.810284920083391e-05,
+      "loss": 2.2627,
       "step": 1470
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003983821934288535,
-      "loss": 2.2008,
+      "epoch": 0.14,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.833449154505443e-05,
+      "loss": 2.2784,
       "step": 1475
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003983409290319237,
-      "loss": 2.202,
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.856613388927496e-05,
+      "loss": 2.2429,
       "step": 1480
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039829914717976156,
-      "loss": 2.1906,
+      "epoch": 0.14,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.879777623349548e-05,
+      "loss": 2.2085,
       "step": 1485
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003982568479813724,
-      "loss": 2.1852,
+      "epoch": 0.14,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.9029418577716e-05,
+      "loss": 2.2074,
       "step": 1490
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039821403154711136,
-      "loss": 2.1686,
+      "epoch": 0.14,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.926106092193654e-05,
+      "loss": 2.2188,
       "step": 1495
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.0003981706979886831,
-      "loss": 2.1896,
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.949270326615705e-05,
+      "loss": 2.1995,
       "step": 1500
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039812684741914137,
-      "loss": 2.1638,
+      "epoch": 0.14,
+      "grad_norm": 0.27734375,
+      "learning_rate": 6.972434561037758e-05,
+      "loss": 2.247,
       "step": 1505
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003980824799528886,
-      "loss": 2.1888,
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 6.99559879545981e-05,
+      "loss": 2.2431,
       "step": 1510
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003980375957056759,
-      "loss": 2.1733,
+      "epoch": 0.14,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.018763029881863e-05,
+      "loss": 2.1659,
       "step": 1515
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.173828125,
-      "learning_rate": 0.0003979921947946026,
-      "loss": 2.1795,
+      "epoch": 0.14,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.041927264303914e-05,
+      "loss": 2.2411,
       "step": 1520
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039794627733811595,
-      "loss": 2.1668,
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.065091498725967e-05,
+      "loss": 2.2206,
       "step": 1525
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003978998434560109,
-      "loss": 2.2003,
+      "epoch": 0.14,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.088255733148019e-05,
+      "loss": 2.2531,
       "step": 1530
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039785289326942953,
-      "loss": 2.18,
+      "epoch": 0.14,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.111419967570072e-05,
+      "loss": 2.208,
       "step": 1535
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003978054269008611,
-      "loss": 2.1842,
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.134584201992125e-05,
+      "loss": 2.2403,
       "step": 1540
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.17578125,
-      "learning_rate": 0.0003977574444741415,
-      "loss": 2.1539,
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.157748436414177e-05,
+      "loss": 2.2139,
       "step": 1545
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039770894611445293,
-      "loss": 2.182,
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.18091267083623e-05,
+      "loss": 2.2306,
       "step": 1550
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0003976599319483236,
-      "loss": 2.1939,
+      "epoch": 0.14,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.204076905258281e-05,
+      "loss": 2.2054,
       "step": 1555
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1728515625,
-      "learning_rate": 0.0003976104021036275,
-      "loss": 2.191,
+      "epoch": 0.14,
+      "grad_norm": 0.25,
+      "learning_rate": 7.227241139680334e-05,
+      "loss": 2.236,
       "step": 1560
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039756035670958384,
-      "loss": 2.1724,
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.250405374102386e-05,
+      "loss": 2.2136,
       "step": 1565
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003975097958967571,
-      "loss": 2.1847,
+      "epoch": 0.15,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.273569608524439e-05,
+      "loss": 2.188,
       "step": 1570
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.177734375,
-      "learning_rate": 0.00039745871979705616,
-      "loss": 2.1144,
+      "epoch": 0.15,
+      "grad_norm": 0.2734375,
+      "learning_rate": 7.296733842946491e-05,
+      "loss": 2.1902,
       "step": 1575
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003974071285437345,
-      "loss": 2.1888,
+      "epoch": 0.15,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.319898077368544e-05,
+      "loss": 2.1739,
       "step": 1580
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003973550222713894,
-      "loss": 2.1943,
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.343062311790597e-05,
+      "loss": 2.2637,
       "step": 1585
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003973024011159618,
-      "loss": 2.2067,
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.366226546212648e-05,
+      "loss": 2.2033,
       "step": 1590
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003972492652147361,
-      "loss": 2.1436,
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.3893907806347e-05,
+      "loss": 2.154,
       "step": 1595
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003971956147063395,
-      "loss": 2.2443,
+      "epoch": 0.15,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.412555015056753e-05,
+      "loss": 2.2276,
       "step": 1600
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003971414497307417,
-      "loss": 2.1936,
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.435719249478805e-05,
+      "loss": 2.1886,
       "step": 1605
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039708677042925473,
-      "loss": 2.173,
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.458883483900858e-05,
+      "loss": 2.235,
       "step": 1610
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.00039703157694453246,
-      "loss": 2.1894,
+      "epoch": 0.15,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.482047718322909e-05,
+      "loss": 2.2307,
       "step": 1615
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003969758694205701,
-      "loss": 2.1947,
+      "epoch": 0.15,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.505211952744962e-05,
+      "loss": 2.2887,
       "step": 1620
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.000396919648002704,
-      "loss": 2.1641,
+      "epoch": 0.15,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.528376187167014e-05,
+      "loss": 2.2161,
       "step": 1625
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039686291283761127,
-      "loss": 2.1706,
+      "epoch": 0.15,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.551540421589067e-05,
+      "loss": 2.1612,
       "step": 1630
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003968056640733092,
-      "loss": 2.1612,
+      "epoch": 0.15,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.574704656011119e-05,
+      "loss": 2.1952,
       "step": 1635
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039674790185915536,
-      "loss": 2.1836,
+      "epoch": 0.15,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.597868890433172e-05,
+      "loss": 2.2389,
       "step": 1640
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003966896263458465,
-      "loss": 2.1739,
+      "epoch": 0.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.621033124855223e-05,
+      "loss": 2.2284,
       "step": 1645
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003966308376854186,
-      "loss": 2.2084,
+      "epoch": 0.15,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.644197359277276e-05,
+      "loss": 2.2425,
       "step": 1650
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003965715360312466,
-      "loss": 2.1563,
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.667361593699328e-05,
+      "loss": 2.2083,
       "step": 1655
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003965117215380435,
-      "loss": 2.2207,
+      "epoch": 0.15,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.690525828121381e-05,
+      "loss": 2.2459,
       "step": 1660
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039645139436186047,
-      "loss": 2.1985,
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.713690062543434e-05,
+      "loss": 2.2322,
       "step": 1665
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003963905546600862,
-      "loss": 2.201,
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.736854296965486e-05,
+      "loss": 2.238,
       "step": 1670
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00039632920259144645,
-      "loss": 2.185,
+      "epoch": 0.16,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.760018531387539e-05,
+      "loss": 2.2325,
       "step": 1675
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003962673383160038,
-      "loss": 2.1761,
+      "epoch": 0.16,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.78318276580959e-05,
+      "loss": 2.2004,
       "step": 1680
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003962049619951571,
-      "loss": 2.1932,
+      "epoch": 0.16,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.806347000231643e-05,
+      "loss": 2.2026,
       "step": 1685
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039614207379164086,
-      "loss": 2.2029,
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.829511234653695e-05,
+      "loss": 2.2512,
       "step": 1690
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003960786738695255,
-      "loss": 2.2166,
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.852675469075748e-05,
+      "loss": 2.215,
       "step": 1695
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039601476239421604,
-      "loss": 2.1965,
+      "epoch": 0.16,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.8758397034978e-05,
+      "loss": 2.2218,
       "step": 1700
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003959503395324524,
-      "loss": 2.1923,
+      "epoch": 0.16,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.899003937919852e-05,
+      "loss": 2.2774,
       "step": 1705
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00039588540545230845,
-      "loss": 2.1747,
+      "epoch": 0.16,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.922168172341905e-05,
+      "loss": 2.2411,
       "step": 1710
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039581996032319195,
-      "loss": 2.1549,
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.945332406763957e-05,
+      "loss": 2.2283,
       "step": 1715
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.0003957540043158439,
-      "loss": 2.1445,
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.96849664118601e-05,
+      "loss": 2.2495,
       "step": 1720
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039568753760233806,
-      "loss": 2.1719,
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.991660875608062e-05,
+      "loss": 2.2103,
       "step": 1725
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00039562056035608067,
-      "loss": 2.1836,
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.014825110030115e-05,
+      "loss": 2.1846,
       "step": 1730
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003955530727518099,
-      "loss": 2.1984,
+      "epoch": 0.16,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.037989344452166e-05,
+      "loss": 2.2528,
       "step": 1735
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003954850749655953,
-      "loss": 2.1842,
+      "epoch": 0.16,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.061153578874218e-05,
+      "loss": 2.2186,
       "step": 1740
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039541656717483753,
-      "loss": 2.1859,
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.084317813296271e-05,
+      "loss": 2.2235,
       "step": 1745
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003953475495582679,
-      "loss": 2.1949,
+      "epoch": 0.16,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.107482047718323e-05,
+      "loss": 2.1931,
       "step": 1750
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039527802229594763,
-      "loss": 2.1892,
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.130646282140376e-05,
+      "loss": 2.2215,
       "step": 1755
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.00039520798556926776,
-      "loss": 2.1656,
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.153810516562427e-05,
+      "loss": 2.2202,
       "step": 1760
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003951374395609483,
-      "loss": 2.1715,
+      "epoch": 0.16,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.17697475098448e-05,
+      "loss": 2.2017,
       "step": 1765
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039506638445503786,
-      "loss": 2.219,
+      "epoch": 0.16,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.200138985406532e-05,
+      "loss": 2.2078,
       "step": 1770
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00039499482043691363,
-      "loss": 2.1862,
+      "epoch": 0.16,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.223303219828585e-05,
+      "loss": 2.2179,
       "step": 1775
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039492274769328014,
-      "loss": 2.1613,
+      "epoch": 0.16,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.246467454250637e-05,
+      "loss": 2.2147,
       "step": 1780
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00039485016641216934,
-      "loss": 2.181,
+      "epoch": 0.17,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.26963168867269e-05,
+      "loss": 2.1889,
       "step": 1785
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003947770767829398,
-      "loss": 2.1521,
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.292795923094741e-05,
+      "loss": 2.2647,
       "step": 1790
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00039470347899627634,
-      "loss": 2.1948,
+      "epoch": 0.17,
+      "grad_norm": 0.25,
+      "learning_rate": 8.315960157516794e-05,
+      "loss": 2.2182,
       "step": 1795
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003946293732441896,
-      "loss": 2.1656,
+      "epoch": 0.17,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.339124391938847e-05,
+      "loss": 2.2501,
       "step": 1800
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039455475972001544,
-      "loss": 2.1913,
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.362288626360899e-05,
+      "loss": 2.1926,
       "step": 1805
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003944796386184144,
-      "loss": 2.198,
+      "epoch": 0.17,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.385452860782952e-05,
+      "loss": 2.1765,
       "step": 1810
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003944040101353713,
-      "loss": 2.1771,
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.408617095205004e-05,
+      "loss": 2.193,
       "step": 1815
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003943278744681945,
-      "loss": 2.2192,
+      "epoch": 0.17,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.431781329627057e-05,
+      "loss": 2.2842,
       "step": 1820
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.000394251231815516,
-      "loss": 2.196,
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.454945564049108e-05,
+      "loss": 2.2242,
       "step": 1825
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039417408237728995,
-      "loss": 2.1658,
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.478109798471161e-05,
+      "loss": 2.2659,
       "step": 1830
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003940964263547931,
-      "loss": 2.1856,
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.501274032893213e-05,
+      "loss": 2.1986,
       "step": 1835
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.00039401826395062356,
-      "loss": 2.1597,
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.524438267315266e-05,
+      "loss": 2.2191,
       "step": 1840
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003939395953687006,
-      "loss": 2.1543,
+      "epoch": 0.17,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.547602501737319e-05,
+      "loss": 2.226,
       "step": 1845
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003938604208142642,
-      "loss": 2.213,
+      "epoch": 0.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 8.57076673615937e-05,
+      "loss": 2.2502,
       "step": 1850
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00039378074049387416,
-      "loss": 2.1782,
+      "epoch": 0.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.593930970581424e-05,
+      "loss": 2.1724,
       "step": 1855
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039370055461541,
-      "loss": 2.1538,
+      "epoch": 0.17,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.617095205003475e-05,
+      "loss": 2.2273,
       "step": 1860
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039361986338807013,
-      "loss": 2.1781,
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.640259439425528e-05,
+      "loss": 2.1974,
       "step": 1865
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1748046875,
-      "learning_rate": 0.00039353866702237123,
-      "loss": 2.1624,
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.66342367384758e-05,
+      "loss": 2.2581,
       "step": 1870
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003934569657301482,
-      "loss": 2.1507,
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.686587908269633e-05,
+      "loss": 2.2007,
       "step": 1875
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039337475972455274,
-      "loss": 2.2238,
+      "epoch": 0.17,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.709752142691684e-05,
+      "loss": 2.2613,
       "step": 1880
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003932920492200537,
-      "loss": 2.1408,
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.732916377113736e-05,
+      "loss": 2.1954,
       "step": 1885
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039320883443243605,
-      "loss": 2.1932,
+      "epoch": 0.18,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.756080611535789e-05,
+      "loss": 2.233,
       "step": 1890
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00039312511557880025,
-      "loss": 2.193,
+      "epoch": 0.18,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.779244845957841e-05,
+      "loss": 2.2335,
       "step": 1895
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003930408928775621,
-      "loss": 2.2054,
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.802409080379894e-05,
+      "loss": 2.2563,
       "step": 1900
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00039295616654845154,
-      "loss": 2.1553,
+      "epoch": 0.18,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.825573314801945e-05,
+      "loss": 2.2321,
       "step": 1905
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1796875,
-      "learning_rate": 0.00039287093681251273,
-      "loss": 2.2483,
+      "epoch": 0.18,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.848737549223998e-05,
+      "loss": 2.2262,
       "step": 1910
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039278520389210306,
-      "loss": 2.2321,
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.87190178364605e-05,
+      "loss": 2.2352,
       "step": 1915
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039269896801089266,
-      "loss": 2.2265,
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.895066018068103e-05,
+      "loss": 2.1554,
       "step": 1920
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003926122293938639,
-      "loss": 2.1578,
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.918230252490155e-05,
+      "loss": 2.2027,
       "step": 1925
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1806640625,
-      "learning_rate": 0.0003925249882673107,
-      "loss": 2.1928,
+      "epoch": 0.18,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.941394486912208e-05,
+      "loss": 2.1676,
       "step": 1930
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.000392437244858838,
-      "loss": 2.1682,
+      "epoch": 0.18,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.964558721334261e-05,
+      "loss": 2.2288,
       "step": 1935
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003923489993973612,
-      "loss": 2.219,
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.987722955756312e-05,
+      "loss": 2.1773,
       "step": 1940
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003922602521131055,
-      "loss": 2.2151,
+      "epoch": 0.18,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.010887190178365e-05,
+      "loss": 2.2169,
       "step": 1945
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00039217100323760525,
-      "loss": 2.1969,
+      "epoch": 0.18,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.034051424600417e-05,
+      "loss": 2.1887,
       "step": 1950
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.00039208125300370345,
-      "loss": 2.1983,
+      "epoch": 0.18,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.05721565902247e-05,
+      "loss": 2.1821,
       "step": 1955
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039199100164555116,
-      "loss": 2.164,
+      "epoch": 0.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.080379893444522e-05,
+      "loss": 2.2021,
       "step": 1960
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00039190024939860673,
-      "loss": 2.1929,
+      "epoch": 0.18,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.103544127866575e-05,
+      "loss": 2.249,
       "step": 1965
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00039180899649963536,
-      "loss": 2.1742,
+      "epoch": 0.18,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.126708362288626e-05,
+      "loss": 2.2102,
       "step": 1970
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003917172431867083,
-      "loss": 2.1718,
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.14987259671068e-05,
+      "loss": 2.2002,
       "step": 1975
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00039162498969920256,
-      "loss": 2.144,
+      "epoch": 0.18,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.173036831132732e-05,
+      "loss": 2.2279,
       "step": 1980
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003915322362777999,
-      "loss": 2.1415,
+      "epoch": 0.18,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.196201065554784e-05,
+      "loss": 2.1884,
       "step": 1985
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00039143898316448627,
-      "loss": 2.1537,
+      "epoch": 0.18,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.219365299976837e-05,
+      "loss": 2.2368,
       "step": 1990
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00039134523060255157,
-      "loss": 2.212,
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.242529534398889e-05,
+      "loss": 2.175,
       "step": 1995
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003912509788365885,
-      "loss": 2.1902,
+      "epoch": 0.19,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.265693768820942e-05,
+      "loss": 2.1861,
       "step": 2000
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003911562281124921,
-      "loss": 2.1683,
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.288858003242993e-05,
+      "loss": 2.2325,
       "step": 2005
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003910609786774593,
-      "loss": 2.1913,
+      "epoch": 0.19,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.312022237665046e-05,
+      "loss": 2.2002,
       "step": 2010
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00039096523077998817,
-      "loss": 2.183,
+      "epoch": 0.19,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.335186472087098e-05,
+      "loss": 2.2053,
       "step": 2015
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00039086898466987705,
-      "loss": 2.2021,
+      "epoch": 0.19,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.358350706509151e-05,
+      "loss": 2.2693,
       "step": 2020
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00039077224059822413,
-      "loss": 2.1907,
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.381514940931203e-05,
+      "loss": 2.1776,
       "step": 2025
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00039067499881742675,
-      "loss": 2.1648,
+      "epoch": 0.19,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.404679175353254e-05,
+      "loss": 2.1938,
       "step": 2030
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00039057725958118085,
-      "loss": 2.1839,
+      "epoch": 0.19,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.427843409775307e-05,
+      "loss": 2.2114,
       "step": 2035
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.00039047902314448,
-      "loss": 2.1565,
+      "epoch": 0.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.451007644197359e-05,
+      "loss": 2.2895,
       "step": 2040
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000390380289763615,
-      "loss": 2.1839,
+      "epoch": 0.19,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.474171878619412e-05,
+      "loss": 2.2338,
       "step": 2045
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00039028105969617327,
-      "loss": 2.2042,
+      "epoch": 0.19,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.497336113041463e-05,
+      "loss": 2.21,
       "step": 2050
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 0.0003901813332010378,
-      "loss": 2.2105,
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.520500347463517e-05,
+      "loss": 2.256,
       "step": 2055
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00039008111053838695,
-      "loss": 2.1962,
+      "epoch": 0.19,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.54366458188557e-05,
+      "loss": 2.2182,
       "step": 2060
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00038998039196969336,
-      "loss": 2.1674,
+      "epoch": 0.19,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.566828816307621e-05,
+      "loss": 2.2088,
       "step": 2065
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0003898791777577235,
-      "loss": 2.2015,
+      "epoch": 0.19,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.589993050729674e-05,
+      "loss": 2.2362,
       "step": 2070
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000389777468166537,
-      "loss": 2.1528,
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.613157285151726e-05,
+      "loss": 2.2191,
       "step": 2075
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003896752634614859,
-      "loss": 2.2207,
+      "epoch": 0.19,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.636321519573779e-05,
+      "loss": 2.1633,
       "step": 2080
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00038957256390921384,
-      "loss": 2.1882,
+      "epoch": 0.19,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.65948575399583e-05,
+      "loss": 2.2223,
       "step": 2085
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003894693697776555,
-      "loss": 2.1701,
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.682649988417883e-05,
+      "loss": 2.2014,
       "step": 2090
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00038936568133603595,
-      "loss": 2.1576,
+      "epoch": 0.19,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.705814222839935e-05,
+      "loss": 2.1771,
       "step": 2095
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003892614988548699,
-      "loss": 2.1723,
+      "epoch": 0.19,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.728978457261988e-05,
+      "loss": 2.2872,
       "step": 2100
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003891568226059608,
-      "loss": 2.2056,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.752142691684041e-05,
+      "loss": 2.1927,
       "step": 2105
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0003890516528624005,
-      "loss": 2.1687,
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.775306926106093e-05,
+      "loss": 2.241,
       "step": 2110
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003889459898985682,
-      "loss": 2.178,
+      "epoch": 0.2,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.798471160528146e-05,
+      "loss": 2.2358,
       "step": 2115
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003888398339901299,
-      "loss": 2.1991,
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.821635394950197e-05,
+      "loss": 2.2333,
       "step": 2120
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00038873318541403774,
-      "loss": 2.1967,
+      "epoch": 0.2,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.84479962937225e-05,
+      "loss": 2.1645,
       "step": 2125
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.000388626044448529,
-      "loss": 2.2305,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.867963863794302e-05,
+      "loss": 2.1647,
       "step": 2130
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003885184113731258,
-      "loss": 2.1956,
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.891128098216355e-05,
+      "loss": 2.1548,
       "step": 2135
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003884102864686339,
-      "loss": 2.1677,
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.914292332638407e-05,
+      "loss": 2.1943,
       "step": 2140
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038830167001714243,
-      "loss": 2.194,
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.93745656706046e-05,
+      "loss": 2.2037,
       "step": 2145
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00038819256230202275,
-      "loss": 2.1536,
+      "epoch": 0.2,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.960620801482511e-05,
+      "loss": 2.2391,
       "step": 2150
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00038808296360792787,
-      "loss": 2.1888,
+      "epoch": 0.2,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.983785035904564e-05,
+      "loss": 2.2119,
       "step": 2155
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00038797287422079194,
-      "loss": 2.1827,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010006949270326616,
+      "loss": 2.1864,
       "step": 2160
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0003878622944278291,
-      "loss": 2.1551,
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010030113504748668,
+      "loss": 2.1818,
       "step": 2165
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.000387751224517533,
-      "loss": 2.1649,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001005327773917072,
+      "loss": 2.2662,
       "step": 2170
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003876396647796759,
-      "loss": 2.1716,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010076441973592774,
+      "loss": 2.28,
       "step": 2175
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000387527615505308,
-      "loss": 2.2026,
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010099606208014825,
+      "loss": 2.2524,
       "step": 2180
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00038741507698675665,
-      "loss": 2.1967,
+      "epoch": 0.2,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00010122770442436877,
+      "loss": 2.2377,
       "step": 2185
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00038730204951762574,
-      "loss": 2.1593,
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001014593467685893,
+      "loss": 2.2419,
       "step": 2190
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00038718853339279457,
-      "loss": 2.1712,
+      "epoch": 0.2,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010169098911280983,
+      "loss": 2.2186,
       "step": 2195
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00038707452890841744,
-      "loss": 2.1664,
+      "epoch": 0.2,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010192263145703036,
+      "loss": 2.2146,
       "step": 2200
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003869600363619227,
-      "loss": 2.1739,
+      "epoch": 0.2,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010215427380125086,
+      "loss": 2.2597,
       "step": 2205
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00038684505605201207,
-      "loss": 2.1687,
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010238591614547139,
+      "loss": 2.2081,
       "step": 2210
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00038672958827865966,
-      "loss": 2.1776,
+      "epoch": 0.21,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010261755848969192,
+      "loss": 2.2238,
       "step": 2215
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0003866136333431116,
-      "loss": 2.1977,
+      "epoch": 0.21,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010284920083391245,
+      "loss": 2.26,
       "step": 2220
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1826171875,
-      "learning_rate": 0.0003864971915478846,
-      "loss": 2.2004,
+      "epoch": 0.21,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00010308084317813296,
+      "loss": 2.2359,
       "step": 2225
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003863802631967659,
-      "loss": 2.2117,
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010331248552235349,
+      "loss": 2.1727,
       "step": 2230
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00038626284859481194,
-      "loss": 2.1687,
+      "epoch": 0.21,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010354412786657402,
+      "loss": 2.1833,
       "step": 2235
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038614494804834783,
-      "loss": 2.1712,
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010377577021079455,
+      "loss": 2.2411,
       "step": 2240
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003860265618649664,
-      "loss": 2.161,
+      "epoch": 0.21,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00010400741255501508,
+      "loss": 2.26,
       "step": 2245
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00038590769035352755,
-      "loss": 2.1636,
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010423905489923558,
+      "loss": 2.2362,
       "step": 2250
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.181640625,
-      "learning_rate": 0.0003857883338241573,
-      "loss": 2.1801,
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010447069724345611,
+      "loss": 2.1814,
       "step": 2255
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.000385668492588247,
-      "loss": 2.1542,
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010470233958767664,
+      "loss": 2.1848,
       "step": 2260
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003855481669584527,
-      "loss": 2.1636,
+      "epoch": 0.21,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010493398193189717,
+      "loss": 2.21,
       "step": 2265
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.000385427357248694,
-      "loss": 2.2045,
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010516562427611767,
+      "loss": 2.1993,
       "step": 2270
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00038530606377415363,
-      "loss": 2.2136,
+      "epoch": 0.21,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001053972666203382,
+      "loss": 2.16,
       "step": 2275
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003851842868512763,
-      "loss": 2.1449,
+      "epoch": 0.21,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010562890896455873,
+      "loss": 2.1781,
       "step": 2280
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.000385062026797768,
-      "loss": 2.173,
+      "epoch": 0.21,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00010586055130877926,
+      "loss": 2.2268,
       "step": 2285
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003849392839325952,
-      "loss": 2.1765,
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010609219365299978,
+      "loss": 2.2153,
       "step": 2290
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00038481605857598397,
-      "loss": 2.1903,
+      "epoch": 0.21,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001063238359972203,
+      "loss": 2.2119,
       "step": 2295
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00038469235104941925,
-      "loss": 2.2198,
+      "epoch": 0.21,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010655547834144082,
+      "loss": 2.1815,
       "step": 2300
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003845681616756437,
-      "loss": 2.1736,
+      "epoch": 0.21,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00010678712068566134,
+      "loss": 2.1753,
       "step": 2305
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003844434907786574,
-      "loss": 2.1704,
+      "epoch": 0.21,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010701876302988187,
+      "loss": 2.2399,
       "step": 2310
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003843183386837163,
-      "loss": 2.2177,
+      "epoch": 0.21,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010725040537410239,
+      "loss": 2.1628,
       "step": 2315
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00038419270571733217,
-      "loss": 2.2014,
+      "epoch": 0.21,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001074820477183229,
+      "loss": 2.1869,
       "step": 2320
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003840665922072709,
-      "loss": 2.1716,
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010771369006254343,
+      "loss": 2.2863,
       "step": 2325
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00038393999848255247,
-      "loss": 2.1645,
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010794533240676396,
+      "loss": 2.2107,
       "step": 2330
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003838129248734495,
-      "loss": 2.1691,
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001081769747509845,
+      "loss": 2.2098,
       "step": 2335
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003836853717114866,
-      "loss": 2.143,
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000108408617095205,
+      "loss": 2.1832,
       "step": 2340
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0003835573393294396,
-      "loss": 2.2016,
+      "epoch": 0.22,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00010864025943942553,
+      "loss": 2.2119,
       "step": 2345
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00038342882806133435,
-      "loss": 2.1799,
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00010887190178364606,
+      "loss": 2.2116,
       "step": 2350
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00038329983824244624,
-      "loss": 2.1505,
-      "step": 2355
+      "epoch": 0.22,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010910354412786659,
+      "loss": 2.2265,
+      "step": 2355
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003831703702092993,
-      "loss": 2.1802,
+      "epoch": 0.22,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010933518647208709,
+      "loss": 2.2605,
       "step": 2360
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038304042429966487,
-      "loss": 2.1758,
+      "epoch": 0.22,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00010956682881630762,
+      "loss": 2.2035,
       "step": 2365
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00038291000085256116,
-      "loss": 2.1753,
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010979847116052815,
+      "loss": 2.2229,
       "step": 2370
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003827791002082524,
-      "loss": 2.1467,
+      "epoch": 0.22,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011003011350474868,
+      "loss": 2.2364,
       "step": 2375
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00038264772270824753,
-      "loss": 2.1684,
+      "epoch": 0.22,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011026175584896921,
+      "loss": 2.1944,
       "step": 2380
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003825158686952997,
-      "loss": 2.1402,
+      "epoch": 0.22,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011049339819318971,
+      "loss": 2.2563,
       "step": 2385
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00038238353851340534,
-      "loss": 2.1557,
+      "epoch": 0.22,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011072504053741024,
+      "loss": 2.2039,
       "step": 2390
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00038225073250780284,
-      "loss": 2.1839,
+      "epoch": 0.22,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011095668288163077,
+      "loss": 2.2322,
       "step": 2395
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003821174510249724,
-      "loss": 2.1392,
+      "epoch": 0.22,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001111883252258513,
+      "loss": 2.2286,
       "step": 2400
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003819836944126343,
-      "loss": 2.1385,
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001114199675700718,
+      "loss": 2.219,
       "step": 2405
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00038184946301974876,
-      "loss": 2.1517,
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011165160991429234,
+      "loss": 2.1879,
       "step": 2410
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003817147571965143,
-      "loss": 2.1759,
+      "epoch": 0.22,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011188325225851287,
+      "loss": 2.1573,
       "step": 2415
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038157957729436753,
-      "loss": 2.1571,
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001121148946027334,
+      "loss": 2.203,
       "step": 2420
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00038144392366598167,
-      "loss": 2.1535,
+      "epoch": 0.22,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011234653694695391,
+      "loss": 2.2012,
       "step": 2425
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.000381307796665266,
-      "loss": 2.1696,
+      "epoch": 0.23,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011257817929117443,
+      "loss": 2.2257,
       "step": 2430
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038117119664736454,
-      "loss": 2.2011,
+      "epoch": 0.23,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011280982163539496,
+      "loss": 2.2242,
       "step": 2435
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00038103412396865575,
-      "loss": 2.2036,
+      "epoch": 0.23,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00011304146397961547,
+      "loss": 2.1507,
       "step": 2440
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00038089657898675083,
-      "loss": 2.1503,
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000113273106323836,
+      "loss": 2.2293,
       "step": 2445
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003807585620604935,
-      "loss": 2.165,
+      "epoch": 0.23,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00011350474866805652,
+      "loss": 2.2078,
       "step": 2450
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003806200735499585,
-      "loss": 2.1836,
+      "epoch": 0.23,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011373639101227704,
+      "loss": 2.1999,
       "step": 2455
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00038048111381645117,
-      "loss": 2.1602,
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011396803335649757,
+      "loss": 2.1265,
       "step": 2460
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00038034168322250594,
-      "loss": 2.1654,
+      "epoch": 0.23,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001141996757007181,
+      "loss": 2.1881,
       "step": 2465
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003802017821318858,
-      "loss": 2.1579,
+      "epoch": 0.23,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011443131804493863,
+      "loss": 2.2739,
       "step": 2470
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00038006141090958127,
-      "loss": 2.1749,
+      "epoch": 0.23,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011466296038915913,
+      "loss": 2.2101,
       "step": 2475
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003799205699218094,
-      "loss": 2.1696,
+      "epoch": 0.23,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011489460273337966,
+      "loss": 2.2168,
       "step": 2480
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003797792595360127,
-      "loss": 2.1507,
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011512624507760019,
+      "loss": 2.2322,
       "step": 2485
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003796374801208585,
-      "loss": 2.1427,
+      "epoch": 0.23,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011535788742182072,
+      "loss": 2.2198,
       "step": 2490
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003794952320462376,
-      "loss": 2.189,
+      "epoch": 0.23,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011558952976604122,
+      "loss": 2.1845,
       "step": 2495
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00037935251568326357,
-      "loss": 2.1501,
+      "epoch": 0.23,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00011582117211026175,
+      "loss": 2.281,
       "step": 2500
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003792093314042718,
-      "loss": 2.2141,
+      "epoch": 0.23,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011605281445448228,
+      "loss": 2.1971,
       "step": 2505
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003790656795828182,
-      "loss": 2.1319,
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011628445679870281,
+      "loss": 2.2335,
       "step": 2510
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003789215605936787,
-      "loss": 2.1639,
+      "epoch": 0.23,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011651609914292334,
+      "loss": 2.1277,
       "step": 2515
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00037877697481284784,
-      "loss": 2.2032,
+      "epoch": 0.23,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00011674774148714385,
+      "loss": 2.1897,
       "step": 2520
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003786319226175382,
-      "loss": 2.1946,
+      "epoch": 0.23,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011697938383136438,
+      "loss": 2.2237,
       "step": 2525
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.000378486404386179,
-      "loss": 2.1725,
+      "epoch": 0.23,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001172110261755849,
+      "loss": 2.2502,
       "step": 2530
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003783404204984154,
-      "loss": 2.1877,
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011744266851980544,
+      "loss": 2.2232,
       "step": 2535
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003781939713351074,
-      "loss": 2.171,
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011767431086402597,
+      "loss": 2.2146,
       "step": 2540
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003780470572783288,
-      "loss": 2.1502,
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011790595320824647,
+      "loss": 2.2477,
       "step": 2545
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00037789967871136647,
-      "loss": 2.1711,
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000118137595552467,
+      "loss": 2.2037,
       "step": 2550
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00037775183601871894,
-      "loss": 2.1952,
+      "epoch": 0.24,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011836923789668753,
+      "loss": 2.2035,
       "step": 2555
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00037760352958609564,
-      "loss": 2.1996,
+      "epoch": 0.24,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011860088024090805,
+      "loss": 2.1922,
       "step": 2560
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.000377454759800416,
-      "loss": 2.1831,
+      "epoch": 0.24,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00011883252258512856,
+      "loss": 2.1737,
       "step": 2565
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00037730552704980813,
-      "loss": 2.1548,
+      "epoch": 0.24,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011906416492934909,
+      "loss": 2.2381,
       "step": 2570
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00037715583172360807,
-      "loss": 2.2046,
+      "epoch": 0.24,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00011929580727356961,
+      "loss": 2.2847,
       "step": 2575
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003770056742123586,
-      "loss": 2.187,
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011952744961779014,
+      "loss": 2.2089,
       "step": 2580
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003768550549078085,
-      "loss": 2.1945,
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011975909196201067,
+      "loss": 2.1496,
       "step": 2585
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003767039742029111,
-      "loss": 2.1916,
+      "epoch": 0.24,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00011999073430623119,
+      "loss": 2.2346,
       "step": 2590
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00037655243249182355,
-      "loss": 2.1468,
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001202223766504517,
+      "loss": 2.1798,
       "step": 2595
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00037640043016990587,
-      "loss": 2.2071,
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012045401899467223,
+      "loss": 2.1939,
       "step": 2600
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00037624796763371953,
-      "loss": 2.1319,
+      "epoch": 0.24,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00012068566133889276,
+      "loss": 2.1727,
       "step": 2605
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00037609504528102693,
-      "loss": 2.2019,
+      "epoch": 0.24,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012091730368311326,
+      "loss": 2.1835,
       "step": 2610
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00037594166351078985,
-      "loss": 2.1725,
+      "epoch": 0.24,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001211489460273338,
+      "loss": 2.2309,
       "step": 2615
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00037578782272316877,
-      "loss": 2.1626,
+      "epoch": 0.24,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00012138058837155432,
+      "loss": 2.2195,
       "step": 2620
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003756335233195218,
-      "loss": 2.1622,
+      "epoch": 0.24,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012161223071577485,
+      "loss": 2.1988,
       "step": 2625
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003754787657024033,
-      "loss": 2.1923,
+      "epoch": 0.24,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012184387305999538,
+      "loss": 2.1423,
       "step": 2630
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00037532355027556334,
-      "loss": 2.176,
+      "epoch": 0.24,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001220755154042159,
+      "loss": 2.1697,
       "step": 2635
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00037516787744394615,
-      "loss": 2.1632,
+      "epoch": 0.24,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00012230715774843643,
+      "loss": 2.1498,
       "step": 2640
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003750117476136895,
-      "loss": 2.2095,
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012253880009265695,
+      "loss": 2.1818,
       "step": 2645
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003748551611921232,
-      "loss": 2.16,
+      "epoch": 0.25,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00012277044243687746,
+      "loss": 2.148,
       "step": 2650
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003746981185877684,
-      "loss": 2.1668,
+      "epoch": 0.25,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012300208478109798,
+      "loss": 2.1947,
       "step": 2655
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003745406202103365,
-      "loss": 2.1289,
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001232337271253185,
+      "loss": 2.2196,
       "step": 2660
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003743826664707279,
-      "loss": 2.1539,
+      "epoch": 0.25,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012346536946953904,
+      "loss": 2.2211,
       "step": 2665
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003742242577810307,
-      "loss": 2.1488,
+      "epoch": 0.25,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012369701181375956,
+      "loss": 2.1992,
       "step": 2670
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003740653945545205,
-      "loss": 2.1595,
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001239286541579801,
+      "loss": 2.1636,
       "step": 2675
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003739060772056583,
-      "loss": 2.2113,
+      "epoch": 0.25,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001241602965022006,
+      "loss": 2.2144,
       "step": 2680
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00037374630615008987,
-      "loss": 2.1642,
+      "epoch": 0.25,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012439193884642113,
+      "loss": 2.2184,
       "step": 2685
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003735860818046451,
-      "loss": 2.2222,
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012462358119064165,
+      "loss": 2.1898,
       "step": 2690
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003734254045873359,
-      "loss": 2.1348,
+      "epoch": 0.25,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001248552235348622,
+      "loss": 2.2049,
       "step": 2695
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00037326427491735613,
-      "loss": 2.1671,
+      "epoch": 0.25,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012508686587908268,
+      "loss": 2.2245,
       "step": 2700
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00037310269321507984,
-      "loss": 2.1556,
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012531850822330323,
+      "loss": 2.2388,
       "step": 2705
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003729406599020605,
-      "loss": 2.1649,
+      "epoch": 0.25,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012555015056752374,
+      "loss": 2.2528,
       "step": 2710
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00037277817540102973,
-      "loss": 2.2039,
+      "epoch": 0.25,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012578179291174429,
+      "loss": 2.2148,
       "step": 2715
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003726152401358962,
-      "loss": 2.1396,
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001260134352559648,
+      "loss": 2.1997,
       "step": 2720
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003724518545317448,
-      "loss": 2.1507,
+      "epoch": 0.25,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012624507760018532,
+      "loss": 2.2232,
       "step": 2725
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00037228801901483503,
-      "loss": 2.1799,
+      "epoch": 0.25,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00012647671994440584,
+      "loss": 2.2382,
       "step": 2730
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00037212373401260046,
-      "loss": 2.187,
+      "epoch": 0.25,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012670836228862638,
+      "loss": 2.1553,
       "step": 2735
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003719589999536471,
-      "loss": 2.1976,
+      "epoch": 0.25,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001269400046328469,
+      "loss": 2.1868,
       "step": 2740
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0003717938172677527,
-      "loss": 2.1753,
+      "epoch": 0.25,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001271716469770674,
+      "loss": 2.218,
       "step": 2745
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00037162818638586533,
-      "loss": 2.1722,
+      "epoch": 0.25,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012740328932128793,
+      "loss": 2.2198,
       "step": 2750
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00037146210774010235,
-      "loss": 2.1647,
+      "epoch": 0.26,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00012763493166550847,
+      "loss": 2.2061,
       "step": 2755
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00037129558176374937,
-      "loss": 2.1425,
+      "epoch": 0.26,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000127866574009729,
+      "loss": 2.2699,
       "step": 2760
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000371128608891259,
-      "loss": 2.1494,
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001280982163539495,
+      "loss": 2.2287,
       "step": 2765
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00037096118955824985,
-      "loss": 2.1378,
+      "epoch": 0.26,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00012832985869817002,
+      "loss": 2.1879,
       "step": 2770
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003707933242015051,
-      "loss": 2.2139,
+      "epoch": 0.26,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00012856150104239057,
+      "loss": 2.163,
       "step": 2775
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00037062501325897185,
-      "loss": 2.1782,
+      "epoch": 0.26,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012879314338661108,
+      "loss": 2.2158,
       "step": 2780
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003704562571697595,
-      "loss": 2.1569,
+      "epoch": 0.26,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001290247857308316,
+      "loss": 2.2131,
       "step": 2785
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003702870563741389,
-      "loss": 2.1816,
+      "epoch": 0.26,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00012925642807505211,
+      "loss": 2.1876,
       "step": 2790
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00037011741131354095,
-      "loss": 2.1699,
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00012948807041927266,
+      "loss": 2.172,
       "step": 2795
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00036994732243055586,
-      "loss": 2.1796,
+      "epoch": 0.26,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00012971971276349317,
+      "loss": 2.1694,
       "step": 2800
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003697767901689315,
-      "loss": 2.1894,
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001299513551077137,
+      "loss": 2.2098,
       "step": 2805
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00036960581497357263,
-      "loss": 2.2078,
+      "epoch": 0.26,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00013018299745193423,
+      "loss": 2.2295,
       "step": 2810
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00036943439729053936,
-      "loss": 2.2227,
+      "epoch": 0.26,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00013041463979615472,
+      "loss": 2.2365,
       "step": 2815
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003692625375670466,
-      "loss": 2.129,
+      "epoch": 0.26,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00013064628214037527,
+      "loss": 2.2187,
       "step": 2820
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00036909023625146216,
-      "loss": 2.1497,
+      "epoch": 0.26,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00013087792448459578,
+      "loss": 2.2166,
       "step": 2825
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0003689174937933061,
-      "loss": 2.1787,
+      "epoch": 0.26,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00013110956682881633,
+      "loss": 2.1975,
       "step": 2830
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003687443106432493,
-      "loss": 2.1767,
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013134120917303682,
+      "loss": 2.1942,
       "step": 2835
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00036857068725311247,
-      "loss": 2.2093,
+      "epoch": 0.26,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013157285151725736,
+      "loss": 2.2058,
       "step": 2840
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00036839662407586477,
-      "loss": 2.192,
+      "epoch": 0.26,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013180449386147788,
+      "loss": 2.2803,
       "step": 2845
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003682221215656227,
-      "loss": 2.1946,
+      "epoch": 0.26,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013203613620569842,
+      "loss": 2.2094,
       "step": 2850
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00036804718017764915,
-      "loss": 2.1645,
+      "epoch": 0.26,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00013226777854991894,
+      "loss": 2.1815,
       "step": 2855
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00036787180036835176,
-      "loss": 2.1705,
+      "epoch": 0.27,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00013249942089413945,
+      "loss": 2.2405,
       "step": 2860
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003676959825952821,
-      "loss": 2.1524,
+      "epoch": 0.27,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00013273106323835997,
+      "loss": 2.1676,
       "step": 2865
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00036751972731713443,
-      "loss": 2.1779,
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001329627055825805,
+      "loss": 2.2528,
       "step": 2870
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003673430349937442,
-      "loss": 2.1696,
+      "epoch": 0.27,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013319434792680103,
+      "loss": 2.1596,
       "step": 2875
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00036716590608608726,
-      "loss": 2.2063,
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013342599027102155,
+      "loss": 2.1556,
       "step": 2880
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00036698834105627833,
-      "loss": 2.1803,
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013365763261524206,
+      "loss": 2.1646,
       "step": 2885
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00036681034036757025,
-      "loss": 2.1442,
+      "epoch": 0.27,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001338892749594626,
+      "loss": 2.251,
       "step": 2890
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00036663190448435197,
-      "loss": 2.1818,
+      "epoch": 0.27,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013412091730368312,
+      "loss": 2.254,
       "step": 2895
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003664530338721483,
-      "loss": 2.1825,
+      "epoch": 0.27,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00013435255964790367,
+      "loss": 2.2205,
       "step": 2900
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00036627372899761783,
-      "loss": 2.1626,
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013458420199212416,
+      "loss": 2.2488,
       "step": 2905
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00036609399032855246,
-      "loss": 2.1804,
+      "epoch": 0.27,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001348158443363447,
+      "loss": 2.1944,
       "step": 2910
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003659138183338755,
-      "loss": 2.2026,
+      "epoch": 0.27,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013504748668056522,
+      "loss": 2.204,
       "step": 2915
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.000365733213483641,
-      "loss": 2.1142,
+      "epoch": 0.27,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00013527912902478573,
+      "loss": 2.1932,
       "step": 2920
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003655521762490322,
-      "loss": 2.1742,
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013551077136900625,
+      "loss": 2.2121,
       "step": 2925
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00036537070710236045,
-      "loss": 2.1832,
+      "epoch": 0.27,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001357424137132268,
+      "loss": 2.175,
       "step": 2930
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003651888065170638,
-      "loss": 2.1556,
+      "epoch": 0.27,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001359740560574473,
+      "loss": 2.2337,
       "step": 2935
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003650064749677061,
-      "loss": 2.1317,
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00013620569840166783,
+      "loss": 2.2156,
       "step": 2940
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003648237129299752,
-      "loss": 2.1326,
+      "epoch": 0.27,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013643734074588837,
+      "loss": 2.2503,
       "step": 2945
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003646405208806825,
-      "loss": 2.1876,
+      "epoch": 0.27,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013666898309010886,
+      "loss": 2.2421,
       "step": 2950
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.000364456899297761,
-      "loss": 2.1772,
+      "epoch": 0.27,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001369006254343294,
+      "loss": 2.2053,
       "step": 2955
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00036427284866026425,
-      "loss": 2.1754,
+      "epoch": 0.27,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013713226777854992,
+      "loss": 2.1722,
       "step": 2960
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00036408836944836544,
-      "loss": 2.1947,
+      "epoch": 0.27,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00013736391012277046,
+      "loss": 2.2072,
       "step": 2965
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00036390346214335565,
-      "loss": 2.1612,
+      "epoch": 0.28,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00013759555246699095,
+      "loss": 2.2156,
       "step": 2970
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003637181272276428,
-      "loss": 2.1577,
+      "epoch": 0.28,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001378271948112115,
+      "loss": 2.1809,
       "step": 2975
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00036353236518475067,
-      "loss": 2.1848,
+      "epoch": 0.28,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000138058837155432,
+      "loss": 2.2127,
       "step": 2980
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00036334617649931713,
-      "loss": 2.1327,
+      "epoch": 0.28,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00013829047949965255,
+      "loss": 2.1664,
       "step": 2985
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003631595616570931,
-      "loss": 2.165,
+      "epoch": 0.28,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013852212184387307,
+      "loss": 2.1366,
       "step": 2990
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003629725211449416,
-      "loss": 2.1891,
+      "epoch": 0.28,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001387537641880936,
+      "loss": 2.1917,
       "step": 2995
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.0003627850554508359,
-      "loss": 2.1425,
+      "epoch": 0.28,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001389854065323141,
+      "loss": 2.1838,
       "step": 3000
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003625971650638586,
-      "loss": 2.2033,
+      "epoch": 0.28,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00013921704887653465,
+      "loss": 2.2292,
       "step": 3005
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00036240885047420044,
-      "loss": 2.1094,
+      "epoch": 0.28,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00013944869122075516,
+      "loss": 2.2082,
       "step": 3010
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00036222011217315863,
-      "loss": 2.167,
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013968033356497568,
+      "loss": 2.2012,
       "step": 3015
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00036203095065313597,
-      "loss": 2.1472,
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001399119759091962,
+      "loss": 2.2236,
       "step": 3020
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003618413664076394,
-      "loss": 2.1544,
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014014361825341674,
+      "loss": 2.2565,
       "step": 3025
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003616513599312787,
-      "loss": 2.1752,
+      "epoch": 0.28,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00014037526059763726,
+      "loss": 2.2026,
       "step": 3030
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.0003614609317197651,
-      "loss": 2.1324,
+      "epoch": 0.28,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001406069029418578,
+      "loss": 2.2016,
       "step": 3035
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00036127008226991036,
-      "loss": 2.1255,
+      "epoch": 0.28,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001408385452860783,
+      "loss": 2.1589,
       "step": 3040
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00036107881207962493,
-      "loss": 2.1587,
+      "epoch": 0.28,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00014107018763029883,
+      "loss": 2.1791,
       "step": 3045
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00036088712164791706,
-      "loss": 2.1735,
+      "epoch": 0.28,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014130182997451935,
+      "loss": 2.2499,
       "step": 3050
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003606950114748914,
-      "loss": 2.2219,
+      "epoch": 0.28,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00014153347231873987,
+      "loss": 2.1763,
       "step": 3055
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003605024820617477,
-      "loss": 2.1603,
+      "epoch": 0.28,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00014176511466296038,
+      "loss": 2.2262,
       "step": 3060
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003603095339107794,
-      "loss": 2.1873,
+      "epoch": 0.28,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00014199675700718093,
+      "loss": 2.2255,
       "step": 3065
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003601161675253724,
-      "loss": 2.1496,
+      "epoch": 0.28,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00014222839935140144,
+      "loss": 2.1754,
       "step": 3070
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00035992238341000377,
-      "loss": 2.174,
+      "epoch": 0.28,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00014246004169562196,
+      "loss": 2.1521,
       "step": 3075
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003597281820702403,
-      "loss": 2.1466,
+      "epoch": 0.29,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001426916840398425,
+      "loss": 2.2273,
       "step": 3080
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00035953356401273754,
-      "loss": 2.2021,
+      "epoch": 0.29,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000142923326384063,
+      "loss": 2.203,
       "step": 3085
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00035933852974523797,
-      "loss": 2.1912,
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014315496872828354,
+      "loss": 2.2151,
       "step": 3090
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00035914307977657,
-      "loss": 2.1559,
+      "epoch": 0.29,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00014338661107250405,
+      "loss": 2.2108,
       "step": 3095
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00035894721461664653,
-      "loss": 2.1599,
+      "epoch": 0.29,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001436182534167246,
+      "loss": 2.1667,
       "step": 3100
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003587509347764639,
-      "loss": 2.1783,
+      "epoch": 0.29,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014384989576094509,
+      "loss": 2.1934,
       "step": 3105
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003585542407681,
-      "loss": 2.1911,
+      "epoch": 0.29,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014408153810516563,
+      "loss": 2.1597,
       "step": 3110
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003583571331047133,
-      "loss": 2.1656,
+      "epoch": 0.29,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014431318044938615,
+      "loss": 2.1986,
       "step": 3115
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003581596123005417,
-      "loss": 2.2032,
+      "epoch": 0.29,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001445448227936067,
+      "loss": 2.2097,
       "step": 3120
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00035796167887090064,
-      "loss": 2.1761,
+      "epoch": 0.29,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001447764651378272,
+      "loss": 2.2052,
       "step": 3125
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003577633333321824,
-      "loss": 2.1613,
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014500810748204772,
+      "loss": 2.2297,
       "step": 3130
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00035756457620185413,
-      "loss": 2.1708,
+      "epoch": 0.29,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00014523974982626824,
+      "loss": 2.1506,
       "step": 3135
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003573654079984569,
-      "loss": 2.1343,
+      "epoch": 0.29,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014547139217048878,
+      "loss": 2.2407,
       "step": 3140
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003571658292416043,
-      "loss": 2.1691,
+      "epoch": 0.29,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001457030345147093,
+      "loss": 2.1839,
       "step": 3145
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00035696584045198085,
-      "loss": 2.1903,
+      "epoch": 0.29,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014593467685892981,
+      "loss": 2.1912,
       "step": 3150
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.000356765442151341,
-      "loss": 2.1999,
+      "epoch": 0.29,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00014616631920315033,
+      "loss": 2.2142,
       "step": 3155
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003565646348625074,
-      "loss": 2.178,
+      "epoch": 0.29,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00014639796154737087,
+      "loss": 2.2034,
       "step": 3160
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00035636341910937,
-      "loss": 2.1452,
+      "epoch": 0.29,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001466296038915914,
+      "loss": 2.194,
       "step": 3165
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00035616179541688396,
-      "loss": 2.1834,
+      "epoch": 0.29,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014686124623581193,
+      "loss": 2.2574,
       "step": 3170
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.1845703125,
-      "learning_rate": 0.00035595976431106917,
-      "loss": 2.1612,
+      "epoch": 0.29,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00014709288858003242,
+      "loss": 2.1939,
       "step": 3175
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00035575732631900817,
-      "loss": 2.1553,
+      "epoch": 0.29,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00014732453092425297,
+      "loss": 2.2151,
       "step": 3180
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003555544819688451,
-      "loss": 2.1546,
+      "epoch": 0.3,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00014755617326847348,
+      "loss": 2.2117,
       "step": 3185
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00035535123178978415,
-      "loss": 2.1749,
+      "epoch": 0.3,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000147787815612694,
+      "loss": 2.2214,
       "step": 3190
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003551475763120886,
-      "loss": 2.1814,
+      "epoch": 0.3,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00014801945795691452,
+      "loss": 2.1911,
       "step": 3195
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003549435160670787,
-      "loss": 2.1969,
+      "epoch": 0.3,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00014825110030113506,
+      "loss": 2.2095,
       "step": 3200
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00035473905158713096,
-      "loss": 2.1151,
+      "epoch": 0.3,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00014848274264535558,
+      "loss": 2.239,
       "step": 3205
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00035453418340567654,
-      "loss": 2.1841,
+      "epoch": 0.3,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001487143849895761,
+      "loss": 2.1851,
       "step": 3210
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003543289120571996,
-      "loss": 2.1678,
+      "epoch": 0.3,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00014894602733379664,
+      "loss": 2.1806,
       "step": 3215
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00035412323807723644,
-      "loss": 2.1795,
+      "epoch": 0.3,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00014917766967801715,
+      "loss": 2.1873,
       "step": 3220
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00035391716200237346,
-      "loss": 2.1592,
+      "epoch": 0.3,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014940931202223767,
+      "loss": 2.1967,
       "step": 3225
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0003537106843702464,
-      "loss": 2.1743,
+      "epoch": 0.3,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00014964095436645819,
+      "loss": 2.2141,
       "step": 3230
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00035350380571953836,
-      "loss": 2.1537,
+      "epoch": 0.3,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014987259671067873,
+      "loss": 2.2022,
       "step": 3235
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00035329652658997883,
-      "loss": 2.1708,
+      "epoch": 0.3,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00015010423905489925,
+      "loss": 2.2128,
       "step": 3240
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00035308884752234207,
-      "loss": 2.1661,
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015033588139911976,
+      "loss": 2.2025,
       "step": 3245
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003528807690584458,
-      "loss": 2.1801,
+      "epoch": 0.3,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015056752374334028,
+      "loss": 2.1742,
       "step": 3250
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003526722917411497,
-      "loss": 2.1754,
+      "epoch": 0.3,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015079916608756082,
+      "loss": 2.2127,
       "step": 3255
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003524634161143539,
-      "loss": 2.1586,
+      "epoch": 0.3,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00015103080843178134,
+      "loss": 2.1809,
       "step": 3260
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00035225414272299784,
-      "loss": 2.1633,
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015126245077600186,
+      "loss": 2.2732,
       "step": 3265
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003520444721130586,
-      "loss": 2.1521,
+      "epoch": 0.3,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015149409312022237,
+      "loss": 2.1859,
       "step": 3270
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00035183440483154977,
-      "loss": 2.1735,
+      "epoch": 0.3,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015172573546444292,
+      "loss": 2.1864,
       "step": 3275
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00035162394142651955,
-      "loss": 2.1732,
+      "epoch": 0.3,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00015195737780866343,
+      "loss": 2.216,
       "step": 3280
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00035141308244704975,
-      "loss": 2.1746,
+      "epoch": 0.3,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015218902015288398,
+      "loss": 2.2082,
       "step": 3285
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0003512018284432542,
-      "loss": 2.1395,
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015242066249710447,
+      "loss": 2.2454,
       "step": 3290
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00035099017996627726,
-      "loss": 2.1977,
+      "epoch": 0.31,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000152652304841325,
+      "loss": 2.1708,
       "step": 3295
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00035077813756829247,
-      "loss": 2.175,
+      "epoch": 0.31,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015288394718554553,
+      "loss": 2.194,
       "step": 3300
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00035056570180250104,
-      "loss": 2.1792,
+      "epoch": 0.31,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00015311558952976607,
+      "loss": 2.2143,
       "step": 3305
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00035035287322313053,
-      "loss": 2.1743,
+      "epoch": 0.31,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00015334723187398656,
+      "loss": 2.1981,
       "step": 3310
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00035013965238543324,
-      "loss": 2.1874,
+      "epoch": 0.31,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001535788742182071,
+      "loss": 2.1964,
       "step": 3315
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003499260398456849,
-      "loss": 2.1612,
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015381051656242762,
+      "loss": 2.2497,
       "step": 3320
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.000349712036161183,
-      "loss": 2.1541,
+      "epoch": 0.31,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015404215890664816,
+      "loss": 2.2482,
       "step": 3325
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003494976418902458,
-      "loss": 2.186,
+      "epoch": 0.31,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015427380125086868,
+      "loss": 2.2535,
       "step": 3330
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00034928285759221027,
-      "loss": 2.1962,
+      "epoch": 0.31,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001545054435950892,
+      "loss": 2.2395,
       "step": 3335
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003490676838274311,
-      "loss": 2.1682,
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001547370859393097,
+      "loss": 2.2444,
       "step": 3340
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0003488521211572789,
-      "loss": 2.1448,
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015496872828353023,
+      "loss": 2.188,
       "step": 3345
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00034863617014413916,
-      "loss": 2.1853,
+      "epoch": 0.31,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00015520037062775077,
+      "loss": 2.1708,
       "step": 3350
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.0003484198313514103,
-      "loss": 2.1469,
+      "epoch": 0.31,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001554320129719713,
+      "loss": 2.1771,
       "step": 3355
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003482031053435025,
-      "loss": 2.1847,
+      "epoch": 0.31,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001556636553161918,
+      "loss": 2.1518,
       "step": 3360
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.18359375,
-      "learning_rate": 0.0003479859926858362,
-      "loss": 2.1409,
+      "epoch": 0.31,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00015589529766041232,
+      "loss": 2.2173,
       "step": 3365
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00034776849394484047,
-      "loss": 2.1633,
+      "epoch": 0.31,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015612694000463286,
+      "loss": 2.2019,
       "step": 3370
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003475506096879518,
-      "loss": 2.1971,
+      "epoch": 0.31,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015635858234885338,
+      "loss": 2.2779,
       "step": 3375
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003473323404836122,
-      "loss": 2.2042,
+      "epoch": 0.31,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001565902246930739,
+      "loss": 2.2017,
       "step": 3380
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00034711368690126836,
-      "loss": 2.186,
+      "epoch": 0.31,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001568218670372944,
+      "loss": 2.1745,
       "step": 3385
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00034689464951136943,
-      "loss": 2.1366,
+      "epoch": 0.31,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00015705350938151496,
+      "loss": 2.1732,
       "step": 3390
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00034667522888536597,
-      "loss": 2.1616,
+      "epoch": 0.31,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00015728515172573547,
+      "loss": 2.2157,
       "step": 3395
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003464554255957085,
-      "loss": 2.135,
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000157516794069956,
+      "loss": 2.2099,
       "step": 3400
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00034623524021584577,
-      "loss": 2.212,
+      "epoch": 0.32,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001577484364141765,
+      "loss": 2.1851,
       "step": 3405
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003460146733202234,
-      "loss": 2.1422,
+      "epoch": 0.32,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015798007875839705,
+      "loss": 2.2642,
       "step": 3410
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00034579372548428235,
-      "loss": 2.1923,
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015821172110261757,
+      "loss": 2.1327,
       "step": 3415
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003455723972844574,
-      "loss": 2.1822,
+      "epoch": 0.32,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001584433634468381,
+      "loss": 2.1935,
       "step": 3420
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003453506892981757,
-      "loss": 2.1625,
+      "epoch": 0.32,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001586750057910586,
+      "loss": 2.2124,
       "step": 3425
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00034512860210385516,
-      "loss": 2.1835,
+      "epoch": 0.32,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00015890664813527914,
+      "loss": 2.1915,
       "step": 3430
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00034490613628090325,
-      "loss": 2.1461,
+      "epoch": 0.32,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015913829047949966,
+      "loss": 2.2042,
       "step": 3435
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.185546875,
-      "learning_rate": 0.00034468329240971486,
-      "loss": 2.1501,
+      "epoch": 0.32,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001593699328237202,
+      "loss": 2.2052,
       "step": 3440
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003444600710716715,
-      "loss": 2.2022,
+      "epoch": 0.32,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001596015751679407,
+      "loss": 2.1956,
       "step": 3445
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00034423647284913934,
-      "loss": 2.1706,
+      "epoch": 0.32,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015983321751216124,
+      "loss": 2.2016,
       "step": 3450
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003440124983254678,
-      "loss": 2.2002,
+      "epoch": 0.32,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00016006485985638175,
+      "loss": 2.2447,
       "step": 3455
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003437881480849881,
-      "loss": 2.1764,
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001602965022006023,
+      "loss": 2.2025,
       "step": 3460
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003435634227130116,
-      "loss": 2.1617,
+      "epoch": 0.32,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001605281445448228,
+      "loss": 2.2128,
       "step": 3465
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003433383227958284,
-      "loss": 2.1917,
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016075978688904333,
+      "loss": 2.2115,
       "step": 3470
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003431128489207058,
-      "loss": 2.161,
+      "epoch": 0.32,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016099142923326385,
+      "loss": 2.2098,
       "step": 3475
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00034288700167588653,
-      "loss": 2.1907,
+      "epoch": 0.32,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016122307157748436,
+      "loss": 2.2214,
       "step": 3480
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003426607816505877,
-      "loss": 2.1946,
+      "epoch": 0.32,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001614547139217049,
+      "loss": 2.1643,
       "step": 3485
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00034243418943499863,
-      "loss": 2.1641,
+      "epoch": 0.32,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016168635626592542,
+      "loss": 2.19,
       "step": 3490
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00034220722562028,
-      "loss": 2.1497,
+      "epoch": 0.32,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00016191799861014594,
+      "loss": 2.1679,
       "step": 3495
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00034197989079856177,
-      "loss": 2.1762,
+      "epoch": 0.32,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00016214964095436645,
+      "loss": 2.1725,
       "step": 3500
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0003417521855629418,
-      "loss": 2.1725,
+      "epoch": 0.32,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.000162381283298587,
+      "loss": 2.1664,
       "step": 3505
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003415241105074845,
-      "loss": 2.1933,
+      "epoch": 0.33,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00016261292564280751,
+      "loss": 2.2045,
       "step": 3510
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00034129566622721893,
-      "loss": 2.1642,
+      "epoch": 0.33,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00016284456798702803,
+      "loss": 2.1669,
       "step": 3515
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003410668533181375,
-      "loss": 2.1441,
+      "epoch": 0.33,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016307621033124855,
+      "loss": 2.157,
       "step": 3520
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00034083767237719437,
-      "loss": 2.1493,
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001633078526754691,
+      "loss": 2.168,
       "step": 3525
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003406081240023038,
-      "loss": 2.1511,
+      "epoch": 0.33,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001635394950196896,
+      "loss": 2.1878,
       "step": 3530
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00034037820879233874,
-      "loss": 2.124,
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016377113736391012,
+      "loss": 2.1989,
       "step": 3535
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003401479273471291,
-      "loss": 2.1946,
+      "epoch": 0.33,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00016400277970813064,
+      "loss": 2.2292,
       "step": 3540
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00033991728026746035,
-      "loss": 2.1686,
+      "epoch": 0.33,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00016423442205235118,
+      "loss": 2.1269,
       "step": 3545
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003396862681550717,
-      "loss": 2.1675,
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001644660643965717,
+      "loss": 2.2122,
       "step": 3550
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00033945489161265496,
-      "loss": 2.1678,
+      "epoch": 0.33,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016469770674079224,
+      "loss": 2.1967,
       "step": 3555
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003392231512438525,
-      "loss": 2.1688,
+      "epoch": 0.33,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00016492934908501273,
+      "loss": 2.1926,
       "step": 3560
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.000338991047653256,
-      "loss": 2.1348,
+      "epoch": 0.33,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016516099142923328,
+      "loss": 2.2155,
       "step": 3565
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003387585814464046,
-      "loss": 2.1784,
+      "epoch": 0.33,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001653926337734538,
+      "loss": 2.1683,
       "step": 3570
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003385257532297837,
-      "loss": 2.1485,
+      "epoch": 0.33,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016562427611767434,
+      "loss": 2.2157,
       "step": 3575
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.000338292563610823,
-      "loss": 2.1796,
+      "epoch": 0.33,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016585591846189483,
+      "loss": 2.1902,
       "step": 3580
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1865234375,
-      "learning_rate": 0.00033805901319789523,
-      "loss": 2.1385,
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016608756080611537,
+      "loss": 2.2501,
       "step": 3585
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00033782510260031397,
-      "loss": 2.1663,
+      "epoch": 0.33,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00016631920315033589,
+      "loss": 2.2006,
       "step": 3590
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00033759083242833316,
-      "loss": 2.1829,
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016655084549455643,
+      "loss": 2.1946,
       "step": 3595
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00033735620329314423,
-      "loss": 2.1654,
+      "epoch": 0.33,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00016678248783877695,
+      "loss": 2.1236,
       "step": 3600
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00033712121580687553,
-      "loss": 2.1817,
+      "epoch": 0.33,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00016701413018299746,
+      "loss": 2.2147,
       "step": 3605
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00033688587058259005,
-      "loss": 2.1509,
+      "epoch": 0.33,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00016724577252721798,
+      "loss": 2.2271,
       "step": 3610
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003366501682342842,
-      "loss": 2.1625,
+      "epoch": 0.33,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016747741487143852,
+      "loss": 2.1912,
       "step": 3615
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00033641410937688623,
-      "loss": 2.1596,
+      "epoch": 0.34,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00016770905721565904,
+      "loss": 2.2131,
       "step": 3620
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00033617769462625414,
-      "loss": 2.1586,
+      "epoch": 0.34,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016794069955987956,
+      "loss": 2.1652,
       "step": 3625
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00033594092459917485,
-      "loss": 2.1758,
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00016817234190410007,
+      "loss": 2.2229,
       "step": 3630
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003357037999133618,
-      "loss": 2.1764,
+      "epoch": 0.34,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001684039842483206,
+      "loss": 2.1651,
       "step": 3635
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.000335466321187454,
-      "loss": 2.1582,
+      "epoch": 0.34,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00016863562659254113,
+      "loss": 2.2002,
       "step": 3640
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00033522848904101387,
-      "loss": 2.2074,
+      "epoch": 0.34,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00016886726893676165,
+      "loss": 2.2457,
       "step": 3645
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003349903040945261,
-      "loss": 2.2073,
+      "epoch": 0.34,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016909891128098217,
+      "loss": 2.1943,
       "step": 3650
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00033475176696939565,
-      "loss": 2.1773,
+      "epoch": 0.34,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00016933055362520268,
+      "loss": 2.2154,
       "step": 3655
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00033451287828794625,
-      "loss": 2.1756,
+      "epoch": 0.34,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00016956219596942323,
+      "loss": 2.2003,
       "step": 3660
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.000334273638673419,
-      "loss": 2.1762,
+      "epoch": 0.34,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00016979383831364374,
+      "loss": 2.1911,
       "step": 3665
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003340340487499705,
-      "loss": 2.1871,
+      "epoch": 0.34,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017002548065786426,
+      "loss": 2.1935,
       "step": 3670
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00033379410914267113,
-      "loss": 2.123,
+      "epoch": 0.34,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017025712300208477,
+      "loss": 2.1958,
       "step": 3675
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003335538204775037,
-      "loss": 2.1596,
+      "epoch": 0.34,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017048876534630532,
+      "loss": 2.1659,
       "step": 3680
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00033331318338136154,
-      "loss": 2.1601,
+      "epoch": 0.34,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017072040769052583,
+      "loss": 2.2091,
       "step": 3685
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00033307219848204727,
-      "loss": 2.1494,
+      "epoch": 0.34,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017095205003474638,
+      "loss": 2.1724,
       "step": 3690
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00033283086640827075,
-      "loss": 2.1872,
+      "epoch": 0.34,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017118369237896687,
+      "loss": 2.2604,
       "step": 3695
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003325891877896475,
-      "loss": 2.1453,
+      "epoch": 0.34,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001714153347231874,
+      "loss": 2.1606,
       "step": 3700
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00033234716325669726,
-      "loss": 2.1355,
+      "epoch": 0.34,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017164697706740793,
+      "loss": 2.24,
       "step": 3705
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003321047934408422,
-      "loss": 2.1521,
+      "epoch": 0.34,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017187861941162847,
+      "loss": 2.212,
       "step": 3710
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00033186207897440535,
-      "loss": 2.1659,
+      "epoch": 0.34,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017211026175584896,
+      "loss": 2.1722,
       "step": 3715
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00033161902049060884,
-      "loss": 2.1289,
+      "epoch": 0.34,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001723419041000695,
+      "loss": 2.1328,
       "step": 3720
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00033137561862357225,
-      "loss": 2.1724,
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017257354644429002,
+      "loss": 2.1967,
       "step": 3725
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00033113187400831133,
-      "loss": 2.1556,
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017280518878851056,
+      "loss": 2.1932,
       "step": 3730
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003308877872807356,
-      "loss": 2.1442,
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017303683113273108,
+      "loss": 2.1755,
       "step": 3735
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00033064335907764743,
-      "loss": 2.1844,
+      "epoch": 0.35,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001732684734769516,
+      "loss": 2.2209,
       "step": 3740
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003303985900367399,
-      "loss": 2.1469,
+      "epoch": 0.35,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001735001158211721,
+      "loss": 2.1952,
       "step": 3745
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00033015348079659543,
-      "loss": 2.1677,
+      "epoch": 0.35,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017373175816539266,
+      "loss": 2.2179,
       "step": 3750
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00032990803199668396,
-      "loss": 2.1505,
+      "epoch": 0.35,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017396340050961317,
+      "loss": 2.2201,
       "step": 3755
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00032966224427736135,
-      "loss": 2.1784,
+      "epoch": 0.35,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001741950428538337,
+      "loss": 2.2669,
       "step": 3760
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003294161182798675,
-      "loss": 2.1666,
+      "epoch": 0.35,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001744266851980542,
+      "loss": 2.2419,
       "step": 3765
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00032916965464632503,
-      "loss": 2.184,
+      "epoch": 0.35,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017465832754227472,
+      "loss": 2.2413,
       "step": 3770
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003289228540197373,
-      "loss": 2.1779,
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017488996988649527,
+      "loss": 2.179,
       "step": 3775
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00032867571704398707,
-      "loss": 2.1613,
+      "epoch": 0.35,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017512161223071578,
+      "loss": 2.218,
       "step": 3780
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003284282443638343,
-      "loss": 2.1666,
+      "epoch": 0.35,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001753532545749363,
+      "loss": 2.147,
       "step": 3785
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003281804366249151,
-      "loss": 2.1374,
+      "epoch": 0.35,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017558489691915682,
+      "loss": 2.2382,
       "step": 3790
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003279322944737394,
-      "loss": 2.1905,
+      "epoch": 0.35,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017581653926337736,
+      "loss": 2.216,
       "step": 3795
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00032768381855768984,
-      "loss": 2.1411,
+      "epoch": 0.35,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017604818160759788,
+      "loss": 2.1772,
       "step": 3800
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00032743500952501966,
-      "loss": 2.1392,
+      "epoch": 0.35,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001762798239518184,
+      "loss": 2.2102,
       "step": 3805
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003271858680248513,
-      "loss": 2.1566,
+      "epoch": 0.35,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001765114662960389,
+      "loss": 2.2489,
       "step": 3810
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003269363947071744,
-      "loss": 2.1638,
+      "epoch": 0.35,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017674310864025945,
+      "loss": 2.1916,
       "step": 3815
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0003266865902228446,
-      "loss": 2.1398,
+      "epoch": 0.35,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017697475098447997,
+      "loss": 2.2097,
       "step": 3820
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00032643645522358115,
-      "loss": 2.1758,
+      "epoch": 0.35,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001772063933287005,
+      "loss": 2.1978,
       "step": 3825
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00032618599036196593,
-      "loss": 2.1802,
+      "epoch": 0.35,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000177438035672921,
+      "loss": 2.1986,
       "step": 3830
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003259351962914411,
-      "loss": 2.1981,
+      "epoch": 0.36,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017766967801714155,
+      "loss": 2.2032,
       "step": 3835
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003256840736663079,
-      "loss": 2.1491,
+      "epoch": 0.36,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017790132036136206,
+      "loss": 2.1963,
       "step": 3840
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00032543262314172465,
-      "loss": 2.1215,
+      "epoch": 0.36,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001781329627055826,
+      "loss": 2.1824,
       "step": 3845
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00032518084537370523,
-      "loss": 2.1359,
+      "epoch": 0.36,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001783646050498031,
+      "loss": 2.2,
       "step": 3850
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00032492874101911715,
-      "loss": 2.1479,
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017859624739402364,
+      "loss": 2.2288,
       "step": 3855
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00032467631073568006,
-      "loss": 2.1548,
+      "epoch": 0.36,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017882788973824415,
+      "loss": 2.2002,
       "step": 3860
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003244235551819639,
-      "loss": 2.1486,
+      "epoch": 0.36,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001790595320824647,
+      "loss": 2.2578,
       "step": 3865
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.0003241704750173871,
-      "loss": 2.141,
+      "epoch": 0.36,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017929117442668521,
+      "loss": 2.2042,
       "step": 3870
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003239170709022151,
-      "loss": 2.183,
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017952281677090573,
+      "loss": 2.1653,
       "step": 3875
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003236633434975586,
-      "loss": 2.1688,
+      "epoch": 0.36,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017975445911512625,
+      "loss": 2.234,
       "step": 3880
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003234092934653714,
-      "loss": 2.1295,
+      "epoch": 0.36,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001799861014593468,
+      "loss": 2.2356,
       "step": 3885
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003231549214684495,
-      "loss": 2.1869,
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001802177438035673,
+      "loss": 2.1619,
       "step": 3890
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003229002281704283,
-      "loss": 2.2002,
+      "epoch": 0.36,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018044938614778782,
+      "loss": 2.1554,
       "step": 3895
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00032264521423578193,
-      "loss": 2.1324,
+      "epoch": 0.36,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018068102849200834,
+      "loss": 2.1572,
       "step": 3900
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003223898803298207,
-      "loss": 2.2215,
+      "epoch": 0.36,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018091267083622886,
+      "loss": 2.2297,
       "step": 3905
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.0003221342271186901,
-      "loss": 2.1823,
+      "epoch": 0.36,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001811443131804494,
+      "loss": 2.2142,
       "step": 3910
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00032187825526936814,
-      "loss": 2.1501,
+      "epoch": 0.36,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018137595552466992,
+      "loss": 2.1869,
       "step": 3915
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00032162196544966457,
-      "loss": 2.1406,
+      "epoch": 0.36,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018160759786889043,
+      "loss": 2.2017,
       "step": 3920
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00032136535832821855,
-      "loss": 2.1025,
+      "epoch": 0.36,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018183924021311095,
+      "loss": 2.2791,
       "step": 3925
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003211084345744968,
-      "loss": 2.196,
+      "epoch": 0.36,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001820708825573315,
+      "loss": 2.2094,
       "step": 3930
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003208511948587926,
-      "loss": 2.163,
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000182302524901552,
+      "loss": 2.192,
       "step": 3935
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00032059363985222316,
-      "loss": 2.1585,
+      "epoch": 0.37,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018253416724577253,
+      "loss": 2.1932,
       "step": 3940
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00032033577022672847,
-      "loss": 2.1706,
+      "epoch": 0.37,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018276580958999304,
+      "loss": 2.1719,
       "step": 3945
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003200775866550692,
-      "loss": 2.1882,
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001829974519342136,
+      "loss": 2.2178,
       "step": 3950
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003198190898108251,
-      "loss": 2.1918,
+      "epoch": 0.37,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001832290942784341,
+      "loss": 2.2268,
       "step": 3955
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00031956028036839336,
-      "loss": 2.1635,
+      "epoch": 0.37,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018346073662265465,
+      "loss": 2.2227,
       "step": 3960
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00031930115900298653,
-      "loss": 2.1744,
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018369237896687514,
+      "loss": 2.1916,
       "step": 3965
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.000319041726390631,
-      "loss": 2.1833,
+      "epoch": 0.37,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018392402131109568,
+      "loss": 2.1937,
       "step": 3970
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003187819832081653,
-      "loss": 2.184,
+      "epoch": 0.37,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001841556636553162,
+      "loss": 2.2053,
       "step": 3975
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031852193013323795,
-      "loss": 2.1278,
+      "epoch": 0.37,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018438730599953674,
+      "loss": 2.2132,
       "step": 3980
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003182615678443063,
-      "loss": 2.1921,
+      "epoch": 0.37,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018461894834375726,
+      "loss": 2.1919,
       "step": 3985
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.00031800089702063413,
-      "loss": 2.1368,
+      "epoch": 0.37,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018485059068797777,
+      "loss": 2.1793,
       "step": 3990
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003177399183422903,
-      "loss": 2.1714,
+      "epoch": 0.37,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001850822330321983,
+      "loss": 2.1686,
       "step": 3995
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003174786324901467,
-      "loss": 2.1578,
+      "epoch": 0.37,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018531387537641883,
+      "loss": 2.2662,
       "step": 4000
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0003172170401458768,
-      "loss": 2.1785,
+      "epoch": 0.37,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018554551772063935,
+      "loss": 2.2001,
       "step": 4005
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00031695514199195357,
-      "loss": 2.1574,
+      "epoch": 0.37,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018577716006485987,
+      "loss": 2.156,
       "step": 4010
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00031669293871164784,
-      "loss": 2.1741,
+      "epoch": 0.37,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018600880240908038,
+      "loss": 2.1674,
       "step": 4015
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031643043098902645,
-      "loss": 2.1515,
+      "epoch": 0.37,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018624044475330093,
+      "loss": 2.1878,
       "step": 4020
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031616761950895066,
-      "loss": 2.1695,
+      "epoch": 0.37,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018647208709752144,
+      "loss": 2.1653,
       "step": 4025
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031590450495707387,
-      "loss": 2.1727,
+      "epoch": 0.37,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018670372944174196,
+      "loss": 2.1554,
       "step": 4030
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003156410880198406,
-      "loss": 2.1336,
+      "epoch": 0.37,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018693537178596247,
+      "loss": 2.1618,
       "step": 4035
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031537736938448387,
-      "loss": 2.1436,
+      "epoch": 0.37,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018716701413018302,
+      "loss": 2.2569,
       "step": 4040
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0003151133497390242,
-      "loss": 2.1573,
+      "epoch": 0.37,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018739865647440353,
+      "loss": 2.2222,
       "step": 4045
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00031484902977226703,
-      "loss": 2.0997,
+      "epoch": 0.38,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018763029881862405,
+      "loss": 2.2366,
       "step": 4050
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0003145844101738017,
-      "loss": 2.1623,
+      "epoch": 0.38,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00018786194116284457,
+      "loss": 2.1781,
       "step": 4055
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003143194916339987,
-      "loss": 2.1201,
+      "epoch": 0.38,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018809358350706508,
+      "loss": 2.2022,
       "step": 4060
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00031405427484400923,
-      "loss": 2.142,
+      "epoch": 0.38,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018832522585128563,
+      "loss": 2.1676,
       "step": 4065
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00031378876049576186,
-      "loss": 2.1642,
+      "epoch": 0.38,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018855686819550614,
+      "loss": 2.209,
       "step": 4070
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003135229492819619,
-      "loss": 2.1755,
+      "epoch": 0.38,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001887885105397267,
+      "loss": 2.1787,
       "step": 4075
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003132568418960889,
-      "loss": 2.161,
+      "epoch": 0.38,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018902015288394718,
+      "loss": 2.1485,
       "step": 4080
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003129904390323954,
-      "loss": 2.1652,
+      "epoch": 0.38,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018925179522816772,
+      "loss": 2.2332,
       "step": 4085
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0003127237413859046,
-      "loss": 2.1814,
+      "epoch": 0.38,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018948343757238824,
+      "loss": 2.2131,
       "step": 4090
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003124567496524087,
-      "loss": 2.1351,
+      "epoch": 0.38,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018971507991660878,
+      "loss": 2.1563,
       "step": 4095
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.1884765625,
-      "learning_rate": 0.00031218946452846733,
-      "loss": 2.167,
+      "epoch": 0.38,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018994672226082927,
+      "loss": 2.21,
       "step": 4100
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00031192188671140547,
-      "loss": 2.1597,
+      "epoch": 0.38,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019017836460504981,
+      "loss": 2.1953,
       "step": 4105
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003116540168993117,
-      "loss": 2.1689,
+      "epoch": 0.38,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019041000694927033,
+      "loss": 2.1509,
       "step": 4110
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00031138585579103647,
-      "loss": 2.1414,
+      "epoch": 0.38,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019064164929349087,
+      "loss": 2.1779,
       "step": 4115
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.0003111174040861901,
-      "loss": 2.1896,
+      "epoch": 0.38,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001908732916377114,
+      "loss": 2.1577,
       "step": 4120
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003108486624851411,
-      "loss": 2.1622,
+      "epoch": 0.38,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001911049339819319,
+      "loss": 2.1897,
       "step": 4125
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00031057963168901426,
-      "loss": 2.1455,
+      "epoch": 0.38,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019133657632615242,
+      "loss": 2.1855,
       "step": 4130
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003103103123996889,
-      "loss": 2.1765,
+      "epoch": 0.38,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019156821867037297,
+      "loss": 2.1912,
       "step": 4135
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.000310040705319797,
-      "loss": 2.1614,
+      "epoch": 0.38,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019179986101459348,
+      "loss": 2.1621,
       "step": 4140
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0003097708111527214,
-      "loss": 2.1687,
+      "epoch": 0.38,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000192031503358814,
+      "loss": 2.1209,
       "step": 4145
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0003095006306025937,
-      "loss": 2.1996,
+      "epoch": 0.38,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019226314570303452,
+      "loss": 2.1399,
       "step": 4150
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00030923016437429295,
-      "loss": 2.1228,
+      "epoch": 0.39,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019249478804725506,
+      "loss": 2.169,
       "step": 4155
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00030895941317344325,
-      "loss": 2.1984,
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019272643039147558,
+      "loss": 2.1754,
       "step": 4160
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003086883777064125,
-      "loss": 2.1719,
+      "epoch": 0.39,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001929580727356961,
+      "loss": 2.1952,
       "step": 4165
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0003084170586803098,
-      "loss": 2.1309,
+      "epoch": 0.39,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001931897150799166,
+      "loss": 2.2457,
       "step": 4170
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00030814545680298445,
-      "loss": 2.1502,
+      "epoch": 0.39,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019342135742413715,
+      "loss": 2.1663,
       "step": 4175
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0003078735727830234,
-      "loss": 2.1748,
+      "epoch": 0.39,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019365299976835767,
+      "loss": 2.179,
       "step": 4180
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003076014073297498,
-      "loss": 2.1534,
+      "epoch": 0.39,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019388464211257819,
+      "loss": 2.1456,
       "step": 4185
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.000307328961153221,
-      "loss": 2.169,
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001941162844567987,
+      "loss": 2.2509,
       "step": 4190
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00030705623496422673,
-      "loss": 2.2074,
+      "epoch": 0.39,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019434792680101922,
+      "loss": 2.1863,
       "step": 4195
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00030678322947428737,
-      "loss": 2.1453,
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019457956914523976,
+      "loss": 2.2068,
       "step": 4200
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00030650994539565174,
-      "loss": 2.1617,
+      "epoch": 0.39,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019481121148946028,
+      "loss": 2.2364,
       "step": 4205
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0003062363834412957,
-      "loss": 2.1778,
+      "epoch": 0.39,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019504285383368082,
+      "loss": 2.1954,
       "step": 4210
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00030596254432492,
-      "loss": 2.129,
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001952744961779013,
+      "loss": 2.1788,
       "step": 4215
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0003056884287609483,
-      "loss": 2.1792,
+      "epoch": 0.39,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019550613852212185,
+      "loss": 2.1561,
       "step": 4220
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00030541403746452584,
-      "loss": 2.192,
+      "epoch": 0.39,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019573778086634237,
+      "loss": 2.1987,
       "step": 4225
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00030513937115151697,
-      "loss": 2.1739,
+      "epoch": 0.39,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019596942321056291,
+      "loss": 2.1705,
       "step": 4230
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003048644305385035,
-      "loss": 2.1651,
+      "epoch": 0.39,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001962010655547834,
+      "loss": 2.1979,
       "step": 4235
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00030458921634278306,
-      "loss": 2.162,
+      "epoch": 0.39,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019643270789900395,
+      "loss": 2.1638,
       "step": 4240
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00030431372928236693,
-      "loss": 2.1479,
+      "epoch": 0.39,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019666435024322446,
+      "loss": 2.2221,
       "step": 4245
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0003040379700759782,
-      "loss": 2.1697,
+      "epoch": 0.39,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000196895992587445,
+      "loss": 2.182,
       "step": 4250
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0003037619394430501,
-      "loss": 2.1917,
+      "epoch": 0.39,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019712763493166552,
+      "loss": 2.1793,
       "step": 4255
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00030348563810372403,
-      "loss": 2.1322,
+      "epoch": 0.39,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019735927727588604,
+      "loss": 2.2096,
       "step": 4260
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00030320906677884735,
-      "loss": 2.2019,
+      "epoch": 0.4,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019759091962010656,
+      "loss": 2.1355,
       "step": 4265
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00030293222618997214,
-      "loss": 2.1271,
+      "epoch": 0.4,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001978225619643271,
+      "loss": 2.2425,
       "step": 4270
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00030265511705935264,
-      "loss": 2.1734,
+      "epoch": 0.4,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019805420430854762,
+      "loss": 2.1507,
       "step": 4275
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0003023777401099442,
-      "loss": 2.1741,
+      "epoch": 0.4,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019828584665276813,
+      "loss": 2.1743,
       "step": 4280
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0003021000960654001,
-      "loss": 2.1295,
+      "epoch": 0.4,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019851748899698865,
+      "loss": 2.1829,
       "step": 4285
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00030182218565007126,
-      "loss": 2.1086,
+      "epoch": 0.4,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001987491313412092,
+      "loss": 2.1904,
       "step": 4290
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00030154400958900304,
-      "loss": 2.1965,
+      "epoch": 0.4,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001989807736854297,
+      "loss": 2.2038,
       "step": 4295
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00030126556860793406,
-      "loss": 2.1739,
+      "epoch": 0.4,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019921241602965023,
+      "loss": 2.1752,
       "step": 4300
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00030098686343329393,
-      "loss": 2.1785,
+      "epoch": 0.4,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019944405837387074,
+      "loss": 2.2076,
       "step": 4305
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00030070789479220164,
-      "loss": 2.1833,
+      "epoch": 0.4,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001996757007180913,
+      "loss": 2.1591,
       "step": 4310
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00030042866341246357,
-      "loss": 2.1742,
+      "epoch": 0.4,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001999073430623118,
+      "loss": 2.1573,
       "step": 4315
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0003001491700225715,
-      "loss": 2.1997,
+      "epoch": 0.4,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001999999970575584,
+      "loss": 2.1913,
       "step": 4320
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002998694153517007,
-      "loss": 2.1871,
+      "epoch": 0.4,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019999997907597141,
+      "loss": 2.1399,
       "step": 4325
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00029958940012970835,
-      "loss": 2.1697,
+      "epoch": 0.4,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019999994474749014,
+      "loss": 2.2364,
       "step": 4330
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.000299309125087131,
-      "loss": 2.1582,
+      "epoch": 0.4,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001999998940721202,
+      "loss": 2.2035,
       "step": 4335
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00029902859095518345,
-      "loss": 2.1734,
+      "epoch": 0.4,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019999982704986988,
+      "loss": 2.1481,
       "step": 4340
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002987477984657562,
-      "loss": 2.1583,
+      "epoch": 0.4,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019999974368075017,
+      "loss": 2.1522,
       "step": 4345
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002984667483514139,
-      "loss": 2.1455,
+      "epoch": 0.4,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019999964396477462,
+      "loss": 2.2107,
       "step": 4350
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002981854413453931,
-      "loss": 2.1834,
+      "epoch": 0.4,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001999995279019596,
+      "loss": 2.1788,
       "step": 4355
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0002979038781816009,
-      "loss": 2.1747,
+      "epoch": 0.4,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019999939549232405,
+      "loss": 2.172,
       "step": 4360
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002976220595946125,
-      "loss": 2.1205,
+      "epoch": 0.4,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019999924673588963,
+      "loss": 2.1656,
       "step": 4365
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00029733998631966947,
-      "loss": 2.1561,
+      "epoch": 0.4,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019999908163268067,
+      "loss": 2.1675,
       "step": 4370
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002970576590926778,
-      "loss": 2.1635,
+      "epoch": 0.41,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001999989001827241,
+      "loss": 2.2213,
       "step": 4375
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002967750786502063,
-      "loss": 2.137,
+      "epoch": 0.41,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019999870238604966,
+      "loss": 2.2871,
       "step": 4380
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00029649224572948394,
-      "loss": 2.1344,
+      "epoch": 0.41,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001999984882426896,
+      "loss": 2.158,
       "step": 4385
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002962091610683988,
-      "loss": 2.1707,
+      "epoch": 0.41,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019999825775267901,
+      "loss": 2.1788,
       "step": 4390
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002959258254054956,
-      "loss": 2.1537,
+      "epoch": 0.41,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019999801091605553,
+      "loss": 2.1534,
       "step": 4395
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0002956422394799737,
-      "loss": 2.1807,
+      "epoch": 0.41,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019999774773285948,
+      "loss": 2.1908,
       "step": 4400
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002953584040316856,
-      "loss": 2.1838,
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019999746820313389,
+      "loss": 2.1904,
       "step": 4405
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00029507431980113483,
-      "loss": 2.1748,
+      "epoch": 0.41,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019999717232692453,
+      "loss": 2.1889,
       "step": 4410
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00029478998752947383,
-      "loss": 2.1416,
+      "epoch": 0.41,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019999686010427968,
+      "loss": 2.205,
       "step": 4415
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002945054079585021,
-      "loss": 2.1579,
+      "epoch": 0.41,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001999965315352504,
+      "loss": 2.1635,
       "step": 4420
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0002942205818306645,
-      "loss": 2.1964,
+      "epoch": 0.41,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019999618661989047,
+      "loss": 2.2079,
       "step": 4425
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002939355098890491,
-      "loss": 2.1878,
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019999582535825617,
+      "loss": 2.179,
       "step": 4430
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00029365019287738513,
-      "loss": 2.1768,
+      "epoch": 0.41,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001999954477504066,
+      "loss": 2.1595,
       "step": 4435
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002933646315400414,
-      "loss": 2.1224,
+      "epoch": 0.41,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001999950537964035,
+      "loss": 2.2086,
       "step": 4440
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.000293078826622024,
-      "loss": 2.1614,
+      "epoch": 0.41,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019999464349631122,
+      "loss": 2.1875,
       "step": 4445
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002927927788689746,
-      "loss": 2.1222,
+      "epoch": 0.41,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019999421685019692,
+      "loss": 2.2034,
       "step": 4450
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00029250648902716815,
-      "loss": 2.1392,
+      "epoch": 0.41,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019999377385813025,
+      "loss": 2.1773,
       "step": 4455
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002922199578435116,
-      "loss": 2.1278,
+      "epoch": 0.41,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001999933145201837,
+      "loss": 2.1655,
       "step": 4460
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00029193318606554126,
-      "loss": 2.1057,
+      "epoch": 0.41,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001999928388364323,
+      "loss": 2.1382,
       "step": 4465
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0002916461744414213,
-      "loss": 2.1295,
+      "epoch": 0.41,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019999234680695383,
+      "loss": 2.2021,
       "step": 4470
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00029135892371994133,
-      "loss": 2.1465,
+      "epoch": 0.41,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019999183843182872,
+      "loss": 2.1529,
       "step": 4475
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002910714346505151,
-      "loss": 2.1864,
+      "epoch": 0.42,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019999131371114008,
+      "loss": 2.2648,
       "step": 4480
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00029078370798317786,
-      "loss": 2.1191,
+      "epoch": 0.42,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019999077264497373,
+      "loss": 2.1969,
       "step": 4485
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00029049574446858513,
-      "loss": 2.1343,
+      "epoch": 0.42,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019999021523341802,
+      "loss": 2.1676,
       "step": 4490
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00029020754485801,
-      "loss": 2.1347,
+      "epoch": 0.42,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019998964147656414,
+      "loss": 2.1717,
       "step": 4495
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0002899191099033417,
-      "loss": 2.1863,
+      "epoch": 0.42,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019998905137450586,
+      "loss": 2.1795,
       "step": 4500
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002896304403570834,
-      "loss": 2.1284,
+      "epoch": 0.42,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019998844492733964,
+      "loss": 2.1815,
       "step": 4505
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002893415369723502,
-      "loss": 2.1697,
+      "epoch": 0.42,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019998782213516463,
+      "loss": 2.1967,
       "step": 4510
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002890524005028674,
-      "loss": 2.1384,
+      "epoch": 0.42,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001999871829980826,
+      "loss": 2.1367,
       "step": 4515
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0002887630317029684,
-      "loss": 2.1453,
+      "epoch": 0.42,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019998652751619807,
+      "loss": 2.1558,
       "step": 4520
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0002884734313275927,
-      "loss": 2.1537,
+      "epoch": 0.42,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019998585568961822,
+      "loss": 2.19,
       "step": 4525
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00028818360013228384,
-      "loss": 2.1534,
+      "epoch": 0.42,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001999851675184528,
+      "loss": 2.2187,
       "step": 4530
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002878935388731878,
-      "loss": 2.2001,
+      "epoch": 0.42,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001999844630028143,
+      "loss": 2.1355,
       "step": 4535
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002876032483070505,
-      "loss": 2.1625,
+      "epoch": 0.42,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000199983742142818,
+      "loss": 2.2004,
       "step": 4540
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00028731272919121643,
-      "loss": 2.1471,
+      "epoch": 0.42,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001999830049385816,
+      "loss": 2.1719,
       "step": 4545
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00028702198228362595,
-      "loss": 2.1882,
+      "epoch": 0.42,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001999822513902257,
+      "loss": 2.1831,
       "step": 4550
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.000286731008342814,
-      "loss": 2.1713,
+      "epoch": 0.42,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001999814814978734,
+      "loss": 2.228,
       "step": 4555
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002864398081279078,
-      "loss": 2.2102,
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019998069526165065,
+      "loss": 2.178,
       "step": 4560
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0002861483823986248,
-      "loss": 2.1601,
+      "epoch": 0.42,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019997989268168595,
+      "loss": 2.136,
       "step": 4565
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00028585673191527087,
-      "loss": 2.1745,
+      "epoch": 0.42,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019997907375811044,
+      "loss": 2.1463,
       "step": 4570
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00028556485743873815,
-      "loss": 2.1265,
+      "epoch": 0.42,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019997823849105805,
+      "loss": 2.2038,
       "step": 4575
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002852727597305034,
-      "loss": 2.1508,
+      "epoch": 0.42,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019997738688066528,
+      "loss": 2.2066,
       "step": 4580
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002849804395526254,
-      "loss": 2.1352,
+      "epoch": 0.42,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019997651892707136,
+      "loss": 2.1287,
       "step": 4585
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00028468789766774373,
-      "loss": 2.2117,
+      "epoch": 0.43,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019997563463041817,
+      "loss": 2.1929,
       "step": 4590
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002843951348390761,
-      "loss": 2.1643,
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019997473399085028,
+      "loss": 2.1554,
       "step": 4595
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002841021518304167,
-      "loss": 2.1697,
+      "epoch": 0.43,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001999738170085149,
+      "loss": 2.1911,
       "step": 4600
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00028380894940613436,
-      "loss": 2.1483,
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019997288368356193,
+      "loss": 2.1829,
       "step": 4605
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00028351552833116996,
-      "loss": 2.1538,
+      "epoch": 0.43,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019997193401614393,
+      "loss": 2.2242,
       "step": 4610
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002832218893710353,
-      "loss": 2.1349,
+      "epoch": 0.43,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019997096800641618,
+      "loss": 2.1572,
       "step": 4615
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00028292803329181006,
-      "loss": 2.1824,
+      "epoch": 0.43,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019996998565453654,
+      "loss": 2.1301,
       "step": 4620
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002826339608601409,
-      "loss": 2.1932,
+      "epoch": 0.43,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001999689869606656,
+      "loss": 2.1655,
       "step": 4625
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00028233967284323864,
-      "loss": 2.1542,
+      "epoch": 0.43,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019996797192496667,
+      "loss": 2.1681,
       "step": 4630
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00028204517000887663,
-      "loss": 2.1764,
+      "epoch": 0.43,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019996694054760566,
+      "loss": 2.1445,
       "step": 4635
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002817504531253885,
-      "loss": 2.1645,
+      "epoch": 0.43,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019996589282875112,
+      "loss": 2.2438,
       "step": 4640
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.00028145552296166666,
-      "loss": 2.1308,
+      "epoch": 0.43,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019996482876857434,
+      "loss": 2.2005,
       "step": 4645
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.0002811603802871596,
-      "loss": 2.1788,
+      "epoch": 0.43,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001999637483672493,
+      "loss": 2.1571,
       "step": 4650
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.0002808650258718704,
-      "loss": 2.1833,
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019996265162495254,
+      "loss": 2.2073,
       "step": 4655
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00028056946048635455,
-      "loss": 2.1497,
+      "epoch": 0.43,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019996153854186344,
+      "loss": 2.1957,
       "step": 4660
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.0002802736849017179,
-      "loss": 2.148,
+      "epoch": 0.43,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019996040911816386,
+      "loss": 2.1765,
       "step": 4665
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0002799776998896147,
-      "loss": 2.1204,
+      "epoch": 0.43,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019995926335403846,
+      "loss": 2.1613,
       "step": 4670
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00027968150622224565,
-      "loss": 2.1634,
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019995810124967455,
+      "loss": 2.2025,
       "step": 4675
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002793851046723557,
-      "loss": 2.1469,
+      "epoch": 0.43,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019995692280526213,
+      "loss": 2.1785,
       "step": 4680
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00027908849601323226,
-      "loss": 2.1578,
+      "epoch": 0.43,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019995572802099376,
+      "loss": 2.2174,
       "step": 4685
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.000278791681018703,
-      "loss": 2.165,
+      "epoch": 0.43,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019995451689706483,
+      "loss": 2.1824,
       "step": 4690
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00027849466046313384,
-      "loss": 2.1491,
+      "epoch": 0.44,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001999532894336732,
+      "loss": 2.1251,
       "step": 4695
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002781974351214272,
-      "loss": 2.1494,
+      "epoch": 0.44,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019995204563101967,
+      "loss": 2.236,
       "step": 4700
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002779000057690196,
-      "loss": 2.1238,
+      "epoch": 0.44,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001999507854893075,
+      "loss": 2.1889,
       "step": 4705
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002776023731818798,
-      "loss": 2.187,
+      "epoch": 0.44,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019994950900874265,
+      "loss": 2.1262,
       "step": 4710
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.000277304538136507,
-      "loss": 2.1481,
+      "epoch": 0.44,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019994821618953384,
+      "loss": 2.2048,
       "step": 4715
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002770065014099282,
-      "loss": 2.1226,
+      "epoch": 0.44,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019994690703189236,
+      "loss": 2.2246,
       "step": 4720
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0002767082637796969,
-      "loss": 2.1702,
+      "epoch": 0.44,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019994558153603226,
+      "loss": 2.1826,
       "step": 4725
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00027640982602389065,
-      "loss": 2.1044,
+      "epoch": 0.44,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001999442397021702,
+      "loss": 2.153,
       "step": 4730
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002761111889211091,
-      "loss": 2.1993,
+      "epoch": 0.44,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001999428815305255,
+      "loss": 2.1816,
       "step": 4735
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.000275812353250472,
-      "loss": 2.1658,
+      "epoch": 0.44,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001999415070213202,
+      "loss": 2.1952,
       "step": 4740
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00027551331979161714,
-      "loss": 2.1546,
+      "epoch": 0.44,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000199940116174779,
+      "loss": 2.1924,
       "step": 4745
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00027521408932469816,
-      "loss": 2.1709,
+      "epoch": 0.44,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019993870899112924,
+      "loss": 2.2239,
       "step": 4750
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00027491466263038304,
-      "loss": 2.1624,
+      "epoch": 0.44,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019993728547060098,
+      "loss": 2.2003,
       "step": 4755
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00027461504048985135,
-      "loss": 2.1396,
+      "epoch": 0.44,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001999358456134269,
+      "loss": 2.2031,
       "step": 4760
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00027431522368479277,
-      "loss": 2.1456,
+      "epoch": 0.44,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001999343894198424,
+      "loss": 2.1906,
       "step": 4765
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002740152129974048,
-      "loss": 2.1567,
+      "epoch": 0.44,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019993291689008547,
+      "loss": 2.1544,
       "step": 4770
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00027371500921039065,
-      "loss": 2.1809,
+      "epoch": 0.44,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019993142802439686,
+      "loss": 2.1915,
       "step": 4775
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00027341461310695744,
-      "loss": 2.1142,
+      "epoch": 0.44,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019992992282301993,
+      "loss": 2.2201,
       "step": 4780
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00027311402547081407,
-      "loss": 2.1756,
+      "epoch": 0.44,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001999284012862008,
+      "loss": 2.159,
       "step": 4785
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00027281324708616897,
-      "loss": 2.1675,
+      "epoch": 0.44,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001999268634141881,
+      "loss": 2.222,
       "step": 4790
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00027251227873772825,
-      "loss": 2.1433,
+      "epoch": 0.44,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001999253092072333,
+      "loss": 2.1971,
       "step": 4795
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00027221112121069384,
-      "loss": 2.168,
+      "epoch": 0.44,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019992373866559038,
+      "loss": 2.1909,
       "step": 4800
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00027190977529076087,
-      "loss": 2.2056,
+      "epoch": 0.45,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019992215178951618,
+      "loss": 2.1398,
       "step": 4805
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00027160824176411633,
-      "loss": 2.1628,
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019992054857927003,
+      "loss": 2.1848,
       "step": 4810
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002713065214174363,
-      "loss": 2.1301,
+      "epoch": 0.45,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019991892903511402,
+      "loss": 2.2205,
       "step": 4815
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002710046150378846,
-      "loss": 2.1805,
+      "epoch": 0.45,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001999172931573129,
+      "loss": 2.201,
       "step": 4820
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0002707025234131101,
-      "loss": 2.1282,
+      "epoch": 0.45,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001999156409461341,
+      "loss": 2.1704,
       "step": 4825
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00027040024733124523,
-      "loss": 2.159,
+      "epoch": 0.45,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001999139724018477,
+      "loss": 2.1392,
       "step": 4830
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00027009778758090345,
-      "loss": 2.1703,
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019991228752472645,
+      "loss": 2.1777,
       "step": 4835
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00026979514495117745,
-      "loss": 2.1592,
+      "epoch": 0.45,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019991058631504575,
+      "loss": 2.1991,
       "step": 4840
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002694923202316371,
-      "loss": 2.1478,
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019990886877308372,
+      "loss": 2.2011,
       "step": 4845
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00026918931421232725,
-      "loss": 2.1303,
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019990713489912115,
+      "loss": 2.1067,
       "step": 4850
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00026888612768376577,
-      "loss": 2.1344,
+      "epoch": 0.45,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001999053846934414,
+      "loss": 2.1788,
       "step": 4855
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.00026858276143694145,
-      "loss": 2.1288,
+      "epoch": 0.45,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001999036181563307,
+      "loss": 2.1945,
       "step": 4860
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00026827921626331207,
-      "loss": 2.1749,
+      "epoch": 0.45,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019990183528807771,
+      "loss": 2.1792,
       "step": 4865
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.000267975492954802,
-      "loss": 2.1649,
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019990003608897388,
+      "loss": 2.1714,
       "step": 4870
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00026767159230380055,
-      "loss": 2.1581,
+      "epoch": 0.45,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001998982205593134,
+      "loss": 2.2763,
       "step": 4875
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002673675151031596,
-      "loss": 2.1654,
+      "epoch": 0.45,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019989638869939298,
+      "loss": 2.2614,
       "step": 4880
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.19140625,
-      "learning_rate": 0.0002670632621461916,
-      "loss": 2.1446,
+      "epoch": 0.45,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001998945405095121,
+      "loss": 2.1713,
       "step": 4885
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002667588342266677,
-      "loss": 2.1597,
+      "epoch": 0.45,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019989267598997286,
+      "loss": 2.1848,
       "step": 4890
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00026645423213881523,
-      "loss": 2.1272,
+      "epoch": 0.45,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019989079514108008,
+      "loss": 2.1693,
       "step": 4895
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002661494566773162,
-      "loss": 2.1691,
+      "epoch": 0.45,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019988889796314124,
+      "loss": 2.1722,
       "step": 4900
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00026584450863730474,
-      "loss": 2.1536,
+      "epoch": 0.45,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001998869844564664,
+      "loss": 2.1608,
       "step": 4905
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002655393888143654,
-      "loss": 2.1609,
+      "epoch": 0.45,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019988505462136845,
+      "loss": 2.1989,
       "step": 4910
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00026523409800453075,
-      "loss": 2.1631,
+      "epoch": 0.46,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001998831084581628,
+      "loss": 2.1835,
       "step": 4915
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00026492863700427945,
-      "loss": 2.1416,
+      "epoch": 0.46,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019988114596716756,
+      "loss": 2.1744,
       "step": 4920
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002646230066105343,
-      "loss": 2.1406,
+      "epoch": 0.46,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001998791671487036,
+      "loss": 2.1912,
       "step": 4925
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00026431720762065995,
-      "loss": 2.1789,
+      "epoch": 0.46,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019987717200309435,
+      "loss": 2.189,
       "step": 4930
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002640112408324608,
-      "loss": 2.1909,
+      "epoch": 0.46,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019987516053066598,
+      "loss": 2.2138,
       "step": 4935
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002637051070441793,
-      "loss": 2.138,
+      "epoch": 0.46,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001998731327317473,
+      "loss": 2.1849,
       "step": 4940
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00026339880705449325,
-      "loss": 2.1424,
+      "epoch": 0.46,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019987108860666978,
+      "loss": 2.1809,
       "step": 4945
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002630923416625143,
-      "loss": 2.1703,
+      "epoch": 0.46,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019986902815576757,
+      "loss": 2.149,
       "step": 4950
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002627857116677855,
-      "loss": 2.1225,
+      "epoch": 0.46,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019986695137937753,
+      "loss": 2.1963,
       "step": 4955
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0002624789178702796,
-      "loss": 2.1497,
+      "epoch": 0.46,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019986485827783908,
+      "loss": 2.1514,
       "step": 4960
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0002621719610703961,
-      "loss": 2.1998,
+      "epoch": 0.46,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019986274885149444,
+      "loss": 2.1602,
       "step": 4965
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00026186484206896046,
-      "loss": 2.1573,
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001998606231006884,
+      "loss": 2.1876,
       "step": 4970
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00026155756166722086,
-      "loss": 2.169,
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019985848102576848,
+      "loss": 2.1446,
       "step": 4975
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002612501206668468,
-      "loss": 2.1583,
+      "epoch": 0.46,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001998563226270848,
+      "loss": 2.1832,
       "step": 4980
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00026094251986992647,
-      "loss": 2.1491,
+      "epoch": 0.46,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019985414790499022,
+      "loss": 2.1019,
       "step": 4985
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00026063476007896525,
-      "loss": 2.1644,
+      "epoch": 0.46,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019985195685984027,
+      "loss": 2.285,
       "step": 4990
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00026032684209688323,
-      "loss": 2.1645,
+      "epoch": 0.46,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019984974949199304,
+      "loss": 2.1708,
       "step": 4995
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0002600187667270132,
-      "loss": 2.1521,
+      "epoch": 0.46,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019984752580180944,
+      "loss": 2.1638,
       "step": 5000
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00025971053477309854,
-      "loss": 2.2114,
+      "epoch": 0.46,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019984528578965294,
+      "loss": 2.1987,
       "step": 5005
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00025940214703929116,
-      "loss": 2.1751,
+      "epoch": 0.46,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019984302945588972,
+      "loss": 2.1655,
       "step": 5010
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00025909360433014935,
-      "loss": 2.1664,
+      "epoch": 0.46,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001998407568008886,
+      "loss": 2.2035,
       "step": 5015
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0002587849074506358,
-      "loss": 2.2008,
+      "epoch": 0.47,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019983846782502113,
+      "loss": 2.2407,
       "step": 5020
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.189453125,
-      "learning_rate": 0.00025847605720611533,
-      "loss": 2.1275,
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019983616252866146,
+      "loss": 2.251,
       "step": 5025
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002581670544023531,
-      "loss": 2.16,
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019983384091218644,
+      "loss": 2.1923,
       "step": 5030
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00025785789984551186,
-      "loss": 2.1662,
+      "epoch": 0.47,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001998315029759756,
+      "loss": 2.1942,
       "step": 5035
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0002575485943421506,
-      "loss": 2.1949,
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019982914872041107,
+      "loss": 2.1522,
       "step": 5040
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002572391386992221,
-      "loss": 2.1761,
+      "epoch": 0.47,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019982677814587776,
+      "loss": 2.1934,
       "step": 5045
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002569295337240708,
-      "loss": 2.1522,
+      "epoch": 0.47,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019982439125276313,
+      "loss": 2.1948,
       "step": 5050
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00025661978022443056,
-      "loss": 2.153,
+      "epoch": 0.47,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001998219880414574,
+      "loss": 2.149,
       "step": 5055
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.000256309879008423,
-      "loss": 2.1931,
+      "epoch": 0.47,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019981956851235338,
+      "loss": 2.1605,
       "step": 5060
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0002559998308845549,
-      "loss": 2.1806,
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019981713266584663,
+      "loss": 2.2068,
       "step": 5065
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002556896366617166,
-      "loss": 2.1721,
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019981468050233534,
+      "loss": 2.1852,
       "step": 5070
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00025537929714917925,
-      "loss": 2.1222,
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019981221202222034,
+      "loss": 2.1843,
       "step": 5075
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0002550688131565932,
-      "loss": 2.1832,
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019980972722590513,
+      "loss": 2.151,
       "step": 5080
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00025475818549398584,
-      "loss": 2.1428,
+      "epoch": 0.47,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019980722611379593,
+      "loss": 2.1672,
       "step": 5085
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00025444741497175925,
-      "loss": 2.1774,
+      "epoch": 0.47,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019980470868630158,
+      "loss": 2.18,
       "step": 5090
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00025413650240068824,
-      "loss": 2.1533,
+      "epoch": 0.47,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019980217494383365,
+      "loss": 2.1718,
       "step": 5095
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.1904296875,
-      "learning_rate": 0.0002538254485919183,
-      "loss": 2.1717,
+      "epoch": 0.47,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019979962488680624,
+      "loss": 2.2245,
       "step": 5100
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00025351425435696316,
-      "loss": 2.1586,
+      "epoch": 0.47,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019979705851563625,
+      "loss": 2.1741,
       "step": 5105
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00025320292050770327,
+      "epoch": 0.47,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001997944758307432,
       "loss": 2.1615,
       "step": 5110
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002528914478563831,
-      "loss": 2.1589,
+      "epoch": 0.47,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001997918768325493,
+      "loss": 2.195,
       "step": 5115
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.0002525798372156092,
-      "loss": 2.1327,
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019978926152147938,
+      "loss": 2.1514,
       "step": 5120
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002522680893983482,
-      "loss": 2.1334,
+      "epoch": 0.47,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019978662989796097,
+      "loss": 2.2567,
       "step": 5125
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002519562052179247,
-      "loss": 2.1986,
+      "epoch": 0.48,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019978398196242425,
+      "loss": 2.2452,
       "step": 5130
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.0002516441854880189,
-      "loss": 2.1506,
+      "epoch": 0.48,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001997813177153021,
+      "loss": 2.1754,
       "step": 5135
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00025133203102266474,
-      "loss": 2.1611,
+      "epoch": 0.48,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019977863715703,
+      "loss": 2.1967,
       "step": 5140
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00025101974263624764,
-      "loss": 2.1453,
+      "epoch": 0.48,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019977594028804617,
+      "loss": 2.1984,
       "step": 5145
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00025070732114350233,
-      "loss": 2.1207,
+      "epoch": 0.48,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019977322710879148,
+      "loss": 2.1317,
       "step": 5150
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.0002503947673595111,
-      "loss": 2.14,
+      "epoch": 0.48,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001997704976197094,
+      "loss": 2.2034,
       "step": 5155
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00025008208209970095,
-      "loss": 2.1846,
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019976775182124613,
+      "loss": 2.0941,
       "step": 5160
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002497692661798422,
-      "loss": 2.1757,
+      "epoch": 0.48,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019976498971385057,
+      "loss": 2.2342,
       "step": 5165
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00024945632041604586,
-      "loss": 2.1336,
+      "epoch": 0.48,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019976221129797418,
+      "loss": 2.1878,
       "step": 5170
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00024914324562476197,
-      "loss": 2.1451,
+      "epoch": 0.48,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019975941657407123,
+      "loss": 2.1837,
       "step": 5175
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002488300426227768,
-      "loss": 2.1725,
+      "epoch": 0.48,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019975660554259845,
+      "loss": 2.1513,
       "step": 5180
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.0002485167122272115,
-      "loss": 2.1376,
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019975377820401544,
+      "loss": 2.1807,
       "step": 5185
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.0002482032552555192,
-      "loss": 2.179,
+      "epoch": 0.48,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019975093455878434,
+      "loss": 2.2179,
       "step": 5190
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00024788967252548365,
-      "loss": 2.1259,
+      "epoch": 0.48,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019974807460737003,
+      "loss": 2.1163,
       "step": 5195
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002475759648552164,
-      "loss": 2.2117,
+      "epoch": 0.48,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019974519835024005,
+      "loss": 2.1572,
       "step": 5200
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002472621330631551,
-      "loss": 2.1336,
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019974230578786447,
+      "loss": 2.1585,
       "step": 5205
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002469481779680611,
-      "loss": 2.1223,
+      "epoch": 0.48,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019973939692071623,
+      "loss": 2.2028,
       "step": 5210
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002466341003890176,
-      "loss": 2.143,
+      "epoch": 0.48,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019973647174927086,
+      "loss": 2.1509,
       "step": 5215
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00024631990114542716,
-      "loss": 2.1461,
+      "epoch": 0.48,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019973353027400643,
+      "loss": 2.1841,
       "step": 5220
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.0002460055810570101,
-      "loss": 2.1894,
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019973057249540385,
+      "loss": 2.1802,
       "step": 5225
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00024569114094380153,
-      "loss": 2.1477,
+      "epoch": 0.48,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019972759841394664,
+      "loss": 2.1982,
       "step": 5230
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00024537658162615,
-      "loss": 2.1513,
+      "epoch": 0.49,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019972460803012094,
+      "loss": 2.1893,
       "step": 5235
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1923828125,
-      "learning_rate": 0.0002450619039247151,
-      "loss": 2.1068,
+      "epoch": 0.49,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019972160134441554,
+      "loss": 2.1709,
       "step": 5240
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.0002447471086604651,
-      "loss": 2.1434,
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019971857835732204,
+      "loss": 2.1343,
       "step": 5245
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00024443219665467496,
-      "loss": 2.1596,
+      "epoch": 0.49,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019971553906933455,
+      "loss": 2.1597,
       "step": 5250
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0002441171687289245,
-      "loss": 2.1451,
+      "epoch": 0.49,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019971248348094988,
+      "loss": 2.1509,
       "step": 5255
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00024380202570509557,
-      "loss": 2.1864,
+      "epoch": 0.49,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019970941159266756,
+      "loss": 2.1629,
       "step": 5260
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00024348676840537064,
-      "loss": 2.15,
+      "epoch": 0.49,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001997063234049897,
+      "loss": 2.2062,
       "step": 5265
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00024317139765223008,
-      "loss": 2.1246,
+      "epoch": 0.49,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001997032189184212,
+      "loss": 2.1538,
       "step": 5270
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00024285591426845044,
-      "loss": 2.1653,
+      "epoch": 0.49,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001997000981334695,
+      "loss": 2.1841,
       "step": 5275
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.0002425403190771018,
-      "loss": 2.1562,
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019969696105064474,
+      "loss": 2.1257,
       "step": 5280
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00024222461290154647,
-      "loss": 2.1499,
+      "epoch": 0.49,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019969380767045978,
+      "loss": 2.1335,
       "step": 5285
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00024190879656543578,
-      "loss": 2.1019,
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019969063799343004,
+      "loss": 2.2365,
       "step": 5290
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00024159287089270875,
-      "loss": 2.1572,
+      "epoch": 0.49,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019968745202007373,
+      "loss": 2.1585,
       "step": 5295
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1982421875,
-      "learning_rate": 0.00024127683670758954,
-      "loss": 2.1302,
+      "epoch": 0.49,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019968424975091157,
+      "loss": 2.2144,
       "step": 5300
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1875,
-      "learning_rate": 0.00024096069483458545,
-      "loss": 2.1094,
+      "epoch": 0.49,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019968103118646712,
+      "loss": 2.186,
       "step": 5305
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00024064444609848474,
-      "loss": 2.1348,
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001996777963272665,
+      "loss": 2.2009,
       "step": 5310
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00024032809132435446,
-      "loss": 2.1543,
+      "epoch": 0.49,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019967454517383845,
+      "loss": 2.1674,
       "step": 5315
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00024001163133753823,
-      "loss": 2.1001,
+      "epoch": 0.49,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001996712777267145,
+      "loss": 2.1758,
       "step": 5320
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.1943359375,
-      "learning_rate": 0.00023969506696365432,
-      "loss": 2.1323,
+      "epoch": 0.49,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019966799398642873,
+      "loss": 2.2056,
       "step": 5325
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00023937839902859315,
-      "loss": 2.1543,
+      "epoch": 0.49,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019966469395351797,
+      "loss": 2.1771,
       "step": 5330
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00023906162835851555,
-      "loss": 2.1806,
+      "epoch": 0.49,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019966137762852162,
+      "loss": 2.1792,
       "step": 5335
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00023874475577985007,
-      "loss": 2.1394,
+      "epoch": 0.49,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019965804501198187,
+      "loss": 2.1963,
       "step": 5340
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002384277821192914,
-      "loss": 2.1701,
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019965469610444344,
+      "loss": 2.1008,
       "step": 5345
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00023811070820379783,
-      "loss": 2.1586,
+      "epoch": 0.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019965133090645376,
+      "loss": 2.1731,
       "step": 5350
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.0002377935348605893,
-      "loss": 2.14,
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019964794941856298,
+      "loss": 2.1741,
       "step": 5355
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00023747626291714498,
-      "loss": 2.1441,
+      "epoch": 0.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019964455164132386,
+      "loss": 2.1969,
       "step": 5360
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00023715889320120135,
-      "loss": 2.1567,
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019964113757529184,
+      "loss": 2.211,
       "step": 5365
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.193359375,
-      "learning_rate": 0.00023684142654075004,
-      "loss": 2.1384,
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019963770722102497,
+      "loss": 2.2007,
       "step": 5370
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00023652386376403562,
-      "loss": 2.1509,
+      "epoch": 0.5,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019963426057908405,
+      "loss": 2.1368,
       "step": 5375
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.1953125,
-      "learning_rate": 0.00023620620569955332,
-      "loss": 2.1294,
+      "epoch": 0.5,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001996307976500325,
+      "loss": 2.2399,
       "step": 5380
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.0002358884531760469,
-      "loss": 2.1489,
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019962731843443634,
+      "loss": 2.1931,
       "step": 5385
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00023557060702250677,
-      "loss": 2.1334,
+      "epoch": 0.5,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001996238229328644,
+      "loss": 2.2267,
       "step": 5390
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00023525266806816744,
-      "loss": 2.1313,
+      "epoch": 0.5,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019962031114588804,
+      "loss": 2.1471,
       "step": 5395
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00023493463714250568,
-      "loss": 2.1266,
+      "epoch": 0.5,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001996167830740813,
+      "loss": 2.2632,
       "step": 5400
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 2.150402307510376,
-      "eval_runtime": 188.8418,
-      "eval_samples_per_second": 25.752,
-      "eval_steps_per_second": 3.22,
-      "step": 5403
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.000234616515075238,
-      "loss": 2.1241,
+      "epoch": 0.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019961323871802098,
+      "loss": 2.1925,
       "step": 5405
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00023429830269631886,
-      "loss": 2.1299,
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019960967807828642,
+      "loss": 2.2276,
       "step": 5410
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.19921875,
-      "learning_rate": 0.00023398000083593829,
-      "loss": 2.106,
+      "epoch": 0.5,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001996061011554597,
+      "loss": 2.2401,
       "step": 5415
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00023366161032451982,
-      "loss": 2.1316,
+      "epoch": 0.5,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019960250795012554,
+      "loss": 2.1842,
       "step": 5420
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00023334313199271808,
-      "loss": 2.1283,
+      "epoch": 0.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019959889846287126,
+      "loss": 2.2122,
       "step": 5425
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00023302456667141704,
-      "loss": 2.1331,
+      "epoch": 0.5,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000199595272694287,
+      "loss": 2.2182,
       "step": 5430
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00023270591519172743,
-      "loss": 2.1087,
+      "epoch": 0.5,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019959163064496536,
+      "loss": 2.2152,
       "step": 5435
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00023238717838498495,
-      "loss": 2.1158,
+      "epoch": 0.5,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019958797231550175,
+      "loss": 2.169,
       "step": 5440
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00023206835708274777,
-      "loss": 2.1277,
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019958429770649421,
+      "loss": 2.1873,
       "step": 5445
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00023174945211679458,
-      "loss": 2.1279,
+      "epoch": 0.51,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019958060681854337,
+      "loss": 2.2116,
       "step": 5450
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.0002314304643191222,
-      "loss": 2.1279,
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019957689965225263,
+      "loss": 2.196,
       "step": 5455
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00023111139452194378,
-      "loss": 2.1358,
+      "epoch": 0.51,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019957317620822798,
+      "loss": 2.1744,
       "step": 5460
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.197265625,
-      "learning_rate": 0.00023079224355768618,
-      "loss": 2.0907,
+      "epoch": 0.51,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019956943648707806,
+      "loss": 2.209,
       "step": 5465
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.2001953125,
-      "learning_rate": 0.00023047301225898826,
-      "loss": 2.0924,
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019956568048941427,
+      "loss": 2.2503,
       "step": 5470
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00023015370145869812,
-      "loss": 2.1152,
+      "epoch": 0.51,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001995619082158505,
+      "loss": 2.191,
       "step": 5475
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002298343119898716,
-      "loss": 2.1402,
+      "epoch": 0.51,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019955811966700348,
+      "loss": 2.1616,
       "step": 5480
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00022951484468576965,
-      "loss": 2.1549,
+      "epoch": 0.51,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019955431484349248,
+      "loss": 2.1801,
       "step": 5485
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.0002291953003798563,
-      "loss": 2.0924,
+      "epoch": 0.51,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001995504937459395,
+      "loss": 2.1944,
       "step": 5490
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.201171875,
-      "learning_rate": 0.00022887567990579637,
-      "loss": 2.1107,
+      "epoch": 0.51,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019954665637496915,
+      "loss": 2.2079,
       "step": 5495
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.2021484375,
-      "learning_rate": 0.00022855598409745356,
-      "loss": 2.117,
+      "epoch": 0.51,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001995428027312087,
+      "loss": 2.2402,
       "step": 5500
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00022823621378888803,
-      "loss": 2.1362,
+      "epoch": 0.51,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019953893281528812,
+      "loss": 2.1976,
       "step": 5505
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0002279163698143544,
-      "loss": 2.0987,
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019953504662784005,
+      "loss": 2.2668,
       "step": 5510
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00022759645300829924,
-      "loss": 2.122,
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019953114416949971,
+      "loss": 2.1224,
       "step": 5515
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00022727646420535934,
-      "loss": 2.0972,
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001995272254409051,
+      "loss": 2.2209,
       "step": 5520
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00022695640424035926,
-      "loss": 2.0972,
+      "epoch": 0.51,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019952329044269673,
+      "loss": 2.2549,
       "step": 5525
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00022663627394830935,
-      "loss": 2.1113,
+      "epoch": 0.51,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001995193391755179,
+      "loss": 2.1786,
       "step": 5530
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00022631607416440311,
-      "loss": 2.0969,
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019951537164001452,
+      "loss": 2.1716,
       "step": 5535
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00022599580572401568,
-      "loss": 2.1307,
+      "epoch": 0.51,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019951138783683514,
+      "loss": 2.1738,
       "step": 5540
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00022567546946270105,
-      "loss": 2.088,
+      "epoch": 0.51,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000199507387766631,
+      "loss": 2.2095,
       "step": 5545
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00022535506621619046,
-      "loss": 2.1045,
+      "epoch": 0.51,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019950337143005597,
+      "loss": 2.1764,
       "step": 5550
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00022503459682038962,
-      "loss": 2.0976,
+      "epoch": 0.51,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001994993388277666,
+      "loss": 2.1664,
       "step": 5555
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00022471406211137693,
-      "loss": 2.125,
+      "epoch": 0.52,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019949528996042216,
+      "loss": 2.2112,
       "step": 5560
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.00022439346292540114,
-      "loss": 2.0827,
+      "epoch": 0.52,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019949122482868442,
+      "loss": 2.2315,
       "step": 5565
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00022407280009887934,
-      "loss": 2.1119,
+      "epoch": 0.52,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019948714343321792,
+      "loss": 2.1666,
       "step": 5570
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00022375207446839453,
-      "loss": 2.1057,
+      "epoch": 0.52,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001994830457746899,
+      "loss": 2.1849,
       "step": 5575
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00022343128687069363,
-      "loss": 2.1151,
+      "epoch": 0.52,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019947893185377012,
+      "loss": 2.1708,
       "step": 5580
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00022311043814268513,
-      "loss": 2.1059,
+      "epoch": 0.52,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019947480167113117,
+      "loss": 2.1967,
       "step": 5585
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002227895291214371,
-      "loss": 2.0808,
+      "epoch": 0.52,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019947065522744813,
+      "loss": 2.1776,
       "step": 5590
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00022246856064417494,
-      "loss": 2.0955,
+      "epoch": 0.52,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019946649252339886,
+      "loss": 2.171,
       "step": 5595
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00022214753354827908,
-      "loss": 2.1399,
+      "epoch": 0.52,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001994623135596638,
+      "loss": 2.1667,
       "step": 5600
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00022182644867128283,
-      "loss": 2.1297,
+      "epoch": 0.52,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001994581183369261,
+      "loss": 2.176,
       "step": 5605
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00022150530685087044,
-      "loss": 2.1046,
+      "epoch": 0.52,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019945390685587158,
+      "loss": 2.1962,
       "step": 5610
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00022118410892487455,
-      "loss": 2.1181,
+      "epoch": 0.52,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019944967911718863,
+      "loss": 2.2323,
       "step": 5615
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00022086285573127423,
-      "loss": 2.1405,
+      "epoch": 0.52,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019944543512156836,
+      "loss": 2.1422,
       "step": 5620
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00022054154810819277,
-      "loss": 2.1203,
+      "epoch": 0.52,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019944117486970456,
+      "loss": 2.168,
       "step": 5625
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00022022018689389535,
-      "loss": 2.1112,
+      "epoch": 0.52,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019943689836229362,
+      "loss": 2.1591,
       "step": 5630
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00021989877292678714,
-      "loss": 2.1051,
+      "epoch": 0.52,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019943260560003468,
+      "loss": 2.1535,
       "step": 5635
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00021957730704541084,
-      "loss": 2.0685,
+      "epoch": 0.52,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019942829658362938,
+      "loss": 2.1755,
       "step": 5640
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00021925579008844453,
-      "loss": 2.1139,
+      "epoch": 0.52,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019942397131378218,
+      "loss": 2.1964,
       "step": 5645
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002189342228946997,
-      "loss": 2.096,
+      "epoch": 0.52,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001994196297912001,
+      "loss": 2.2564,
       "step": 5650
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.00021861260630311863,
-      "loss": 2.1158,
+      "epoch": 0.52,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019941527201659286,
+      "loss": 2.1885,
       "step": 5655
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00021829094115277287,
-      "loss": 2.1403,
+      "epoch": 0.52,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019941089799067275,
+      "loss": 2.195,
       "step": 5660
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002179692282828603,
-      "loss": 2.1077,
+      "epoch": 0.52,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019940650771415493,
+      "loss": 2.1804,
       "step": 5665
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00021764746853270356,
-      "loss": 2.0722,
+      "epoch": 0.53,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019940210118775696,
+      "loss": 2.1499,
       "step": 5670
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00021732566274174726,
-      "loss": 2.146,
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001993976784121992,
+      "loss": 2.1301,
       "step": 5675
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00021700381174955658,
-      "loss": 2.1231,
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019939323938820462,
+      "loss": 2.2017,
       "step": 5680
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00021668191639581425,
-      "loss": 2.1488,
+      "epoch": 0.53,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019938878411649894,
+      "loss": 2.1195,
       "step": 5685
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00021635997752031893,
-      "loss": 2.0689,
+      "epoch": 0.53,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019938431259781034,
+      "loss": 2.1901,
       "step": 5690
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0002160379959629828,
-      "loss": 2.1011,
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019937982483286986,
+      "loss": 2.2306,
       "step": 5695
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00021571597256382917,
-      "loss": 2.0967,
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019937532082241107,
+      "loss": 2.2096,
       "step": 5700
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.000215393908162991,
-      "loss": 2.0587,
+      "epoch": 0.53,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019937080056717027,
+      "loss": 2.1816,
       "step": 5705
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00021507180360070765,
-      "loss": 2.091,
+      "epoch": 0.53,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019936626406788635,
+      "loss": 2.2181,
       "step": 5710
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.0002147496597173237,
-      "loss": 2.0682,
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001993617113253009,
+      "loss": 2.1061,
       "step": 5715
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00021442747735328608,
-      "loss": 2.0907,
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019935714234015816,
+      "loss": 2.1042,
       "step": 5720
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.1962890625,
-      "learning_rate": 0.00021410525734914218,
-      "loss": 2.0557,
+      "epoch": 0.53,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000199352557113205,
+      "loss": 2.1815,
       "step": 5725
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00021378300054553754,
-      "loss": 2.1218,
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019934795564519097,
+      "loss": 2.1752,
       "step": 5730
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00021346070778321393,
-      "loss": 2.1368,
+      "epoch": 0.53,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001993433379368683,
+      "loss": 2.1828,
       "step": 5735
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00021313837990300657,
-      "loss": 2.1157,
+      "epoch": 0.53,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019933870398899177,
+      "loss": 2.2148,
       "step": 5740
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.00021281601774584256,
-      "loss": 2.1341,
+      "epoch": 0.53,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019933405380231895,
+      "loss": 2.2041,
       "step": 5745
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00021249362215273836,
-      "loss": 2.1038,
+      "epoch": 0.53,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019932938737760996,
+      "loss": 2.1874,
       "step": 5750
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.0002121711939647977,
-      "loss": 2.1066,
+      "epoch": 0.53,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019932470471562764,
+      "loss": 2.2331,
       "step": 5755
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0002118487340232092,
-      "loss": 2.0586,
+      "epoch": 0.53,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019932000581713747,
+      "loss": 2.1795,
       "step": 5760
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0002115262431692445,
-      "loss": 2.0977,
+      "epoch": 0.53,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019931529068290755,
+      "loss": 2.1301,
       "step": 5765
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0002112037222442558,
-      "loss": 2.1219,
+      "epoch": 0.53,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019931055931370864,
+      "loss": 2.1838,
       "step": 5770
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.22265625,
-      "learning_rate": 0.0002108811720896738,
-      "loss": 2.1481,
+      "epoch": 0.54,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019930581171031424,
+      "loss": 2.1947,
       "step": 5775
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.203125,
-      "learning_rate": 0.00021055859354700537,
-      "loss": 2.097,
+      "epoch": 0.54,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019930104787350037,
+      "loss": 2.1723,
       "step": 5780
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00021023598745783145,
-      "loss": 2.1014,
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001992962678040458,
+      "loss": 2.24,
       "step": 5785
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00020991335466380502,
-      "loss": 2.1153,
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019929147150273192,
+      "loss": 2.2488,
       "step": 5790
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2041015625,
-      "learning_rate": 0.0002095906960066486,
-      "loss": 2.0794,
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019928665897034277,
+      "loss": 2.1651,
       "step": 5795
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00020926801232815204,
-      "loss": 2.1081,
+      "epoch": 0.54,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019928183020766506,
+      "loss": 2.2131,
       "step": 5800
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00020894530447017077,
-      "loss": 2.1151,
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019927698521548815,
+      "loss": 2.1263,
       "step": 5805
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00020862257327462297,
-      "loss": 2.1014,
+      "epoch": 0.54,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000199272123994604,
+      "loss": 2.2273,
       "step": 5810
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00020829981958348813,
-      "loss": 2.1248,
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019926724654580733,
+      "loss": 2.1712,
       "step": 5815
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.000207977044238804,
-      "loss": 2.1307,
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001992623528698954,
+      "loss": 2.147,
       "step": 5820
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0002076542480826651,
-      "loss": 2.114,
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001992574429676682,
+      "loss": 2.2098,
       "step": 5825
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00020733143195722002,
-      "loss": 2.1148,
+      "epoch": 0.54,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019925251683992835,
+      "loss": 2.2039,
       "step": 5830
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00020700859670466981,
-      "loss": 2.1523,
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001992475744874811,
+      "loss": 2.1757,
       "step": 5835
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2060546875,
-      "learning_rate": 0.000206685743167265,
-      "loss": 2.1059,
+      "epoch": 0.54,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001992426159111344,
+      "loss": 2.2252,
       "step": 5840
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.0002063628721873041,
-      "loss": 2.1124,
+      "epoch": 0.54,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001992376411116988,
+      "loss": 2.187,
       "step": 5845
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.000206039984607131,
-      "loss": 2.1202,
+      "epoch": 0.54,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019923265008998752,
+      "loss": 2.213,
       "step": 5850
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00020571708126913295,
-      "loss": 2.1126,
+      "epoch": 0.54,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001992276428468165,
+      "loss": 2.1854,
       "step": 5855
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00020539416301573832,
-      "loss": 2.11,
+      "epoch": 0.54,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019922261938300415,
+      "loss": 2.1402,
       "step": 5860
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00020507123068941438,
-      "loss": 2.1197,
+      "epoch": 0.54,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019921757969937177,
+      "loss": 2.2142,
       "step": 5865
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00020474828513266505,
-      "loss": 2.1129,
+      "epoch": 0.54,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001992125237967431,
+      "loss": 2.1431,
       "step": 5870
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002044253271880289,
-      "loss": 2.1116,
+      "epoch": 0.54,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001992074516759447,
+      "loss": 2.221,
       "step": 5875
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00020410235769807668,
-      "loss": 2.0783,
+      "epoch": 0.54,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019920236333780564,
+      "loss": 2.1922,
       "step": 5880
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00020377937750540946,
-      "loss": 2.1268,
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019919725878315777,
+      "loss": 2.2017,
       "step": 5885
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00020345638745265592,
-      "loss": 2.1024,
+      "epoch": 0.55,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019919213801283548,
+      "loss": 2.1486,
       "step": 5890
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0002031333883824707,
-      "loss": 2.0943,
+      "epoch": 0.55,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019918700102767586,
+      "loss": 2.197,
       "step": 5895
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00020281038113753198,
-      "loss": 2.1171,
+      "epoch": 0.55,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019918184782851864,
+      "loss": 2.17,
       "step": 5900
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00020248736656053914,
-      "loss": 2.1008,
+      "epoch": 0.55,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001991766784162062,
+      "loss": 2.1589,
       "step": 5905
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00020216434549421073,
-      "loss": 2.0961,
+      "epoch": 0.55,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019917149279158366,
+      "loss": 2.1901,
       "step": 5910
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0002018413187812823,
-      "loss": 2.09,
+      "epoch": 0.55,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019916629095549862,
+      "loss": 2.1579,
       "step": 5915
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00020151828726450396,
-      "loss": 2.1428,
+      "epoch": 0.55,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019916107290880147,
+      "loss": 2.2051,
       "step": 5920
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00020119525178663862,
-      "loss": 2.1225,
+      "epoch": 0.55,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019915583865234517,
+      "loss": 2.1991,
       "step": 5925
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.0002008722131904593,
-      "loss": 2.1276,
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019915058818698538,
+      "loss": 2.1797,
       "step": 5930
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00020054917231874722,
-      "loss": 2.0879,
+      "epoch": 0.55,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019914532151358034,
+      "loss": 2.1601,
       "step": 5935
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00020022613001428953,
-      "loss": 2.1028,
+      "epoch": 0.55,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019914003863299103,
+      "loss": 2.1936,
       "step": 5940
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001999030871198772,
-      "loss": 2.1489,
+      "epoch": 0.55,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019913473954608107,
+      "loss": 2.1413,
       "step": 5945
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0001995800444783027,
-      "loss": 2.1126,
+      "epoch": 0.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019912942425371661,
+      "loss": 2.1734,
       "step": 5950
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00019925700293235764,
-      "loss": 2.1148,
+      "epoch": 0.55,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001991240927567666,
+      "loss": 2.1766,
       "step": 5955
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00019893396332483108,
-      "loss": 2.1339,
+      "epoch": 0.55,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019911874505610256,
+      "loss": 2.1456,
       "step": 5960
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00019861092649850682,
-      "loss": 2.106,
+      "epoch": 0.55,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019911338115259866,
+      "loss": 2.1284,
       "step": 5965
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00019828789329616163,
-      "loss": 2.1064,
+      "epoch": 0.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019910800104713173,
+      "loss": 2.1551,
       "step": 5970
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00019796486456056245,
-      "loss": 2.1179,
+      "epoch": 0.55,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019910260474058128,
+      "loss": 2.1556,
       "step": 5975
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00019764184113446487,
-      "loss": 2.1449,
+      "epoch": 0.55,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019909719223382944,
+      "loss": 2.1812,
       "step": 5980
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00019731882386061055,
-      "loss": 2.0949,
+      "epoch": 0.55,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019909176352776094,
+      "loss": 2.2374,
       "step": 5985
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00019699581358172507,
-      "loss": 2.0989,
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001990863186232632,
+      "loss": 2.1847,
       "step": 5990
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00019667281114051592,
-      "loss": 2.1064,
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019908085752122636,
+      "loss": 2.1396,
       "step": 5995
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0001963498173796698,
-      "loss": 2.144,
+      "epoch": 0.56,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019907538022254312,
+      "loss": 2.1796,
       "step": 6000
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00019602683314185105,
-      "loss": 2.1148,
+      "epoch": 0.56,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019906988672810881,
+      "loss": 2.1054,
       "step": 6005
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00019570385926969902,
-      "loss": 2.1193,
+      "epoch": 0.56,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019906437703882151,
+      "loss": 2.1307,
       "step": 6010
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001953808966058263,
-      "loss": 2.1129,
+      "epoch": 0.56,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001990588511555818,
+      "loss": 2.1996,
       "step": 6015
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00019505794599281577,
-      "loss": 2.128,
+      "epoch": 0.56,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019905330907929304,
+      "loss": 2.2247,
       "step": 6020
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001947350082732192,
-      "loss": 2.116,
+      "epoch": 0.56,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001990477508108612,
+      "loss": 2.1691,
       "step": 6025
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00019441208428955467,
-      "loss": 2.0824,
+      "epoch": 0.56,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019904217635119485,
+      "loss": 2.2253,
       "step": 6030
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00019408917488430438,
-      "loss": 2.0985,
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019903658570120524,
+      "loss": 2.1687,
       "step": 6035
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00019376628089991246,
-      "loss": 2.1691,
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001990309788618063,
+      "loss": 2.1832,
       "step": 6040
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00019344340317878293,
-      "loss": 2.1106,
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019902535583391458,
+      "loss": 2.2267,
       "step": 6045
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00019312054256327729,
-      "loss": 2.115,
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019901971661844922,
+      "loss": 2.1529,
       "step": 6050
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001927976998957124,
-      "loss": 2.1005,
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001990140612163321,
+      "loss": 2.1948,
       "step": 6055
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00019247487601835835,
-      "loss": 2.1078,
+      "epoch": 0.56,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019900838962848768,
+      "loss": 2.1743,
       "step": 6060
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00019215207177343617,
-      "loss": 2.1068,
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019900270185584313,
+      "loss": 2.1919,
       "step": 6065
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0001918292880031157,
-      "loss": 2.0899,
+      "epoch": 0.56,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019899699789932814,
+      "loss": 2.1828,
       "step": 6070
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001915065255495134,
-      "loss": 2.1286,
-      "step": 6075
+      "epoch": 0.56,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019899127775987525,
+      "loss": 2.1522,
+      "step": 6075
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00019118378525468996,
-      "loss": 2.1497,
+      "epoch": 0.56,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001989855414384194,
+      "loss": 2.18,
       "step": 6080
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001908610679606484,
-      "loss": 2.0953,
+      "epoch": 0.56,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019897978893589836,
+      "loss": 2.2101,
       "step": 6085
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00019053837450933175,
-      "loss": 2.1246,
+      "epoch": 0.56,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019897402025325254,
+      "loss": 2.2278,
       "step": 6090
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00019021570574262076,
-      "loss": 2.124,
+      "epoch": 0.56,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019896823539142483,
+      "loss": 2.1861,
       "step": 6095
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00018989306250233177,
-      "loss": 2.1067,
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019896243435136093,
+      "loss": 2.1936,
       "step": 6100
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00018957044563021455,
-      "loss": 2.1097,
+      "epoch": 0.57,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019895661713400915,
+      "loss": 2.1793,
       "step": 6105
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00018924785596795014,
-      "loss": 2.1008,
+      "epoch": 0.57,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001989507837403204,
+      "loss": 2.2026,
       "step": 6110
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.20703125,
-      "learning_rate": 0.00018892529435714854,
-      "loss": 2.0854,
+      "epoch": 0.57,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019894493417124828,
+      "loss": 2.1459,
       "step": 6115
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.205078125,
-      "learning_rate": 0.0001886027616393465,
-      "loss": 2.1099,
+      "epoch": 0.57,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019893906842774895,
+      "loss": 2.1758,
       "step": 6120
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001882802586560055,
-      "loss": 2.0994,
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001989331865107814,
+      "loss": 2.1951,
       "step": 6125
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00018795778624850935,
-      "loss": 2.1248,
+      "epoch": 0.57,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019892728842130702,
+      "loss": 2.1699,
       "step": 6130
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00018763534525816224,
-      "loss": 2.125,
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019892137416029,
+      "loss": 2.1865,
       "step": 6135
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.0001873129365261862,
-      "loss": 2.0986,
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019891544372869713,
+      "loss": 2.1988,
       "step": 6140
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001869905608937192,
-      "loss": 2.1644,
+      "epoch": 0.57,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001989094971274979,
+      "loss": 2.1463,
       "step": 6145
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001866682192018129,
-      "loss": 2.0913,
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019890353435766435,
+      "loss": 2.1485,
       "step": 6150
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001863459122914304,
-      "loss": 2.1306,
+      "epoch": 0.57,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019889755542017124,
+      "loss": 2.2024,
       "step": 6155
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00018602364100344382,
-      "loss": 2.1249,
+      "epoch": 0.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019889156031599592,
+      "loss": 2.1613,
       "step": 6160
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00018570140617863275,
-      "loss": 2.1154,
+      "epoch": 0.57,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001988855490461184,
+      "loss": 2.1898,
       "step": 6165
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00018537920865768136,
-      "loss": 2.1396,
+      "epoch": 0.57,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019887952161152134,
+      "loss": 2.1673,
       "step": 6170
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001850570492811767,
-      "loss": 2.1388,
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019887347801319006,
+      "loss": 2.2118,
       "step": 6175
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.000184734928889606,
-      "loss": 2.1193,
+      "epoch": 0.57,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019886741825211245,
+      "loss": 2.1794,
       "step": 6180
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00018441284832335496,
-      "loss": 2.1099,
+      "epoch": 0.57,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019886134232927912,
+      "loss": 2.128,
       "step": 6185
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00018409080842270558,
-      "loss": 2.1267,
+      "epoch": 0.57,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019885525024568334,
+      "loss": 2.1873,
       "step": 6190
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00018376881002783355,
-      "loss": 2.0948,
+      "epoch": 0.57,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001988491420023209,
+      "loss": 2.1835,
       "step": 6195
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00018344685397880613,
-      "loss": 2.0771,
+      "epoch": 0.57,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019884301760019036,
+      "loss": 2.1646,
       "step": 6200
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00018312494111558033,
-      "loss": 2.0696,
+      "epoch": 0.57,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019883687704029285,
+      "loss": 2.178,
       "step": 6205
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001828030722780005,
-      "loss": 2.0965,
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001988307203236322,
+      "loss": 2.1591,
       "step": 6210
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00018248124830579608,
-      "loss": 2.1058,
+      "epoch": 0.58,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019882454745121474,
+      "loss": 2.1151,
       "step": 6215
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001821594700385794,
-      "loss": 2.1195,
+      "epoch": 0.58,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019881835842404966,
+      "loss": 2.1471,
       "step": 6220
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0001818377383158435,
-      "loss": 2.1058,
+      "epoch": 0.58,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019881215324314862,
+      "loss": 2.192,
       "step": 6225
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001815160539769602,
-      "loss": 2.1028,
+      "epoch": 0.58,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019880593190952595,
+      "loss": 2.12,
       "step": 6230
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00018119441786117763,
-      "loss": 2.1162,
+      "epoch": 0.58,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001987996944241987,
+      "loss": 2.1543,
       "step": 6235
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00018087283080761793,
-      "loss": 2.1048,
+      "epoch": 0.58,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001987934407881865,
+      "loss": 2.1637,
       "step": 6240
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00018055129365527543,
-      "loss": 2.1006,
+      "epoch": 0.58,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019878717100251154,
+      "loss": 2.2296,
       "step": 6245
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00018022980724301423,
-      "loss": 2.0785,
+      "epoch": 0.58,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019878088506819884,
+      "loss": 2.1445,
       "step": 6250
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017990837240956602,
-      "loss": 2.0936,
+      "epoch": 0.58,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019877458298627593,
+      "loss": 2.1591,
       "step": 6255
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00017958698999352794,
-      "loss": 2.1514,
+      "epoch": 0.58,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019876826475777296,
+      "loss": 2.1666,
       "step": 6260
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00017926566083336035,
-      "loss": 2.0913,
+      "epoch": 0.58,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019876193038372283,
+      "loss": 2.1503,
       "step": 6265
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.0001789443857673848,
-      "loss": 2.104,
+      "epoch": 0.58,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019875557986516094,
+      "loss": 2.1884,
       "step": 6270
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00017862316563378144,
-      "loss": 2.1296,
+      "epoch": 0.58,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019874921320312547,
+      "loss": 2.1998,
       "step": 6275
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00017830200127058756,
-      "loss": 2.11,
+      "epoch": 0.58,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019874283039865713,
+      "loss": 2.1369,
       "step": 6280
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.0001779808935156944,
-      "loss": 2.0885,
+      "epoch": 0.58,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019873643145279935,
+      "loss": 2.1651,
       "step": 6285
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0001776598432068459,
-      "loss": 2.0768,
+      "epoch": 0.58,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001987300163665981,
+      "loss": 2.2258,
       "step": 6290
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017733885118163595,
-      "loss": 2.114,
+      "epoch": 0.58,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019872358514110206,
+      "loss": 2.2156,
       "step": 6295
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017701791827750654,
-      "loss": 2.0768,
+      "epoch": 0.58,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001987171377773626,
+      "loss": 2.1914,
       "step": 6300
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001766970453317452,
-      "loss": 2.1185,
+      "epoch": 0.58,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019871067427643362,
+      "loss": 2.2089,
       "step": 6305
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00017637623318148324,
-      "loss": 2.0814,
+      "epoch": 0.58,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001987041946393717,
+      "loss": 2.1339,
       "step": 6310
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017605548266369321,
-      "loss": 2.1291,
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019869769886723604,
+      "loss": 2.1573,
       "step": 6315
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00017573479461518708,
-      "loss": 2.1129,
+      "epoch": 0.59,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019869118696108854,
+      "loss": 2.1841,
       "step": 6320
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001754141698726135,
-      "loss": 2.099,
+      "epoch": 0.59,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019868465892199363,
+      "loss": 2.1555,
       "step": 6325
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00017509360927245622,
-      "loss": 2.1072,
+      "epoch": 0.59,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001986781147510185,
+      "loss": 2.2064,
       "step": 6330
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00017477311365103162,
-      "loss": 2.1402,
+      "epoch": 0.59,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019867155444923293,
+      "loss": 2.1249,
       "step": 6335
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001744526838444865,
-      "loss": 2.0942,
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001986649780177093,
+      "loss": 2.1624,
       "step": 6340
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00017413232068879596,
-      "loss": 2.1288,
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019865838545752264,
+      "loss": 2.1947,
       "step": 6345
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001738120250197612,
-      "loss": 2.086,
+      "epoch": 0.59,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019865177676975067,
+      "loss": 2.1751,
       "step": 6350
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017349179767300738,
-      "loss": 2.0644,
+      "epoch": 0.59,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019864515195547363,
+      "loss": 2.1976,
       "step": 6355
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.0001731716394839815,
-      "loss": 2.0885,
+      "epoch": 0.59,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019863851101577454,
+      "loss": 2.2264,
       "step": 6360
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00017285155128794993,
-      "loss": 2.1015,
+      "epoch": 0.59,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019863185395173897,
+      "loss": 2.1852,
       "step": 6365
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00017253153391999656,
-      "loss": 2.0885,
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019862518076445515,
+      "loss": 2.1409,
       "step": 6370
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00017221158821502055,
-      "loss": 2.133,
-      "step": 6375
+      "epoch": 0.59,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001986184914550139,
+      "loss": 2.1435,
+      "step": 6375
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.000171891715007734,
-      "loss": 2.0947,
+      "epoch": 0.59,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019861178602450876,
+      "loss": 2.1624,
       "step": 6380
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00017157191513265984,
-      "loss": 2.1282,
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019860506447403587,
+      "loss": 2.185,
       "step": 6385
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.0001712521894241298,
-      "loss": 2.0747,
+      "epoch": 0.59,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019859832680469393,
+      "loss": 2.2671,
       "step": 6390
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00017093253871628206,
-      "loss": 2.108,
+      "epoch": 0.59,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001985915730175844,
+      "loss": 2.0822,
       "step": 6395
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00017061296384305913,
-      "loss": 2.1235,
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019858480311381123,
+      "loss": 2.1295,
       "step": 6400
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001702934656382056,
-      "loss": 2.0921,
+      "epoch": 0.59,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019857801709448122,
+      "loss": 2.125,
       "step": 6405
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00016997404493526618,
-      "loss": 2.147,
+      "epoch": 0.59,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019857121496070362,
+      "loss": 2.2151,
       "step": 6410
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00016965470256758328,
-      "loss": 2.1055,
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019856439671359031,
+      "loss": 2.1514,
       "step": 6415
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00016933543936829506,
-      "loss": 2.1194,
+      "epoch": 0.59,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019855756235425594,
+      "loss": 2.1534,
       "step": 6420
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00016901625617033283,
-      "loss": 2.1074,
+      "epoch": 0.6,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019855071188381763,
+      "loss": 2.1329,
       "step": 6425
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0001686971538064196,
-      "loss": 2.0911,
+      "epoch": 0.6,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019854384530339531,
+      "loss": 2.1554,
       "step": 6430
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001683781331090672,
-      "loss": 2.1184,
+      "epoch": 0.6,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001985369626141114,
+      "loss": 2.1305,
       "step": 6435
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00016805919491057463,
-      "loss": 2.131,
+      "epoch": 0.6,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019853006381709103,
+      "loss": 2.2181,
       "step": 6440
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00016774034004302525,
-      "loss": 2.1019,
+      "epoch": 0.6,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019852314891346192,
+      "loss": 2.1835,
       "step": 6445
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001674215693382855,
-      "loss": 2.0983,
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019851621790435445,
+      "loss": 2.1905,
       "step": 6450
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00016710288362800196,
-      "loss": 2.1232,
+      "epoch": 0.6,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019850927079090164,
+      "loss": 2.1715,
       "step": 6455
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.22265625,
-      "learning_rate": 0.00016678428374359964,
-      "loss": 2.1064,
+      "epoch": 0.6,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001985023075742391,
+      "loss": 2.1422,
       "step": 6460
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00016646577051627936,
-      "loss": 2.089,
+      "epoch": 0.6,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001984953282555051,
+      "loss": 2.1595,
       "step": 6465
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00016614734477701605,
-      "loss": 2.1403,
+      "epoch": 0.6,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001984883328358406,
+      "loss": 2.1361,
       "step": 6470
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00016582900735655649,
-      "loss": 2.1017,
+      "epoch": 0.6,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019848132131638905,
+      "loss": 2.1894,
       "step": 6475
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00016551075908541693,
-      "loss": 2.1058,
+      "epoch": 0.6,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019847429369829666,
+      "loss": 2.1524,
       "step": 6480
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001651926007938809,
-      "loss": 2.1283,
+      "epoch": 0.6,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019846724998271226,
+      "loss": 2.1709,
       "step": 6485
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00016487453331199737,
-      "loss": 2.0858,
+      "epoch": 0.6,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001984601901707872,
+      "loss": 2.2066,
       "step": 6490
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00016455655746957834,
-      "loss": 2.0983,
+      "epoch": 0.6,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001984531142636756,
+      "loss": 2.2124,
       "step": 6495
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00016423867409619687,
-      "loss": 2.1359,
+      "epoch": 0.6,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019844602226253412,
+      "loss": 2.2054,
       "step": 6500
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001639208840211844,
-      "loss": 2.1151,
+      "epoch": 0.6,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001984389141685221,
+      "loss": 2.1896,
       "step": 6505
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001636031880736293,
-      "loss": 2.108,
+      "epoch": 0.6,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019843178998280153,
+      "loss": 2.2624,
       "step": 6510
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.0001632855870823743,
-      "loss": 2.0985,
+      "epoch": 0.6,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001984246497065369,
+      "loss": 2.1674,
       "step": 6515
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00016296808187601437,
-      "loss": 2.0595,
+      "epoch": 0.6,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001984174933408955,
+      "loss": 2.206,
       "step": 6520
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00016265067328289447,
-      "loss": 2.0978,
+      "epoch": 0.6,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019841032088704715,
+      "loss": 2.1365,
       "step": 6525
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00016233336213110773,
-      "loss": 2.1,
+      "epoch": 0.61,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019840313234616432,
+      "loss": 2.1966,
       "step": 6530
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00016201614924849287,
-      "loss": 2.1164,
+      "epoch": 0.61,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001983959277194221,
+      "loss": 2.1516,
       "step": 6535
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00016169903546263238,
-      "loss": 2.0785,
+      "epoch": 0.61,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019838870700799826,
+      "loss": 2.1915,
       "step": 6540
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00016138202160085005,
-      "loss": 2.161,
+      "epoch": 0.61,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019838147021307312,
+      "loss": 2.1796,
       "step": 6545
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00016106510849020914,
-      "loss": 2.1207,
+      "epoch": 0.61,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001983742173358297,
+      "loss": 2.2526,
       "step": 6550
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00016074829695750996,
-      "loss": 2.1187,
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019836694837745364,
+      "loss": 2.1876,
       "step": 6555
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0001604315878292879,
-      "loss": 2.1297,
+      "epoch": 0.61,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019835966333913314,
+      "loss": 2.1832,
       "step": 6560
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00016011498193181108,
-      "loss": 2.1197,
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001983523622220591,
+      "loss": 2.1784,
       "step": 6565
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00015979848009107828,
-      "loss": 2.1178,
+      "epoch": 0.61,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019834504502742504,
+      "loss": 2.1593,
       "step": 6570
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00015948208313281694,
-      "loss": 2.1231,
+      "epoch": 0.61,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019833771175642708,
+      "loss": 2.1687,
       "step": 6575
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0001591657918824808,
-      "loss": 2.0752,
+      "epoch": 0.61,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019833036241026397,
+      "loss": 2.1963,
       "step": 6580
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00015884960716524787,
-      "loss": 2.1095,
+      "epoch": 0.61,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001983229969901371,
+      "loss": 2.1028,
       "step": 6585
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00015853352980601802,
-      "loss": 2.1059,
+      "epoch": 0.61,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019831561549725054,
+      "loss": 2.1799,
       "step": 6590
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00015821756062941132,
-      "loss": 2.113,
+      "epoch": 0.61,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001983082179328109,
+      "loss": 2.2115,
       "step": 6595
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.0001579017004597654,
-      "loss": 2.1061,
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001983008042980274,
+      "loss": 2.1922,
       "step": 6600
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001575859501211336,
-      "loss": 2.1151,
+      "epoch": 0.61,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019829337459411202,
+      "loss": 2.2001,
       "step": 6605
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001572703104372826,
-      "loss": 2.0957,
+      "epoch": 0.61,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019828592882227926,
+      "loss": 2.1922,
       "step": 6610
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.22265625,
-      "learning_rate": 0.00015695478223169053,
-      "loss": 2.1191,
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019827846698374627,
+      "loss": 2.1075,
       "step": 6615
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00015663936632754464,
-      "loss": 2.0812,
+      "epoch": 0.61,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019827098907973279,
+      "loss": 2.1045,
       "step": 6620
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015632406354773918,
-      "loss": 2.1666,
+      "epoch": 0.61,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001982634951114613,
+      "loss": 2.1808,
       "step": 6625
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00015600887471487322,
-      "loss": 2.0995,
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019825598508015676,
+      "loss": 2.203,
       "step": 6630
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015569380065124863,
-      "loss": 2.1118,
+      "epoch": 0.61,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001982484589870469,
+      "loss": 2.1302,
       "step": 6635
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001553788421788678,
-      "loss": 2.1483,
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019824091683336192,
+      "loss": 2.1907,
       "step": 6640
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0001550640001194317,
-      "loss": 2.076,
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001982333586203348,
+      "loss": 2.1585,
       "step": 6645
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001547492752943373,
-      "loss": 2.1103,
+      "epoch": 0.62,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019822578434920102,
+      "loss": 2.1765,
       "step": 6650
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00015443466852467594,
-      "loss": 2.1059,
+      "epoch": 0.62,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019821819402119877,
+      "loss": 2.1872,
       "step": 6655
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00015412018063123091,
-      "loss": 2.1148,
+      "epoch": 0.62,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019821058763756883,
+      "loss": 2.193,
       "step": 6660
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001538058124344754,
-      "loss": 2.1381,
-      "step": 6665
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001982029651995546,
+      "loss": 2.1371,
+      "step": 6665
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.00015349156475457021,
-      "loss": 2.0915,
+      "epoch": 0.62,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019819532670840214,
+      "loss": 2.1943,
       "step": 6670
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001531774384113618,
-      "loss": 2.0751,
+      "epoch": 0.62,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019818767216536005,
+      "loss": 2.1893,
       "step": 6675
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0001528634342243801,
-      "loss": 2.0934,
+      "epoch": 0.62,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019818000157167967,
+      "loss": 2.1916,
       "step": 6680
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015254955301283635,
-      "loss": 2.127,
+      "epoch": 0.62,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019817231492861486,
+      "loss": 2.1868,
       "step": 6685
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001522357955956207,
-      "loss": 2.1216,
+      "epoch": 0.62,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019816461223742216,
+      "loss": 2.1412,
       "step": 6690
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00015192216279130076,
-      "loss": 2.112,
+      "epoch": 0.62,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019815689349936074,
+      "loss": 2.1834,
       "step": 6695
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015160865541811872,
-      "loss": 2.0921,
+      "epoch": 0.62,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019814915871569234,
+      "loss": 2.1998,
       "step": 6700
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015129527429398975,
-      "loss": 2.146,
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019814140788768135,
+      "loss": 2.2292,
       "step": 6705
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001509820202364993,
-      "loss": 2.095,
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019813364101659486,
+      "loss": 2.1457,
       "step": 6710
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001506688940629017,
-      "loss": 2.0786,
+      "epoch": 0.62,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019812585810370247,
+      "loss": 2.1835,
       "step": 6715
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00015035589659011746,
-      "loss": 2.0724,
+      "epoch": 0.62,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019811805915027645,
+      "loss": 2.1601,
       "step": 6720
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.0001500430286347314,
-      "loss": 2.1372,
+      "epoch": 0.62,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019811024415759162,
+      "loss": 2.1702,
       "step": 6725
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00014973029101299023,
-      "loss": 2.0978,
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019810241312692562,
+      "loss": 2.2098,
       "step": 6730
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0001494176845408008,
-      "loss": 2.1331,
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019809456605955847,
+      "loss": 2.1775,
       "step": 6735
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00014910521003372792,
-      "loss": 2.1272,
+      "epoch": 0.62,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019808670295677296,
+      "loss": 2.1848,
       "step": 6740
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00014879286830699195,
-      "loss": 2.1223,
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001980788238198545,
+      "loss": 2.1891,
       "step": 6745
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00014848066017546674,
-      "loss": 2.093,
+      "epoch": 0.63,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019807092865009102,
+      "loss": 2.131,
       "step": 6750
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2080078125,
-      "learning_rate": 0.00014816858645367778,
-      "loss": 2.1193,
+      "epoch": 0.63,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019806301744877317,
+      "loss": 2.216,
       "step": 6755
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00014785664795579983,
-      "loss": 2.081,
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019805509021719416,
+      "loss": 2.0969,
       "step": 6760
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.000147544845495655,
-      "loss": 2.1271,
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019804714695664992,
+      "loss": 2.186,
       "step": 6765
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.0001472331798867102,
-      "loss": 2.0703,
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019803918766843882,
+      "loss": 2.1518,
       "step": 6770
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001469216519420755,
-      "loss": 2.1123,
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019803121235386205,
+      "loss": 2.2037,
       "step": 6775
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001466102624745018,
-      "loss": 2.1405,
+      "epoch": 0.63,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019802322101422325,
+      "loss": 2.2104,
       "step": 6780
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00014629901229637878,
-      "loss": 2.1119,
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019801521365082883,
+      "loss": 2.1534,
       "step": 6785
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00014598790221973247,
-      "loss": 2.1046,
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019800719026498767,
+      "loss": 2.1533,
       "step": 6790
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00014567693305622377,
-      "loss": 2.1295,
+      "epoch": 0.63,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019799915085801144,
+      "loss": 2.1711,
       "step": 6795
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00014536610561714558,
-      "loss": 2.0998,
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019799109543121426,
+      "loss": 2.182,
       "step": 6800
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0001450554207134214,
-      "loss": 2.1193,
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019798302398591294,
+      "loss": 2.1303,
       "step": 6805
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00014474487915560252,
-      "loss": 2.101,
+      "epoch": 0.63,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001979749365234269,
+      "loss": 2.1279,
       "step": 6810
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001444344817538665,
-      "loss": 2.0741,
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019796683304507828,
+      "loss": 2.185,
       "step": 6815
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00014412422931801467,
-      "loss": 2.0999,
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019795871355219168,
+      "loss": 2.1867,
       "step": 6820
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00014381412265747034,
-      "loss": 2.1131,
+      "epoch": 0.63,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019795057804609435,
+      "loss": 2.1692,
       "step": 6825
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00014350416258127624,
-      "loss": 2.0859,
+      "epoch": 0.63,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001979424265281163,
+      "loss": 2.1676,
       "step": 6830
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00014319434989809287,
-      "loss": 2.0994,
+      "epoch": 0.63,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019793425899958993,
+      "loss": 2.1581,
       "step": 6835
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00014288468541619613,
-      "loss": 2.1219,
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019792607546185043,
+      "loss": 2.1327,
       "step": 6840
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00014257516994347532,
-      "loss": 2.1255,
+      "epoch": 0.63,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019791787591623558,
+      "loss": 2.2203,
       "step": 6845
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00014226580428743088,
-      "loss": 2.1276,
+      "epoch": 0.63,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019790966036408573,
+      "loss": 2.1103,
       "step": 6850
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00014195658925517246,
-      "loss": 2.0905,
+      "epoch": 0.64,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019790142880674388,
+      "loss": 2.2175,
       "step": 6855
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001416475256534168,
-      "loss": 2.1211,
+      "epoch": 0.64,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019789318124555558,
+      "loss": 2.1466,
       "step": 6860
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00014133861428848547,
-      "loss": 2.1013,
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001978849176818691,
+      "loss": 2.1602,
       "step": 6865
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00014102985596630297,
-      "loss": 2.0874,
+      "epoch": 0.64,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001978766381170353,
+      "loss": 2.1272,
       "step": 6870
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00014072125149239438,
-      "loss": 2.1041,
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019786834255240758,
+      "loss": 2.183,
       "step": 6875
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00014041280167188352,
-      "loss": 2.1036,
+      "epoch": 0.64,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019786003098934204,
+      "loss": 2.1508,
       "step": 6880
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.0001401045073094907,
-      "loss": 2.1149,
+      "epoch": 0.64,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019785170342919732,
+      "loss": 2.1375,
       "step": 6885
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.0001397963692095307,
-      "loss": 2.0599,
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019784335987333478,
+      "loss": 2.1678,
       "step": 6890
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.0001394883881759105,
-      "loss": 2.1005,
+      "epoch": 0.64,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001978350003231183,
+      "loss": 2.1935,
       "step": 6895
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.0001391805650121274,
-      "loss": 2.1037,
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019782662477991437,
+      "loss": 2.1489,
       "step": 6900
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00013887290052126683,
-      "loss": 2.0939,
+      "epoch": 0.64,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019781823324509218,
+      "loss": 2.1942,
       "step": 6905
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00013856539550600026,
-      "loss": 2.0898,
+      "epoch": 0.64,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019780982572002351,
+      "loss": 2.1909,
       "step": 6910
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00013825805076858299,
-      "loss": 2.0876,
+      "epoch": 0.64,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019780140220608265,
+      "loss": 2.1683,
       "step": 6915
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00013795086711085235,
-      "loss": 2.0711,
+      "epoch": 0.64,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019779296270464665,
+      "loss": 2.1705,
       "step": 6920
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00013764384533422526,
-      "loss": 2.1145,
+      "epoch": 0.64,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001977845072170951,
+      "loss": 2.1572,
       "step": 6925
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.0001373369862396965,
-      "loss": 2.1386,
+      "epoch": 0.64,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019777603574481017,
+      "loss": 2.1784,
       "step": 6930
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.00013703029062783615,
-      "loss": 2.1496,
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001977675482891767,
+      "loss": 2.1717,
       "step": 6935
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 0.00013672375929878801,
-      "loss": 2.1117,
+      "epoch": 0.64,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019775904485158215,
+      "loss": 2.203,
       "step": 6940
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00013641739305226723,
-      "loss": 2.1243,
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001977505254334165,
+      "loss": 2.1594,
       "step": 6945
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013611119268755825,
-      "loss": 2.0966,
+      "epoch": 0.64,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019774199003607255,
+      "loss": 2.1927,
       "step": 6950
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00013580515900351258,
-      "loss": 2.0824,
-      "step": 6955
+      "epoch": 0.64,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019773343866094542,
+      "loss": 2.1327,
+      "step": 6955
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00013549929279854718,
-      "loss": 2.1118,
+      "epoch": 0.64,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019772487130943308,
+      "loss": 2.1664,
       "step": 6960
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001351935948706419,
-      "loss": 2.1052,
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000197716287982936,
+      "loss": 2.1765,
       "step": 6965
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001348880660173376,
-      "loss": 2.1367,
+      "epoch": 0.65,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001977076886828573,
+      "loss": 2.1809,
       "step": 6970
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013458270703573387,
-      "loss": 2.0904,
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001976990734106027,
+      "loss": 2.253,
       "step": 6975
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00013427751872248744,
-      "loss": 2.0932,
+      "epoch": 0.65,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019769044216758048,
+      "loss": 2.1193,
       "step": 6980
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013397250187380957,
-      "loss": 2.1145,
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019768179495520166,
+      "loss": 2.1336,
       "step": 6985
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.0001336676572854643,
-      "loss": 2.1086,
+      "epoch": 0.65,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019767313177487973,
+      "loss": 2.227,
       "step": 6990
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.000133362985752766,
-      "loss": 2.1105,
+      "epoch": 0.65,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019766445262803088,
+      "loss": 2.1748,
       "step": 6995
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00013305848807057782,
-      "loss": 2.1029,
+      "epoch": 0.65,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019765575751607388,
+      "loss": 2.1907,
       "step": 7000
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.0001327541650333093,
-      "loss": 2.0997,
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001976470464404301,
+      "loss": 2.182,
       "step": 7005
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00013245001743491437,
-      "loss": 2.1089,
+      "epoch": 0.65,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019763831940252353,
+      "loss": 2.1538,
       "step": 7010
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013214604606888902,
-      "loss": 2.1435,
+      "epoch": 0.65,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001976295764037808,
+      "loss": 2.1218,
       "step": 7015
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00013184225172826973,
-      "loss": 2.1185,
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019762081744563106,
+      "loss": 2.1755,
       "step": 7020
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00013153863520563101,
-      "loss": 2.1492,
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019761204252950618,
+      "loss": 2.1679,
       "step": 7025
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013123519729308364,
-      "loss": 2.1143,
+      "epoch": 0.65,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001976032516568406,
+      "loss": 2.1598,
       "step": 7030
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001309319387822721,
-      "loss": 2.0802,
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019759444482907128,
+      "loss": 2.1308,
       "step": 7035
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00013062886046437302,
-      "loss": 2.1069,
+      "epoch": 0.65,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019758562204763794,
+      "loss": 2.21,
       "step": 7040
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00013032596313009286,
-      "loss": 2.1407,
+      "epoch": 0.65,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019757678331398278,
+      "loss": 2.2228,
       "step": 7045
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00013002324756966618,
-      "loss": 2.0927,
+      "epoch": 0.65,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019756792862955067,
+      "loss": 2.1182,
       "step": 7050
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001297207145728528,
-      "loss": 2.1295,
+      "epoch": 0.65,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001975590579957891,
+      "loss": 2.2277,
       "step": 7055
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.0001294183649289366,
-      "loss": 2.1743,
+      "epoch": 0.65,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001975501714141481,
+      "loss": 2.2003,
       "step": 7060
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.0001291161994267231,
-      "loss": 2.075,
+      "epoch": 0.65,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001975412688860804,
+      "loss": 2.1509,
       "step": 7065
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00012881421885453732,
-      "loss": 2.1188,
+      "epoch": 0.66,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019753235041304127,
+      "loss": 2.2001,
       "step": 7070
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00012851242400022177,
-      "loss": 2.1046,
+      "epoch": 0.66,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001975234159964886,
+      "loss": 2.1696,
       "step": 7075
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00012821081565113452,
-      "loss": 2.1022,
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019751446563788289,
+      "loss": 2.132,
       "step": 7080
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00012790939459414704,
-      "loss": 2.0762,
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019750549933868723,
+      "loss": 2.1759,
       "step": 7085
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00012760816161564223,
-      "loss": 2.1178,
+      "epoch": 0.66,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019749651710036735,
+      "loss": 2.2248,
       "step": 7090
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00012730711750151213,
-      "loss": 2.1288,
+      "epoch": 0.66,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019748751892439156,
+      "loss": 2.1701,
       "step": 7095
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00012700626303715625,
-      "loss": 2.0667,
+      "epoch": 0.66,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001974785048122308,
+      "loss": 2.2207,
       "step": 7100
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.0001267055990074792,
-      "loss": 2.0965,
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019746947476535857,
+      "loss": 2.1951,
       "step": 7105
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00012640512619688888,
-      "loss": 2.1014,
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000197460428785251,
+      "loss": 2.1696,
       "step": 7110
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00012610484538929412,
-      "loss": 2.1169,
+      "epoch": 0.66,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019745136687338689,
+      "loss": 2.1977,
       "step": 7115
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00012580475736810302,
-      "loss": 2.1372,
+      "epoch": 0.66,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019744228903124751,
+      "loss": 2.1562,
       "step": 7120
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.0001255048629162206,
-      "loss": 2.0957,
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019743319526031682,
+      "loss": 2.1671,
       "step": 7125
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00012520516281604692,
-      "loss": 2.1229,
+      "epoch": 0.66,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019742408556208137,
+      "loss": 2.2158,
       "step": 7130
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.000124905657849475,
-      "loss": 2.1027,
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019741495993803035,
+      "loss": 2.1793,
       "step": 7135
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00012460634879788874,
-      "loss": 2.0916,
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001974058183896555,
+      "loss": 2.1231,
       "step": 7140
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00012430723644216092,
-      "loss": 2.1181,
+      "epoch": 0.66,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019739666091845112,
+      "loss": 2.2025,
       "step": 7145
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00012400832156265126,
-      "loss": 2.1138,
+      "epoch": 0.66,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019738748752591424,
+      "loss": 2.1675,
       "step": 7150
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00012370960493920404,
-      "loss": 2.1201,
+      "epoch": 0.66,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019737829821354444,
+      "loss": 2.1378,
       "step": 7155
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.208984375,
-      "learning_rate": 0.00012341108735114653,
-      "loss": 2.0924,
+      "epoch": 0.66,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019736909298284384,
+      "loss": 2.151,
       "step": 7160
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00012311276957728665,
-      "loss": 2.1257,
+      "epoch": 0.66,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019735987183531722,
+      "loss": 2.1726,
       "step": 7165
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00012281465239591106,
-      "loss": 2.1157,
+      "epoch": 0.66,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019735063477247194,
+      "loss": 2.1385,
       "step": 7170
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00012251673658478305,
-      "loss": 2.1112,
+      "epoch": 0.66,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000197341381795818,
+      "loss": 2.1699,
       "step": 7175
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00012221902292114053,
-      "loss": 2.097,
+      "epoch": 0.67,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019733211290686795,
+      "loss": 2.2571,
       "step": 7180
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00012192151218169407,
-      "loss": 2.0966,
+      "epoch": 0.67,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000197322828107137,
+      "loss": 2.1364,
       "step": 7185
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.0001216242051426248,
-      "loss": 2.0823,
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019731352739814288,
+      "loss": 2.2002,
       "step": 7190
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00012132710257958248,
-      "loss": 2.1166,
+      "epoch": 0.67,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000197304210781406,
+      "loss": 2.1607,
       "step": 7195
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00012103020526768327,
-      "loss": 2.1522,
+      "epoch": 0.67,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019729487825844935,
+      "loss": 2.1846,
       "step": 7200
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00012073351398150796,
-      "loss": 2.1473,
+      "epoch": 0.67,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019728552983079846,
+      "loss": 2.1898,
       "step": 7205
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00012043702949509983,
-      "loss": 2.1095,
+      "epoch": 0.67,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019727616549998156,
+      "loss": 2.1855,
       "step": 7210
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00012014075258196266,
-      "loss": 2.107,
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001972667852675294,
+      "loss": 2.128,
       "step": 7215
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00011984468401505849,
-      "loss": 2.1213,
+      "epoch": 0.67,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019725738913497538,
+      "loss": 2.179,
       "step": 7220
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00011954882456680603,
-      "loss": 2.1018,
+      "epoch": 0.67,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019724797710385543,
+      "loss": 2.1329,
       "step": 7225
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00011925317500907837,
-      "loss": 2.1361,
+      "epoch": 0.67,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019723854917570813,
+      "loss": 2.1418,
       "step": 7230
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00011895773611320104,
-      "loss": 2.1327,
+      "epoch": 0.67,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019722910535207474,
+      "loss": 2.208,
       "step": 7235
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00011866250864994973,
-      "loss": 2.0931,
+      "epoch": 0.67,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019721964563449893,
+      "loss": 2.1486,
       "step": 7240
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00011836749338954885,
-      "loss": 2.1342,
+      "epoch": 0.67,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001972101700245271,
+      "loss": 2.174,
       "step": 7245
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00011807269110166897,
-      "loss": 2.0836,
+      "epoch": 0.67,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019720067852370824,
+      "loss": 2.186,
       "step": 7250
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00011777810255542521,
-      "loss": 2.0761,
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001971911711335939,
+      "loss": 2.1615,
       "step": 7255
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00011748372851937479,
-      "loss": 2.1275,
+      "epoch": 0.67,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019718164785573823,
+      "loss": 2.2389,
       "step": 7260
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00011718956976151552,
-      "loss": 2.1099,
+      "epoch": 0.67,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019717210869169805,
+      "loss": 2.1738,
       "step": 7265
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0001168956270492835,
-      "loss": 2.0851,
+      "epoch": 0.67,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019716255364303264,
+      "loss": 2.107,
       "step": 7270
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00011660190114955129,
-      "loss": 2.0925,
+      "epoch": 0.67,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000197152982711304,
+      "loss": 2.1923,
       "step": 7275
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00011630839282862546,
-      "loss": 2.0861,
+      "epoch": 0.67,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019714339589807666,
+      "loss": 2.1311,
       "step": 7280
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2265625,
-      "learning_rate": 0.00011601510285224539,
-      "loss": 2.1035,
+      "epoch": 0.68,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019713379320491776,
+      "loss": 2.2318,
       "step": 7285
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00011572203198558043,
-      "loss": 2.0903,
+      "epoch": 0.68,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019712417463339706,
+      "loss": 2.2133,
       "step": 7290
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00011542918099322862,
-      "loss": 2.0787,
+      "epoch": 0.68,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019711454018508695,
+      "loss": 2.1259,
       "step": 7295
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00011513655063921412,
-      "loss": 2.1227,
+      "epoch": 0.68,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019710488986156226,
+      "loss": 2.1426,
       "step": 7300
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00011484414168698547,
-      "loss": 2.0865,
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019709522366440058,
+      "loss": 2.1305,
       "step": 7305
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.0001145519548994138,
-      "loss": 2.072,
+      "epoch": 0.68,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019708554159518203,
+      "loss": 2.1581,
       "step": 7310
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00011425999103879048,
-      "loss": 2.1382,
+      "epoch": 0.68,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001970758436554893,
+      "loss": 2.1243,
       "step": 7315
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00011396825086682516,
-      "loss": 2.094,
+      "epoch": 0.68,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019706612984690772,
+      "loss": 2.1674,
       "step": 7320
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001136767351446442,
-      "loss": 2.1178,
+      "epoch": 0.68,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001970564001710252,
+      "loss": 2.1554,
       "step": 7325
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00011338544463278815,
-      "loss": 2.092,
+      "epoch": 0.68,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019704665462943227,
+      "loss": 2.1803,
       "step": 7330
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00011309438009121025,
-      "loss": 2.0974,
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019703689322372197,
+      "loss": 2.2612,
       "step": 7335
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00011280354227927374,
-      "loss": 2.1064,
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019702711595548998,
+      "loss": 2.1468,
       "step": 7340
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00011251293195575084,
-      "loss": 2.0531,
+      "epoch": 0.68,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019701732282633466,
+      "loss": 2.216,
       "step": 7345
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0001122225498788201,
-      "loss": 2.1161,
+      "epoch": 0.68,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019700751383785678,
+      "loss": 2.1247,
       "step": 7350
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00011193239680606451,
-      "loss": 2.1124,
+      "epoch": 0.68,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001969976889916599,
+      "loss": 2.1458,
       "step": 7355
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00011164247349446957,
-      "loss": 2.0824,
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019698784828935,
+      "loss": 2.1693,
       "step": 7360
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00011135278070042163,
-      "loss": 2.0939,
+      "epoch": 0.68,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019697799173253578,
+      "loss": 2.2257,
       "step": 7365
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00011106331917970527,
-      "loss": 2.0857,
+      "epoch": 0.68,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019696811932282845,
+      "loss": 2.156,
       "step": 7370
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00011077408968750214,
-      "loss": 2.1444,
+      "epoch": 0.68,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019695823106184187,
+      "loss": 2.2244,
       "step": 7375
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00011048509297838791,
-      "loss": 2.1137,
+      "epoch": 0.68,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019694832695119248,
+      "loss": 2.2523,
       "step": 7380
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00011019632980633156,
-      "loss": 2.0799,
+      "epoch": 0.68,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001969384069924992,
+      "loss": 2.157,
       "step": 7385
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00010990780092469259,
-      "loss": 2.0915,
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019692847118738376,
+      "loss": 2.1788,
       "step": 7390
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 0.0001096195070862191,
-      "loss": 2.1333,
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019691851953747028,
+      "loss": 2.1996,
       "step": 7395
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2109375,
-      "learning_rate": 0.0001093314490430461,
-      "loss": 2.1179,
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019690855204438555,
+      "loss": 2.1707,
       "step": 7400
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00010904362754669342,
-      "loss": 2.0775,
+      "epoch": 0.69,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019689856870975898,
+      "loss": 2.1413,
       "step": 7405
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001087560433480637,
-      "loss": 2.1088,
+      "epoch": 0.69,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001968885695352225,
+      "loss": 2.1704,
       "step": 7410
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00010846869719744086,
-      "loss": 2.1282,
+      "epoch": 0.69,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019687855452241068,
+      "loss": 2.2277,
       "step": 7415
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00010818158984448703,
-      "loss": 2.0788,
+      "epoch": 0.69,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019686852367296068,
+      "loss": 2.1473,
       "step": 7420
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00010789472203824207,
-      "loss": 2.145,
+      "epoch": 0.69,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001968584769885122,
+      "loss": 2.1569,
       "step": 7425
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00010760809452712041,
-      "loss": 2.1578,
+      "epoch": 0.69,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019684841447070757,
+      "loss": 2.1778,
       "step": 7430
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00010732170805890987,
-      "loss": 2.0883,
+      "epoch": 0.69,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019683833612119174,
+      "loss": 2.1925,
       "step": 7435
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00010703556338076924,
-      "loss": 2.1002,
+      "epoch": 0.69,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019682824194161214,
+      "loss": 2.1235,
       "step": 7440
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00010674966123922645,
-      "loss": 2.1353,
+      "epoch": 0.69,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019681813193361892,
+      "loss": 2.1776,
       "step": 7445
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00010646400238017691,
-      "loss": 2.1553,
+      "epoch": 0.69,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001968080060988647,
+      "loss": 2.1747,
       "step": 7450
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 0.00010617858754888103,
-      "loss": 2.1264,
+      "epoch": 0.69,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00019679786443900477,
+      "loss": 2.1767,
       "step": 7455
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00010589341748996285,
-      "loss": 2.0861,
+      "epoch": 0.69,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019678770695569697,
+      "loss": 2.1821,
       "step": 7460
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00010560849294740765,
-      "loss": 2.1787,
+      "epoch": 0.69,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019677753365060174,
+      "loss": 2.1577,
       "step": 7465
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00010532381466456009,
-      "loss": 2.0844,
+      "epoch": 0.69,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019676734452538205,
+      "loss": 2.1694,
       "step": 7470
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00010503938338412269,
-      "loss": 2.1122,
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001967571395817036,
+      "loss": 2.1764,
       "step": 7475
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00010475519984815325,
-      "loss": 2.1068,
+      "epoch": 0.69,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019674691882123454,
+      "loss": 2.1407,
       "step": 7480
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00010447126479806326,
-      "loss": 2.124,
+      "epoch": 0.69,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001967366822456456,
+      "loss": 2.1427,
       "step": 7485
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 0.00010418757897461614,
-      "loss": 2.1239,
+      "epoch": 0.69,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019672642985661017,
+      "loss": 2.1346,
       "step": 7490
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00010390414311792485,
-      "loss": 2.1004,
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019671616165580424,
+      "loss": 2.1172,
       "step": 7495
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 0.00010362095796745039,
-      "loss": 2.0709,
+      "epoch": 0.69,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001967058776449063,
+      "loss": 2.1959,
       "step": 7500
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00010333802426199963,
-      "loss": 2.1201,
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019669557782559744,
+      "loss": 2.216,
       "step": 7505
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00010305534273972326,
-      "loss": 2.1216,
+      "epoch": 0.7,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019668526219956145,
+      "loss": 2.1694,
       "step": 7510
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00010277291413811445,
-      "loss": 2.1217,
+      "epoch": 0.7,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019667493076848454,
+      "loss": 2.1743,
       "step": 7515
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00010249073919400611,
-      "loss": 2.0801,
+      "epoch": 0.7,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019666458353405562,
+      "loss": 2.1446,
       "step": 7520
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.21484375,
-      "learning_rate": 0.00010220881864356955,
-      "loss": 2.0881,
+      "epoch": 0.7,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001966542204979661,
+      "loss": 2.1382,
       "step": 7525
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00010192715322231253,
-      "loss": 2.1399,
+      "epoch": 0.7,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019664384166191004,
+      "loss": 2.1258,
       "step": 7530
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00010164574366507685,
-      "loss": 2.1091,
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019663344702758404,
+      "loss": 2.2213,
       "step": 7535
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.00010136459070603719,
-      "loss": 2.1038,
+      "epoch": 0.7,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019662303659668731,
+      "loss": 2.155,
       "step": 7540
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00010108369507869848,
-      "loss": 2.0849,
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001966126103709217,
+      "loss": 2.1502,
       "step": 7545
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00010080305751589434,
-      "loss": 2.1102,
+      "epoch": 0.7,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019660216835199146,
+      "loss": 2.2091,
       "step": 7550
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00010052267874978531,
-      "loss": 2.1513,
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001965917105416036,
+      "loss": 2.1821,
       "step": 7555
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.216796875,
-      "learning_rate": 0.00010024255951185657,
-      "loss": 2.0989,
+      "epoch": 0.7,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019658123694146765,
+      "loss": 2.183,
       "step": 7560
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 9.996270053291614e-05,
-      "loss": 2.1153,
+      "epoch": 0.7,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019657074755329568,
+      "loss": 2.2143,
       "step": 7565
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.216796875,
-      "learning_rate": 9.968310254309335e-05,
-      "loss": 2.1246,
+      "epoch": 0.7,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001965602423788024,
+      "loss": 2.2226,
       "step": 7570
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.22265625,
-      "learning_rate": 9.940376627183628e-05,
-      "loss": 2.1371,
+      "epoch": 0.7,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001965497214197051,
+      "loss": 2.1954,
       "step": 7575
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.912469244791054e-05,
-      "loss": 2.1117,
+      "epoch": 0.7,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019653918467772363,
+      "loss": 2.1578,
       "step": 7580
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.216796875,
-      "learning_rate": 9.884588179939677e-05,
-      "loss": 2.1116,
+      "epoch": 0.7,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019652863215458038,
+      "loss": 2.1396,
       "step": 7585
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.856733505368905e-05,
-      "loss": 2.0977,
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019651806385200037,
+      "loss": 2.1581,
       "step": 7590
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.828905293749318e-05,
-      "loss": 2.0909,
+      "epoch": 0.7,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019650747977171124,
+      "loss": 2.1478,
       "step": 7595
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 9.80110361768243e-05,
-      "loss": 2.1109,
+      "epoch": 0.7,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019649687991544313,
+      "loss": 2.1832,
       "step": 7600
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.224609375,
-      "learning_rate": 9.773328549700526e-05,
-      "loss": 2.0799,
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019648626428492877,
+      "loss": 2.1733,
       "step": 7605
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 9.745580162266505e-05,
-      "loss": 2.1182,
+      "epoch": 0.71,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019647563288190348,
+      "loss": 2.1844,
       "step": 7610
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 9.717858527773616e-05,
-      "loss": 2.1301,
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001964649857081052,
+      "loss": 2.1123,
       "step": 7615
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.216796875,
-      "learning_rate": 9.690163718545356e-05,
-      "loss": 2.1144,
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019645432276527437,
+      "loss": 2.1578,
       "step": 7620
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 9.662495806835179e-05,
-      "loss": 2.0999,
+      "epoch": 0.71,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019644364405515413,
+      "loss": 2.1505,
       "step": 7625
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.634854864826421e-05,
-      "loss": 2.1331,
+      "epoch": 0.71,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019643294957949,
+      "loss": 2.1787,
       "step": 7630
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 9.607240964632039e-05,
-      "loss": 2.1587,
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001964222393400303,
+      "loss": 2.1284,
       "step": 7635
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.228515625,
-      "learning_rate": 9.579654178294429e-05,
-      "loss": 2.1276,
+      "epoch": 0.71,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019641151333852575,
+      "loss": 2.1413,
       "step": 7640
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.552094577785255e-05,
-      "loss": 2.1238,
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019640077157672976,
+      "loss": 2.109,
       "step": 7645
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 9.524562235005252e-05,
-      "loss": 2.1393,
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019639001405639828,
+      "loss": 2.1301,
       "step": 7650
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.220703125,
-      "learning_rate": 9.497057221784051e-05,
-      "loss": 2.0863,
+      "epoch": 0.71,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001963792407792898,
+      "loss": 2.2609,
       "step": 7655
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 9.469579609879993e-05,
-      "loss": 2.1032,
+      "epoch": 0.71,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019636845174716543,
+      "loss": 2.1725,
       "step": 7660
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.442129470979888e-05,
-      "loss": 2.09,
+      "epoch": 0.71,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019635764696178883,
+      "loss": 2.1706,
       "step": 7665
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.21484375,
-      "learning_rate": 9.414706876698922e-05,
-      "loss": 2.0932,
+      "epoch": 0.71,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001963468264249263,
+      "loss": 2.212,
       "step": 7670
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.212890625,
-      "learning_rate": 9.387311898580378e-05,
-      "loss": 2.098,
+      "epoch": 0.71,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001963359901383466,
+      "loss": 2.1558,
       "step": 7675
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.359944608095528e-05,
-      "loss": 2.0938,
+      "epoch": 0.71,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019632513810382112,
+      "loss": 2.1838,
       "step": 7680
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.234375,
-      "learning_rate": 9.332605076643379e-05,
-      "loss": 2.0881,
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019631427032312386,
+      "loss": 2.1724,
       "step": 7685
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.305293375550525e-05,
-      "loss": 2.1616,
+      "epoch": 0.71,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019630338679803142,
+      "loss": 2.2162,
       "step": 7690
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 9.278009576070963e-05,
-      "loss": 2.0904,
+      "epoch": 0.71,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019629248753032282,
+      "loss": 2.1736,
       "step": 7695
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.250753749385903e-05,
-      "loss": 2.0911,
+      "epoch": 0.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001962815725217798,
+      "loss": 2.1388,
       "step": 7700
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.223525966603536e-05,
-      "loss": 2.1187,
+      "epoch": 0.71,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019627064177418663,
+      "loss": 2.1522,
       "step": 7705
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.220703125,
-      "learning_rate": 9.196326298758937e-05,
-      "loss": 2.1144,
+      "epoch": 0.71,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019625969528933013,
+      "loss": 2.186,
       "step": 7710
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.220703125,
-      "learning_rate": 9.169154816813798e-05,
-      "loss": 2.0919,
+      "epoch": 0.71,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001962487330689997,
+      "loss": 2.1858,
       "step": 7715
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.220703125,
-      "learning_rate": 9.142011591656301e-05,
-      "loss": 2.0698,
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001962377551149874,
+      "loss": 2.1425,
       "step": 7720
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.212890625,
-      "learning_rate": 9.114896694100891e-05,
-      "loss": 2.0918,
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001962267614290877,
+      "loss": 2.1514,
       "step": 7725
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.22265625,
-      "learning_rate": 9.087810194888107e-05,
-      "loss": 2.1143,
+      "epoch": 0.72,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019621575201309775,
+      "loss": 2.1388,
       "step": 7730
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 9.06075216468442e-05,
-      "loss": 2.1348,
+      "epoch": 0.72,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019620472686881723,
+      "loss": 2.1682,
       "step": 7735
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.21484375,
-      "learning_rate": 9.033722674082006e-05,
-      "loss": 2.119,
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019619368599804846,
+      "loss": 2.1538,
       "step": 7740
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 9.006721793598584e-05,
-      "loss": 2.1144,
+      "epoch": 0.72,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019618262940259624,
+      "loss": 2.2009,
       "step": 7745
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.979749593677253e-05,
-      "loss": 2.0731,
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000196171557084268,
+      "loss": 2.1334,
       "step": 7750
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2265625,
-      "learning_rate": 8.952806144686261e-05,
-      "loss": 2.1358,
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019616046904487368,
+      "loss": 2.1664,
       "step": 7755
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.21875,
-      "learning_rate": 8.92589151691887e-05,
-      "loss": 2.1075,
+      "epoch": 0.72,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001961493652862259,
+      "loss": 2.1992,
       "step": 7760
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.224609375,
-      "learning_rate": 8.89900578059313e-05,
-      "loss": 2.0871,
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019613824581013974,
+      "loss": 2.1496,
       "step": 7765
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 8.872149005851722e-05,
-      "loss": 2.107,
+      "epoch": 0.72,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019612711061843289,
+      "loss": 2.1719,
       "step": 7770
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.21875,
-      "learning_rate": 8.845321262761781e-05,
-      "loss": 2.11,
+      "epoch": 0.72,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019611595971292557,
+      "loss": 2.206,
       "step": 7775
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 8.818522621314678e-05,
-      "loss": 2.1028,
+      "epoch": 0.72,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001961047930954407,
+      "loss": 2.1675,
       "step": 7780
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.791753151425888e-05,
-      "loss": 2.1378,
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019609361076780362,
+      "loss": 2.1219,
       "step": 7785
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.765012922934755e-05,
-      "loss": 2.1208,
+      "epoch": 0.72,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019608241273184226,
+      "loss": 2.1514,
       "step": 7790
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 8.738302005604335e-05,
-      "loss": 2.1132,
+      "epoch": 0.72,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019607119898938722,
+      "loss": 2.1567,
       "step": 7795
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.21875,
-      "learning_rate": 8.711620469121245e-05,
-      "loss": 2.1078,
-      "step": 7800
+      "epoch": 0.72,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019605996954227153,
+      "loss": 2.1716,
+      "step": 7800
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.22265625,
-      "learning_rate": 8.684968383095411e-05,
-      "loss": 2.0819,
+      "epoch": 0.72,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019604872439233094,
+      "loss": 2.197,
       "step": 7805
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.228515625,
-      "learning_rate": 8.658345817059941e-05,
-      "loss": 2.1066,
+      "epoch": 0.72,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001960374635414036,
+      "loss": 2.173,
       "step": 7810
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 8.631752840470941e-05,
-      "loss": 2.1254,
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019602618699133035,
+      "loss": 2.1976,
       "step": 7815
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.605189522707291e-05,
-      "loss": 2.0947,
+      "epoch": 0.72,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019601489474395457,
+      "loss": 2.1625,
       "step": 7820
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 8.57865593307053e-05,
-      "loss": 2.1196,
+      "epoch": 0.73,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019600358680112217,
+      "loss": 2.1062,
       "step": 7825
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.552152140784607e-05,
-      "loss": 2.1187,
+      "epoch": 0.73,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019599226316468165,
+      "loss": 2.2011,
       "step": 7830
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.21484375,
-      "learning_rate": 8.525678214995738e-05,
-      "loss": 2.0997,
+      "epoch": 0.73,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019598092383648404,
+      "loss": 2.1465,
       "step": 7835
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.499234224772246e-05,
-      "loss": 2.0727,
+      "epoch": 0.73,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019596956881838304,
+      "loss": 2.1614,
       "step": 7840
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 8.47282023910432e-05,
-      "loss": 2.1121,
+      "epoch": 0.73,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001959581981122348,
+      "loss": 2.1652,
       "step": 7845
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.446436326903882e-05,
-      "loss": 2.133,
+      "epoch": 0.73,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001959468117198981,
+      "loss": 2.2084,
       "step": 7850
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 8.420082557004409e-05,
-      "loss": 2.0483,
+      "epoch": 0.73,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019593540964323424,
+      "loss": 2.1897,
       "step": 7855
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.224609375,
-      "learning_rate": 8.393758998160717e-05,
-      "loss": 2.1356,
+      "epoch": 0.73,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001959239918841071,
+      "loss": 2.2209,
       "step": 7860
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.367465719048831e-05,
-      "loss": 2.1049,
+      "epoch": 0.73,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019591255844438315,
+      "loss": 2.1493,
       "step": 7865
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.212890625,
-      "learning_rate": 8.341202788265734e-05,
-      "loss": 2.0773,
+      "epoch": 0.73,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019590110932593137,
+      "loss": 2.165,
       "step": 7870
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 8.314970274329277e-05,
-      "loss": 2.117,
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001958896445306234,
+      "loss": 2.1902,
       "step": 7875
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.224609375,
-      "learning_rate": 8.288768245677945e-05,
-      "loss": 2.131,
+      "epoch": 0.73,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019587816406033332,
+      "loss": 2.1745,
       "step": 7880
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 8.262596770670676e-05,
-      "loss": 2.0954,
+      "epoch": 0.73,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019586666791693785,
+      "loss": 2.1646,
       "step": 7885
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 8.236455917586699e-05,
-      "loss": 2.0664,
+      "epoch": 0.73,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019585515610231623,
+      "loss": 2.1523,
       "step": 7890
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.21034575462537e-05,
-      "loss": 2.1112,
+      "epoch": 0.73,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019584362861835034,
+      "loss": 2.1992,
       "step": 7895
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.184266349905951e-05,
-      "loss": 2.1066,
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019583208546692453,
+      "loss": 2.1439,
       "step": 7900
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 8.158217771467496e-05,
-      "loss": 2.1083,
+      "epoch": 0.73,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019582052664992574,
+      "loss": 2.152,
       "step": 7905
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.132200087268577e-05,
-      "loss": 2.1021,
+      "epoch": 0.73,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001958089521692435,
+      "loss": 2.1833,
       "step": 7910
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 8.106213365187223e-05,
-      "loss": 2.1463,
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001957973620267698,
+      "loss": 2.1738,
       "step": 7915
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.080257673020663e-05,
-      "loss": 2.1233,
+      "epoch": 0.73,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019578575622439942,
+      "loss": 2.2025,
       "step": 7920
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.21875,
-      "learning_rate": 8.054333078485168e-05,
-      "loss": 2.1229,
+      "epoch": 0.73,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019577413476402943,
+      "loss": 2.1699,
       "step": 7925
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.028439649215877e-05,
-      "loss": 2.104,
+      "epoch": 0.73,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001957624976475596,
+      "loss": 2.1708,
       "step": 7930
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 8.002577452766623e-05,
-      "loss": 2.0927,
+      "epoch": 0.74,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019575084487689228,
+      "loss": 2.216,
       "step": 7935
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 7.976746556609759e-05,
-      "loss": 2.0963,
+      "epoch": 0.74,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019573917645393228,
+      "loss": 2.2305,
       "step": 7940
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 7.950947028135997e-05,
-      "loss": 2.1209,
+      "epoch": 0.74,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019572749238058705,
+      "loss": 2.1134,
       "step": 7945
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.21484375,
-      "learning_rate": 7.925178934654161e-05,
-      "loss": 2.0654,
+      "epoch": 0.74,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019571579265876662,
+      "loss": 2.1577,
       "step": 7950
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 7.899442343391114e-05,
-      "loss": 2.0789,
+      "epoch": 0.74,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019570407729038345,
+      "loss": 2.1803,
       "step": 7955
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.873737321491506e-05,
-      "loss": 2.1175,
+      "epoch": 0.74,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019569234627735265,
+      "loss": 2.2069,
       "step": 7960
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.216796875,
-      "learning_rate": 7.848063936017642e-05,
-      "loss": 2.086,
+      "epoch": 0.74,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019568059962159193,
+      "loss": 2.1483,
       "step": 7965
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 7.822422253949278e-05,
-      "loss": 2.1389,
+      "epoch": 0.74,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001956688373250215,
+      "loss": 2.1355,
       "step": 7970
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 7.79681234218345e-05,
-      "loss": 2.0908,
+      "epoch": 0.74,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019565705938956408,
+      "loss": 2.2307,
       "step": 7975
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 7.771234267534337e-05,
-      "loss": 2.0564,
+      "epoch": 0.74,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000195645265817145,
+      "loss": 2.1419,
       "step": 7980
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.745688096733032e-05,
-      "loss": 2.1368,
+      "epoch": 0.74,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019563345660969218,
+      "loss": 2.1812,
       "step": 7985
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.720173896427394e-05,
-      "loss": 2.0946,
+      "epoch": 0.74,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019562163176913607,
+      "loss": 2.1467,
       "step": 7990
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.694691733181896e-05,
-      "loss": 2.1287,
+      "epoch": 0.74,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001956097912974096,
+      "loss": 2.172,
       "step": 7995
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 7.669241673477399e-05,
-      "loss": 2.1275,
+      "epoch": 0.74,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019559793519644843,
+      "loss": 2.1615,
       "step": 8000
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 7.643823783711037e-05,
-      "loss": 2.0847,
+      "epoch": 0.74,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019558606346819054,
+      "loss": 2.1419,
       "step": 8005
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 7.618438130195998e-05,
-      "loss": 2.065,
+      "epoch": 0.74,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019557417611457668,
+      "loss": 2.121,
       "step": 8010
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.593084779161359e-05,
-      "loss": 2.1041,
+      "epoch": 0.74,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019556227313754997,
+      "loss": 2.1829,
       "step": 8015
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.216796875,
-      "learning_rate": 7.567763796751954e-05,
-      "loss": 2.1025,
+      "epoch": 0.74,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019555035453905627,
+      "loss": 2.1596,
       "step": 8020
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.2109375,
-      "learning_rate": 7.542475249028145e-05,
-      "loss": 2.069,
+      "epoch": 0.74,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019553842032104388,
+      "loss": 2.1999,
       "step": 8025
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 7.51721920196567e-05,
-      "loss": 2.074,
+      "epoch": 0.74,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019552647048546364,
+      "loss": 2.1749,
       "step": 8030
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.491995721455505e-05,
-      "loss": 2.1351,
+      "epoch": 0.74,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000195514505034269,
+      "loss": 2.1908,
       "step": 8035
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.2421875,
-      "learning_rate": 7.466804873303625e-05,
-      "loss": 2.1296,
+      "epoch": 0.74,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019550252396941596,
+      "loss": 2.1597,
       "step": 8040
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.212890625,
-      "learning_rate": 7.441646723230906e-05,
-      "loss": 2.1207,
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000195490527292863,
+      "loss": 2.1797,
       "step": 8045
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.216796875,
-      "learning_rate": 7.416521336872894e-05,
-      "loss": 2.0797,
+      "epoch": 0.75,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019547851500657126,
+      "loss": 2.1792,
       "step": 8050
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.391428779779655e-05,
-      "loss": 2.1238,
+      "epoch": 0.75,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019546648711250432,
+      "loss": 2.197,
       "step": 8055
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.228515625,
-      "learning_rate": 7.366369117415626e-05,
-      "loss": 2.0916,
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019545444361262845,
+      "loss": 2.1861,
       "step": 8060
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 7.341342415159403e-05,
-      "loss": 2.0776,
+      "epoch": 0.75,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019544238450891228,
+      "loss": 2.1665,
       "step": 8065
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 7.316348738303611e-05,
-      "loss": 2.14,
+      "epoch": 0.75,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019543030980332716,
+      "loss": 2.1424,
       "step": 8070
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.224609375,
-      "learning_rate": 7.291388152054704e-05,
-      "loss": 2.0778,
+      "epoch": 0.75,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019541821949784697,
+      "loss": 2.1612,
       "step": 8075
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.266460721532793e-05,
-      "loss": 2.0561,
+      "epoch": 0.75,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019540611359444805,
+      "loss": 2.1856,
       "step": 8080
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 7.241566511771521e-05,
-      "loss": 2.1152,
+      "epoch": 0.75,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001953939920951093,
+      "loss": 2.1991,
       "step": 8085
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 7.216705587717831e-05,
-      "loss": 2.12,
+      "epoch": 0.75,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001953818550018123,
+      "loss": 2.1939,
       "step": 8090
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.21484375,
-      "learning_rate": 7.191878014231833e-05,
-      "loss": 2.0983,
+      "epoch": 0.75,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.000195369702316541,
+      "loss": 2.2318,
       "step": 8095
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.16708385608664e-05,
-      "loss": 2.0884,
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001953575340412821,
+      "loss": 2.1298,
       "step": 8100
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.216796875,
-      "learning_rate": 7.142323177968173e-05,
-      "loss": 2.1197,
+      "epoch": 0.75,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001953453501780246,
+      "loss": 2.2079,
       "step": 8105
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 7.117596044475022e-05,
-      "loss": 2.1033,
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001953331507287603,
+      "loss": 2.1956,
       "step": 8110
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.2265625,
-      "learning_rate": 7.092902520118243e-05,
-      "loss": 2.1059,
+      "epoch": 0.75,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019532093569548336,
+      "loss": 2.1565,
       "step": 8115
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.068242669321215e-05,
-      "loss": 2.1148,
+      "epoch": 0.75,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019530870508019054,
+      "loss": 2.1829,
       "step": 8120
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 7.043616556419479e-05,
-      "loss": 2.1426,
+      "epoch": 0.75,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019529645888488126,
+      "loss": 2.2037,
       "step": 8125
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.21484375,
-      "learning_rate": 7.019024245660542e-05,
-      "loss": 2.0757,
+      "epoch": 0.75,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019528419711155732,
+      "loss": 2.1791,
       "step": 8130
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 6.99446580120372e-05,
-      "loss": 2.1089,
+      "epoch": 0.75,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019527191976222318,
+      "loss": 2.1666,
       "step": 8135
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.969941287119994e-05,
-      "loss": 2.0967,
+      "epoch": 0.75,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001952596268388857,
+      "loss": 2.2082,
       "step": 8140
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.945450767391799e-05,
-      "loss": 2.0861,
+      "epoch": 0.75,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001952473183435546,
+      "loss": 2.1794,
       "step": 8145
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.920994305912915e-05,
-      "loss": 2.1158,
+      "epoch": 0.76,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001952349942782417,
+      "loss": 2.1407,
       "step": 8150
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.21484375,
-      "learning_rate": 6.896571966488219e-05,
-      "loss": 2.0883,
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019522265464496182,
+      "loss": 2.097,
       "step": 8155
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.872183812833608e-05,
-      "loss": 2.1185,
+      "epoch": 0.76,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019521029944573193,
+      "loss": 2.1514,
       "step": 8160
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.847829908575782e-05,
-      "loss": 2.1195,
+      "epoch": 0.76,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001951979286825718,
+      "loss": 2.2008,
       "step": 8165
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.823510317252075e-05,
-      "loss": 2.1461,
+      "epoch": 0.76,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019518554235750367,
+      "loss": 2.1551,
       "step": 8170
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.79922510231031e-05,
-      "loss": 2.0658,
+      "epoch": 0.76,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019517314047255233,
+      "loss": 2.2398,
       "step": 8175
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.774974327108614e-05,
-      "loss": 2.0876,
+      "epoch": 0.76,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019516072302974507,
+      "loss": 2.1861,
       "step": 8180
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.750758054915278e-05,
-      "loss": 2.0941,
+      "epoch": 0.76,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019514829003111178,
+      "loss": 2.1464,
       "step": 8185
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.22265625,
-      "learning_rate": 6.726576348908589e-05,
-      "loss": 2.1018,
+      "epoch": 0.76,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019513584147868484,
+      "loss": 2.2192,
       "step": 8190
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.702429272176607e-05,
-      "loss": 2.117,
+      "epoch": 0.76,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019512337737449923,
+      "loss": 2.1546,
       "step": 8195
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.6783168877171e-05,
-      "loss": 2.1286,
+      "epoch": 0.76,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019511089772059244,
+      "loss": 2.1722,
       "step": 8200
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.654239258437285e-05,
-      "loss": 2.0733,
+      "epoch": 0.76,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019509840251900448,
+      "loss": 2.1811,
       "step": 8205
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 6.63019644715374e-05,
-      "loss": 2.0924,
+      "epoch": 0.76,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019508589177177798,
+      "loss": 2.1972,
       "step": 8210
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.606188516592185e-05,
-      "loss": 2.1021,
+      "epoch": 0.76,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019507336548095806,
+      "loss": 2.1876,
       "step": 8215
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.582215529387335e-05,
-      "loss": 2.1414,
+      "epoch": 0.76,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019506082364859227,
+      "loss": 2.1693,
       "step": 8220
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.236328125,
-      "learning_rate": 6.558277548082754e-05,
-      "loss": 2.1095,
+      "epoch": 0.76,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019504826627673096,
+      "loss": 2.2061,
       "step": 8225
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.22265625,
-      "learning_rate": 6.534374635130692e-05,
-      "loss": 2.0879,
+      "epoch": 0.76,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019503569336742677,
+      "loss": 2.1745,
       "step": 8230
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.510506852891858e-05,
-      "loss": 2.086,
+      "epoch": 0.76,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019502310492273498,
+      "loss": 2.16,
       "step": 8235
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 6.486674263635364e-05,
-      "loss": 2.0797,
+      "epoch": 0.76,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001950105009447135,
+      "loss": 2.1762,
       "step": 8240
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 6.462876929538468e-05,
-      "loss": 2.1481,
+      "epoch": 0.76,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019499788143542262,
+      "loss": 2.1189,
       "step": 8245
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.220703125,
-      "learning_rate": 6.439114912686479e-05,
-      "loss": 2.1469,
+      "epoch": 0.76,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019498524639692522,
+      "loss": 2.1611,
       "step": 8250
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.415388275072544e-05,
-      "loss": 2.1036,
+      "epoch": 0.76,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019497259583128678,
+      "loss": 2.1824,
       "step": 8255
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.391697078597512e-05,
-      "loss": 2.0785,
+      "epoch": 0.77,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019495992974057527,
+      "loss": 2.2104,
       "step": 8260
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.21484375,
-      "learning_rate": 6.368041385069785e-05,
-      "loss": 2.0955,
+      "epoch": 0.77,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019494724812686118,
+      "loss": 2.1471,
       "step": 8265
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 6.344421256205122e-05,
-      "loss": 2.1202,
+      "epoch": 0.77,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019493455099221758,
+      "loss": 2.1724,
       "step": 8270
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.3208367536265e-05,
-      "loss": 2.1314,
+      "epoch": 0.77,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019492183833872003,
+      "loss": 2.121,
       "step": 8275
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.297287938863963e-05,
-      "loss": 2.1448,
+      "epoch": 0.77,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019490911016844674,
+      "loss": 2.1625,
       "step": 8280
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.224609375,
-      "learning_rate": 6.273774873354427e-05,
-      "loss": 2.079,
+      "epoch": 0.77,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019489636648347823,
+      "loss": 2.1843,
       "step": 8285
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.25029761844157e-05,
-      "loss": 2.0911,
+      "epoch": 0.77,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019488360728589784,
+      "loss": 2.2019,
       "step": 8290
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.226856235375615e-05,
-      "loss": 2.0911,
+      "epoch": 0.77,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019487083257779122,
+      "loss": 2.1859,
       "step": 8295
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.203450785313208e-05,
-      "loss": 2.1259,
+      "epoch": 0.77,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019485804236124663,
+      "loss": 2.1987,
       "step": 8300
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.180081329317259e-05,
-      "loss": 2.1198,
+      "epoch": 0.77,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001948452366383549,
+      "loss": 2.2313,
       "step": 8305
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.15674792835676e-05,
-      "loss": 2.1614,
+      "epoch": 0.77,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001948324154112094,
+      "loss": 2.1786,
       "step": 8310
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.224609375,
-      "learning_rate": 6.133450643306637e-05,
-      "loss": 2.0845,
+      "epoch": 0.77,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019481957868190593,
+      "loss": 2.1161,
       "step": 8315
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.110189534947606e-05,
-      "loss": 2.1205,
+      "epoch": 0.77,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019480672645254298,
+      "loss": 2.2223,
       "step": 8320
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.086964663965984e-05,
-      "loss": 2.096,
+      "epoch": 0.77,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019479385872522143,
+      "loss": 2.1935,
       "step": 8325
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 6.063776090953568e-05,
-      "loss": 2.1206,
+      "epoch": 0.77,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019478097550204479,
+      "loss": 2.1937,
       "step": 8330
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.0406238764074386e-05,
-      "loss": 2.0934,
+      "epoch": 0.77,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019476807678511904,
+      "loss": 2.1997,
       "step": 8335
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2109375,
-      "learning_rate": 6.017508080729819e-05,
-      "loss": 2.087,
+      "epoch": 0.77,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001947551625765527,
+      "loss": 2.2125,
       "step": 8340
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.9944287642279415e-05,
-      "loss": 2.1131,
+      "epoch": 0.77,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001947422328784569,
+      "loss": 2.1796,
       "step": 8345
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.971385987113836e-05,
-      "loss": 2.128,
+      "epoch": 0.77,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019472928769294522,
+      "loss": 2.1744,
       "step": 8350
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.9483798095042364e-05,
-      "loss": 2.1452,
+      "epoch": 0.77,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019471632702213382,
+      "loss": 2.1339,
       "step": 8355
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.216796875,
-      "learning_rate": 5.9254102914203634e-05,
-      "loss": 2.1297,
+      "epoch": 0.77,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001947033508681413,
+      "loss": 2.1978,
       "step": 8360
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.9024774927878054e-05,
-      "loss": 2.0584,
+      "epoch": 0.78,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019469035923308888,
+      "loss": 2.1683,
       "step": 8365
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.879581473436368e-05,
-      "loss": 2.1078,
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019467735211910038,
+      "loss": 2.1588,
       "step": 8370
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.856722293099881e-05,
-      "loss": 2.1047,
+      "epoch": 0.78,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019466432952830194,
+      "loss": 2.1595,
       "step": 8375
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.8339000114160666e-05,
-      "loss": 2.0854,
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019465129146282236,
+      "loss": 2.1986,
       "step": 8380
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.811114687926402e-05,
-      "loss": 2.1011,
+      "epoch": 0.78,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019463823792479308,
+      "loss": 2.1999,
       "step": 8385
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.788366382075916e-05,
-      "loss": 2.1448,
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019462516891634782,
+      "loss": 2.2308,
       "step": 8390
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.7656551532130945e-05,
-      "loss": 2.1345,
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000194612084439623,
+      "loss": 2.2289,
       "step": 8395
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.742981060589643e-05,
-      "loss": 2.0944,
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019459898449675756,
+      "loss": 2.1207,
       "step": 8400
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.7203441633604294e-05,
-      "loss": 2.0984,
+      "epoch": 0.78,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019458586908989287,
+      "loss": 2.1695,
       "step": 8405
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.697744520583268e-05,
-      "loss": 2.086,
+      "epoch": 0.78,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019457273822117297,
+      "loss": 2.2032,
       "step": 8410
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.675182191218769e-05,
-      "loss": 2.1013,
+      "epoch": 0.78,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019455959189274425,
+      "loss": 2.1554,
       "step": 8415
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 5.652657234130194e-05,
-      "loss": 2.0664,
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019454643010675582,
+      "loss": 2.1915,
       "step": 8420
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.630169708083324e-05,
-      "loss": 2.1041,
+      "epoch": 0.78,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019453325286535917,
+      "loss": 2.224,
       "step": 8425
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.607719671746259e-05,
-      "loss": 2.0922,
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001945200601707084,
+      "loss": 2.1369,
       "step": 8430
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.5853071836893214e-05,
-      "loss": 2.1008,
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019450685202496008,
+      "loss": 2.1081,
       "step": 8435
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.562932302384834e-05,
-      "loss": 2.0781,
+      "epoch": 0.78,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019449362843027338,
+      "loss": 2.1662,
       "step": 8440
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 5.5405950862070385e-05,
-      "loss": 2.1269,
+      "epoch": 0.78,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019448038938880985,
+      "loss": 2.1653,
       "step": 8445
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.51829559343191e-05,
-      "loss": 2.117,
+      "epoch": 0.78,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001944671349027338,
+      "loss": 2.1657,
       "step": 8450
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.49603388223699e-05,
-      "loss": 2.146,
+      "epoch": 0.78,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019445386497421175,
+      "loss": 2.1766,
       "step": 8455
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.4738100107012615e-05,
-      "loss": 2.0968,
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001944405796054131,
+      "loss": 2.1962,
       "step": 8460
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.451624036804976e-05,
-      "loss": 2.1071,
+      "epoch": 0.78,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019442727879850953,
+      "loss": 2.1296,
       "step": 8465
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.429476018429529e-05,
-      "loss": 2.151,
+      "epoch": 0.78,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019441396255567525,
+      "loss": 2.1488,
       "step": 8470
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.21484375,
-      "learning_rate": 5.407366013357298e-05,
-      "loss": 2.1123,
+      "epoch": 0.79,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001944006308790871,
+      "loss": 2.1376,
       "step": 8475
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.3852940792714476e-05,
-      "loss": 2.1213,
+      "epoch": 0.79,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019438728377092445,
+      "loss": 2.1895,
       "step": 8480
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.363260273755866e-05,
-      "loss": 2.1184,
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019437392123336907,
+      "loss": 2.1791,
       "step": 8485
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.21484375,
-      "learning_rate": 5.341264654294935e-05,
-      "loss": 2.1055,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019436054326860532,
+      "loss": 2.1119,
       "step": 8490
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.319307278273433e-05,
-      "loss": 2.1069,
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001943471498788201,
+      "loss": 2.1261,
       "step": 8495
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.297388202976352e-05,
-      "loss": 2.1162,
+      "epoch": 0.79,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019433374106620277,
+      "loss": 2.1432,
       "step": 8500
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.275507485588753e-05,
-      "loss": 2.104,
+      "epoch": 0.79,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019432031683294535,
+      "loss": 2.1404,
       "step": 8505
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.22265625,
-      "learning_rate": 5.253665183195653e-05,
-      "loss": 2.1353,
+      "epoch": 0.79,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001943068771812422,
+      "loss": 2.2191,
       "step": 8510
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.231861352781817e-05,
-      "loss": 2.0981,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019429342211329034,
+      "loss": 2.1238,
       "step": 8515
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.21009605123165e-05,
-      "loss": 2.106,
+      "epoch": 0.79,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019427995163128923,
+      "loss": 2.1968,
       "step": 8520
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.224609375,
-      "learning_rate": 5.188369335329051e-05,
-      "loss": 2.1331,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019426646573744087,
+      "loss": 2.1514,
       "step": 8525
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.1666812617572316e-05,
-      "loss": 2.0954,
+      "epoch": 0.79,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001942529644339498,
+      "loss": 2.1589,
       "step": 8530
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.145031887098606e-05,
-      "loss": 2.1032,
+      "epoch": 0.79,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019423944772302307,
+      "loss": 2.177,
       "step": 8535
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 5.1234212678346184e-05,
-      "loss": 2.0879,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019422591560687023,
+      "loss": 2.1838,
       "step": 8540
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.1018494603455934e-05,
-      "loss": 2.0836,
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019421236808770336,
+      "loss": 2.1842,
       "step": 8545
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 5.080316520910622e-05,
-      "loss": 2.1151,
+      "epoch": 0.79,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019419880516773706,
+      "loss": 2.1425,
       "step": 8550
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.0588225057073744e-05,
-      "loss": 2.1116,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019418522684918843,
+      "loss": 2.1671,
       "step": 8555
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.037367470811969e-05,
-      "loss": 2.1097,
+      "epoch": 0.79,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019417163313427716,
+      "loss": 2.1678,
       "step": 8560
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.015951472198845e-05,
-      "loss": 2.1463,
+      "epoch": 0.79,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019415802402522532,
+      "loss": 2.1974,
       "step": 8565
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 4.994574565740577e-05,
-      "loss": 2.1102,
+      "epoch": 0.79,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019414439952425766,
+      "loss": 2.1889,
       "step": 8570
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.973236807207773e-05,
-      "loss": 2.1069,
+      "epoch": 0.79,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019413075963360131,
+      "loss": 2.1439,
       "step": 8575
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.212890625,
-      "learning_rate": 4.9519382522688903e-05,
-      "loss": 2.1303,
+      "epoch": 0.8,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000194117104355486,
+      "loss": 2.203,
       "step": 8580
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.930678956490109e-05,
-      "loss": 2.1173,
+      "epoch": 0.8,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001941034336921439,
+      "loss": 2.1759,
       "step": 8585
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.21484375,
-      "learning_rate": 4.909458975335202e-05,
-      "loss": 2.1001,
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019408974764580978,
+      "loss": 2.2157,
       "step": 8590
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.888278364165355e-05,
-      "loss": 2.0999,
+      "epoch": 0.8,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019407604621872086,
+      "loss": 2.1716,
       "step": 8595
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.8671371782390405e-05,
-      "loss": 2.1242,
+      "epoch": 0.8,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00019406232941311693,
+      "loss": 2.1967,
       "step": 8600
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.212890625,
-      "learning_rate": 4.846035472711894e-05,
-      "loss": 2.1004,
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019404859723124023,
+      "loss": 2.1332,
       "step": 8605
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 4.824973302636522e-05,
-      "loss": 2.1379,
+      "epoch": 0.8,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019403484967533556,
+      "loss": 2.1246,
       "step": 8610
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.803950722962414e-05,
-      "loss": 2.1155,
+      "epoch": 0.8,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001940210867476502,
+      "loss": 2.1593,
       "step": 8615
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.7829677885357526e-05,
-      "loss": 2.0763,
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000194007308450434,
+      "loss": 2.1461,
       "step": 8620
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.762024554099285e-05,
-      "loss": 2.1104,
+      "epoch": 0.8,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019399351478593927,
+      "loss": 2.1063,
       "step": 8625
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 4.741121074292209e-05,
-      "loss": 2.1256,
+      "epoch": 0.8,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019397970575642083,
+      "loss": 2.1947,
       "step": 8630
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.720257403649977e-05,
-      "loss": 2.085,
+      "epoch": 0.8,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000193965881364136,
+      "loss": 2.1553,
       "step": 8635
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.69943359660421e-05,
-      "loss": 2.1261,
+      "epoch": 0.8,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001939520416113447,
+      "loss": 2.145,
       "step": 8640
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 4.6786497074825056e-05,
-      "loss": 2.1117,
+      "epoch": 0.8,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019393818650030928,
+      "loss": 2.1647,
       "step": 8645
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 4.657905790508324e-05,
-      "loss": 2.1095,
+      "epoch": 0.8,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001939243160332946,
+      "loss": 2.1716,
       "step": 8650
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.637201899800856e-05,
-      "loss": 2.0897,
+      "epoch": 0.8,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001939104302125681,
+      "loss": 2.1484,
       "step": 8655
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.616538089374851e-05,
-      "loss": 2.0951,
+      "epoch": 0.8,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001938965290403996,
+      "loss": 2.1198,
       "step": 8660
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.5959144131404896e-05,
-      "loss": 2.1346,
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019388261251906158,
+      "loss": 2.1514,
       "step": 8665
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 4.575330924903265e-05,
-      "loss": 2.125,
+      "epoch": 0.8,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019386868065082897,
+      "loss": 2.1342,
       "step": 8670
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 4.554787678363801e-05,
-      "loss": 2.1235,
+      "epoch": 0.8,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019385473343797914,
+      "loss": 2.2072,
       "step": 8675
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2109375,
-      "learning_rate": 4.5342847271177616e-05,
-      "loss": 2.0816,
+      "epoch": 0.8,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019384077088279206,
+      "loss": 2.188,
       "step": 8680
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.23046875,
-      "learning_rate": 4.513822124655644e-05,
-      "loss": 2.133,
+      "epoch": 0.8,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019382679298755016,
+      "loss": 2.2,
       "step": 8685
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2265625,
-      "learning_rate": 4.493399924362713e-05,
-      "loss": 2.0935,
+      "epoch": 0.81,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019381279975453839,
+      "loss": 2.1445,
       "step": 8690
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 4.47301817951882e-05,
-      "loss": 2.1127,
+      "epoch": 0.81,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019379879118604424,
+      "loss": 2.1995,
       "step": 8695
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.452676943298264e-05,
-      "loss": 2.1171,
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019378476728435767,
+      "loss": 2.1377,
       "step": 8700
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.4323762687696604e-05,
-      "loss": 2.1185,
+      "epoch": 0.81,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001937707280517711,
+      "loss": 2.1786,
       "step": 8705
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.4121162088958026e-05,
-      "loss": 2.088,
+      "epoch": 0.81,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019375667349057956,
+      "loss": 2.1527,
       "step": 8710
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.391896816533532e-05,
-      "loss": 2.1312,
+      "epoch": 0.81,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019374260360308053,
+      "loss": 2.1399,
       "step": 8715
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.3717181444335964e-05,
-      "loss": 2.0713,
+      "epoch": 0.81,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019372851839157397,
+      "loss": 2.1904,
       "step": 8720
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.351580245240476e-05,
-      "loss": 2.1042,
+      "epoch": 0.81,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001937144178583624,
+      "loss": 2.1465,
       "step": 8725
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2099609375,
-      "learning_rate": 4.331483171492319e-05,
-      "loss": 2.0868,
+      "epoch": 0.81,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019370030200575086,
+      "loss": 2.1376,
       "step": 8730
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2109375,
-      "learning_rate": 4.3114269756207315e-05,
-      "loss": 2.0766,
+      "epoch": 0.81,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019368617083604677,
+      "loss": 2.1552,
       "step": 8735
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.2914117099506926e-05,
-      "loss": 2.1004,
+      "epoch": 0.81,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001936720243515602,
+      "loss": 2.1935,
       "step": 8740
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.271437426700391e-05,
-      "loss": 2.0826,
+      "epoch": 0.81,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019365786255460364,
+      "loss": 2.1631,
       "step": 8745
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.251504177981085e-05,
-      "loss": 2.1315,
+      "epoch": 0.81,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019364368544749208,
+      "loss": 2.1797,
       "step": 8750
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.231612015796995e-05,
-      "loss": 2.1251,
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019362949303254308,
+      "loss": 2.1294,
       "step": 8755
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.211760992045153e-05,
-      "loss": 2.1379,
+      "epoch": 0.81,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019361528531207665,
+      "loss": 2.1606,
       "step": 8760
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.19195115851523e-05,
-      "loss": 2.0842,
+      "epoch": 0.81,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001936010622884153,
+      "loss": 2.1826,
       "step": 8765
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 4.1721825668894776e-05,
-      "loss": 2.1471,
+      "epoch": 0.81,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019358682396388403,
+      "loss": 2.2168,
       "step": 8770
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.152455268742517e-05,
-      "loss": 2.0909,
+      "epoch": 0.81,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019357257034081042,
+      "loss": 2.1742,
       "step": 8775
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 4.132769315541267e-05,
-      "loss": 2.1121,
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019355830142152444,
+      "loss": 2.1691,
       "step": 8780
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.1131247586447597e-05,
-      "loss": 2.1032,
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019354401720835868,
+      "loss": 2.2055,
       "step": 8785
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.0935216493040285e-05,
-      "loss": 2.0749,
+      "epoch": 0.81,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001935297177036481,
+      "loss": 2.0923,
       "step": 8790
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.07396003866199e-05,
-      "loss": 2.107,
+      "epoch": 0.81,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00019351540290973027,
+      "loss": 2.1975,
       "step": 8795
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 4.054439977753282e-05,
-      "loss": 2.1109,
+      "epoch": 0.82,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019350107282894516,
+      "loss": 2.1858,
       "step": 8800
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.034961517504132e-05,
-      "loss": 2.1148,
+      "epoch": 0.82,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019348672746363538,
+      "loss": 2.1437,
       "step": 8805
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.220703125,
-      "learning_rate": 4.015524708732263e-05,
-      "loss": 2.1391,
+      "epoch": 0.82,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019347236681614588,
+      "loss": 2.1789,
       "step": 8810
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 3.9961296021467056e-05,
-      "loss": 2.1037,
+      "epoch": 0.82,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001934579908888242,
+      "loss": 2.1476,
       "step": 8815
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.224609375,
-      "learning_rate": 3.976776248347709e-05,
-      "loss": 2.1215,
+      "epoch": 0.82,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019344359968402036,
+      "loss": 2.166,
       "step": 8820
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 3.9574646978265874e-05,
-      "loss": 2.0873,
+      "epoch": 0.82,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019342919320408688,
+      "loss": 2.1579,
       "step": 8825
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 3.938195000965588e-05,
-      "loss": 2.0711,
+      "epoch": 0.82,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019341477145137876,
+      "loss": 2.2329,
       "step": 8830
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.918967208037778e-05,
-      "loss": 2.0757,
+      "epoch": 0.82,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019340033442825355,
+      "loss": 2.1866,
       "step": 8835
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 3.8997813692068944e-05,
-      "loss": 2.0977,
+      "epoch": 0.82,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019338588213707117,
+      "loss": 2.1257,
       "step": 8840
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.880637534527207e-05,
-      "loss": 2.1179,
+      "epoch": 0.82,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019337141458019422,
+      "loss": 2.1262,
       "step": 8845
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 3.861535753943426e-05,
-      "loss": 2.0717,
+      "epoch": 0.82,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019335693175998766,
+      "loss": 2.1905,
       "step": 8850
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.842476077290522e-05,
-      "loss": 2.0851,
+      "epoch": 0.82,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019334243367881895,
+      "loss": 2.1048,
       "step": 8855
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.8234585542936376e-05,
-      "loss": 2.1166,
+      "epoch": 0.82,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001933279203390581,
+      "loss": 2.1634,
       "step": 8860
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.804483234567928e-05,
-      "loss": 2.0894,
+      "epoch": 0.82,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019331339174307762,
+      "loss": 2.2218,
       "step": 8865
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 3.7855501676184416e-05,
-      "loss": 2.0819,
+      "epoch": 0.82,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001932988478932524,
+      "loss": 2.1271,
       "step": 8870
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 3.766659402840011e-05,
-      "loss": 2.1043,
+      "epoch": 0.82,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019328428879196002,
+      "loss": 2.1248,
       "step": 8875
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.234375,
-      "learning_rate": 3.747810989517087e-05,
-      "loss": 2.1277,
+      "epoch": 0.82,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019326971444158034,
+      "loss": 2.1441,
       "step": 8880
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.7290049768236316e-05,
-      "loss": 2.1355,
+      "epoch": 0.82,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001932551248444959,
+      "loss": 2.187,
       "step": 8885
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.710241413822999e-05,
-      "loss": 2.0989,
+      "epoch": 0.82,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001932405200030916,
+      "loss": 2.1877,
       "step": 8890
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.691520349467781e-05,
-      "loss": 2.1278,
+      "epoch": 0.82,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019322589991975487,
+      "loss": 2.1991,
       "step": 8895
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 3.672841832599712e-05,
-      "loss": 2.112,
+      "epoch": 0.82,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019321126459687565,
+      "loss": 2.2131,
       "step": 8900
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.6542059119494954e-05,
-      "loss": 2.1045,
+      "epoch": 0.83,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019319661403684638,
+      "loss": 2.178,
       "step": 8905
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.635612636136727e-05,
-      "loss": 2.0608,
+      "epoch": 0.83,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019318194824206194,
+      "loss": 2.1562,
       "step": 8910
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.617062053669749e-05,
-      "loss": 2.1293,
+      "epoch": 0.83,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019316726721491976,
+      "loss": 2.2305,
       "step": 8915
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.212890625,
-      "learning_rate": 3.5985542129455e-05,
-      "loss": 2.1204,
+      "epoch": 0.83,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019315257095781974,
+      "loss": 2.1503,
       "step": 8920
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 3.580089162249429e-05,
-      "loss": 2.1028,
+      "epoch": 0.83,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019313785947316423,
+      "loss": 2.1152,
       "step": 8925
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.5616669497553336e-05,
-      "loss": 2.1302,
+      "epoch": 0.83,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019312313276335813,
+      "loss": 2.1782,
       "step": 8930
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.543287623525256e-05,
-      "loss": 2.0837,
+      "epoch": 0.83,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019310839083080876,
+      "loss": 2.1189,
       "step": 8935
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.524951231509359e-05,
-      "loss": 2.1037,
+      "epoch": 0.83,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.000193093633677926,
+      "loss": 2.1981,
       "step": 8940
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.23046875,
-      "learning_rate": 3.506657821545785e-05,
-      "loss": 2.1115,
+      "epoch": 0.83,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001930788613071222,
+      "loss": 2.1657,
       "step": 8945
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.488407441360535e-05,
-      "loss": 2.1232,
+      "epoch": 0.83,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019306407372081218,
+      "loss": 2.1373,
       "step": 8950
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.4702001385673634e-05,
-      "loss": 2.1757,
+      "epoch": 0.83,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001930492709214132,
+      "loss": 2.1533,
       "step": 8955
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.452035960667623e-05,
-      "loss": 2.1113,
+      "epoch": 0.83,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019303445291134511,
+      "loss": 2.1715,
       "step": 8960
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 3.4339149550501836e-05,
-      "loss": 2.1785,
+      "epoch": 0.83,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019301961969303018,
+      "loss": 2.1122,
       "step": 8965
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.415837168991238e-05,
-      "loss": 2.0961,
+      "epoch": 0.83,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001930047712688932,
+      "loss": 2.1325,
       "step": 8970
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 3.397802649654265e-05,
-      "loss": 2.0961,
+      "epoch": 0.83,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001929899076413614,
+      "loss": 2.2475,
       "step": 8975
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.379811444089849e-05,
-      "loss": 2.1169,
+      "epoch": 0.83,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019297502881286454,
+      "loss": 2.1966,
       "step": 8980
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.224609375,
-      "learning_rate": 3.3618635992355686e-05,
-      "loss": 2.1381,
+      "epoch": 0.83,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019296013478583478,
+      "loss": 2.1531,
       "step": 8985
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.343959161915877e-05,
-      "loss": 2.1424,
+      "epoch": 0.83,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019294522556270697,
+      "loss": 2.1944,
       "step": 8990
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.326098178841983e-05,
-      "loss": 2.1002,
+      "epoch": 0.83,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019293030114591816,
+      "loss": 2.1965,
       "step": 8995
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.30828069661173e-05,
-      "loss": 2.1073,
+      "epoch": 0.83,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019291536153790814,
+      "loss": 2.2275,
       "step": 9000
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.2905067617094796e-05,
-      "loss": 2.1393,
+      "epoch": 0.83,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019290040674111903,
+      "loss": 2.1475,
       "step": 9005
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.272776420505954e-05,
-      "loss": 2.1097,
+      "epoch": 0.83,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019288543675799545,
+      "loss": 2.1354,
       "step": 9010
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 3.255089719258173e-05,
-      "loss": 2.0935,
+      "epoch": 0.84,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019287045159098454,
+      "loss": 2.2582,
       "step": 9015
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.212890625,
-      "learning_rate": 3.237446704109284e-05,
-      "loss": 2.1066,
+      "epoch": 0.84,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019285545124253593,
+      "loss": 2.1205,
       "step": 9020
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.219847421088473e-05,
-      "loss": 2.0793,
+      "epoch": 0.84,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001928404357151017,
+      "loss": 2.1876,
       "step": 9025
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.202291916110827e-05,
-      "loss": 2.1244,
+      "epoch": 0.84,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019282540501113643,
+      "loss": 2.1627,
       "step": 9030
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.184780234977214e-05,
-      "loss": 2.0899,
+      "epoch": 0.84,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019281035913309714,
+      "loss": 2.1737,
       "step": 9035
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.1673124233741824e-05,
-      "loss": 2.0989,
+      "epoch": 0.84,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019279529808344345,
+      "loss": 2.1521,
       "step": 9040
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.21484375,
-      "learning_rate": 3.149888526873821e-05,
-      "loss": 2.1033,
+      "epoch": 0.84,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001927802218646373,
+      "loss": 2.1624,
       "step": 9045
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 3.132508590933638e-05,
-      "loss": 2.1003,
+      "epoch": 0.84,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019276513047914318,
+      "loss": 2.1407,
       "step": 9050
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.115172660896477e-05,
-      "loss": 2.1204,
+      "epoch": 0.84,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019275002392942807,
+      "loss": 2.1457,
       "step": 9055
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.097880781990348e-05,
-      "loss": 2.1247,
+      "epoch": 0.84,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019273490221796146,
+      "loss": 2.1907,
       "step": 9060
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.2109375,
-      "learning_rate": 3.080632999328352e-05,
-      "loss": 2.0663,
+      "epoch": 0.84,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019271976534721526,
+      "loss": 2.1547,
       "step": 9065
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.0634293579085404e-05,
-      "loss": 2.1224,
+      "epoch": 0.84,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019270461331966386,
+      "loss": 2.1317,
       "step": 9070
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.0462699026137966e-05,
-      "loss": 2.1124,
+      "epoch": 0.84,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019268944613778417,
+      "loss": 2.1524,
       "step": 9075
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.224609375,
-      "learning_rate": 3.0291546782117452e-05,
-      "loss": 2.0772,
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019267426380405552,
+      "loss": 2.1987,
       "step": 9080
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.012083729354598e-05,
-      "loss": 2.0876,
+      "epoch": 0.84,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001926590663209598,
+      "loss": 2.1827,
       "step": 9085
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.9950571005790573e-05,
-      "loss": 2.0978,
+      "epoch": 0.84,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019264385369098131,
+      "loss": 2.1163,
       "step": 9090
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.978074836306213e-05,
-      "loss": 2.1272,
+      "epoch": 0.84,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019262862591660685,
+      "loss": 2.1526,
       "step": 9095
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 2.9611369808413904e-05,
-      "loss": 2.1163,
+      "epoch": 0.84,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019261338300032566,
+      "loss": 2.1926,
       "step": 9100
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 2.9442435783740775e-05,
-      "loss": 2.1526,
+      "epoch": 0.84,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019259812494462947,
+      "loss": 2.185,
       "step": 9105
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.9273946729777702e-05,
-      "loss": 2.1216,
+      "epoch": 0.84,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019258285175201257,
+      "loss": 2.1416,
       "step": 9110
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 2.9105903086098775e-05,
-      "loss": 2.1364,
+      "epoch": 0.84,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019256756342497157,
+      "loss": 2.1877,
       "step": 9115
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.232421875,
-      "learning_rate": 2.893830529111621e-05,
-      "loss": 2.0959,
+      "epoch": 0.85,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019255225996600572,
+      "loss": 2.2229,
       "step": 9120
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.8771153782078864e-05,
-      "loss": 2.0946,
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001925369413776166,
+      "loss": 2.1976,
       "step": 9125
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.8604448995071285e-05,
-      "loss": 2.1365,
+      "epoch": 0.85,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019252160766230833,
+      "loss": 2.1635,
       "step": 9130
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.8438191365012667e-05,
-      "loss": 2.14,
+      "epoch": 0.85,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019250625882258752,
+      "loss": 2.1934,
       "step": 9135
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.8272381325655483e-05,
-      "loss": 2.0476,
+      "epoch": 0.85,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019249089486096324,
+      "loss": 2.1948,
       "step": 9140
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 2.8107019309584636e-05,
-      "loss": 2.0812,
+      "epoch": 0.85,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019247551577994697,
+      "loss": 2.1516,
       "step": 9145
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.794210574821601e-05,
-      "loss": 2.0666,
+      "epoch": 0.85,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019246012158205276,
+      "loss": 2.1766,
       "step": 9150
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.7777641071795545e-05,
-      "loss": 2.1169,
+      "epoch": 0.85,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019244471226979705,
+      "loss": 2.1841,
       "step": 9155
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 2.761362570939816e-05,
-      "loss": 2.0937,
+      "epoch": 0.85,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001924292878456988,
+      "loss": 2.1728,
       "step": 9160
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.74500600889265e-05,
-      "loss": 2.0642,
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019241384831227945,
+      "loss": 2.15,
       "step": 9165
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.7286944637109812e-05,
-      "loss": 2.1106,
+      "epoch": 0.85,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019239839367206284,
+      "loss": 2.2386,
       "step": 9170
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.7124279779503005e-05,
-      "loss": 2.1315,
+      "epoch": 0.85,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00019238292392757535,
+      "loss": 2.176,
       "step": 9175
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.6962065940485294e-05,
-      "loss": 2.0988,
+      "epoch": 0.85,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019236743908134584,
+      "loss": 2.1694,
       "step": 9180
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.2109375,
-      "learning_rate": 2.680030354325944e-05,
-      "loss": 2.0955,
+      "epoch": 0.85,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001923519391359055,
+      "loss": 2.1998,
       "step": 9185
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.2265625,
-      "learning_rate": 2.663899300985009e-05,
-      "loss": 2.1067,
+      "epoch": 0.85,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019233642409378815,
+      "loss": 2.1707,
       "step": 9190
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.22265625,
-      "learning_rate": 2.647813476110328e-05,
-      "loss": 2.1418,
+      "epoch": 0.85,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019232089395753005,
+      "loss": 2.0698,
       "step": 9195
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.6317729216685095e-05,
-      "loss": 2.0693,
+      "epoch": 0.85,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019230534872966985,
+      "loss": 2.117,
       "step": 9200
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.224609375,
-      "learning_rate": 2.6157776795080357e-05,
-      "loss": 2.1104,
+      "epoch": 0.85,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019228978841274872,
+      "loss": 2.142,
       "step": 9205
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.5998277913591863e-05,
-      "loss": 2.0949,
+      "epoch": 0.85,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001922742130093103,
+      "loss": 2.1773,
       "step": 9210
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.583923298833906e-05,
-      "loss": 2.1104,
+      "epoch": 0.85,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001922586225219007,
+      "loss": 2.1977,
       "step": 9215
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.224609375,
-      "learning_rate": 2.5680642434257208e-05,
-      "loss": 2.113,
+      "epoch": 0.85,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019224301695306843,
+      "loss": 2.1377,
       "step": 9220
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.552250666509606e-05,
-      "loss": 2.1012,
+      "epoch": 0.85,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019222739630536456,
+      "loss": 2.1539,
       "step": 9225
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.536482609341888e-05,
-      "loss": 2.0663,
+      "epoch": 0.86,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019221176058134258,
+      "loss": 2.155,
       "step": 9230
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.224609375,
-      "learning_rate": 2.5207601130601322e-05,
-      "loss": 2.1224,
+      "epoch": 0.86,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001921961097835584,
+      "loss": 2.177,
       "step": 9235
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.224609375,
-      "learning_rate": 2.5050832186830442e-05,
-      "loss": 2.1241,
+      "epoch": 0.86,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019218044391457051,
+      "loss": 2.213,
       "step": 9240
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.4894519671103546e-05,
-      "loss": 2.1329,
+      "epoch": 0.86,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019216476297693978,
+      "loss": 2.2192,
       "step": 9245
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.21484375,
-      "learning_rate": 2.473866399122733e-05,
-      "loss": 2.1264,
+      "epoch": 0.86,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001921490669732295,
+      "loss": 2.113,
       "step": 9250
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.458326555381627e-05,
-      "loss": 2.091,
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001921333559060055,
+      "loss": 2.1809,
       "step": 9255
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 2.4428324764292265e-05,
-      "loss": 2.1031,
+      "epoch": 0.86,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019211762977783609,
+      "loss": 2.1594,
       "step": 9260
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.4273842026883165e-05,
-      "loss": 2.1216,
+      "epoch": 0.86,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019210188859129196,
+      "loss": 2.1447,
       "step": 9265
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 2.4119817744621777e-05,
-      "loss": 2.0704,
+      "epoch": 0.86,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019208613234894637,
+      "loss": 2.23,
       "step": 9270
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.3966252319344774e-05,
-      "loss": 2.0667,
+      "epoch": 0.86,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001920703610533749,
+      "loss": 2.1734,
       "step": 9275
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.3813146151691766e-05,
-      "loss": 2.1076,
+      "epoch": 0.86,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019205457470715574,
+      "loss": 2.2056,
       "step": 9280
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.3660499641104217e-05,
-      "loss": 2.1104,
+      "epoch": 0.86,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019203877331286937,
+      "loss": 2.1412,
       "step": 9285
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.3508313185824472e-05,
-      "loss": 2.131,
+      "epoch": 0.86,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019202295687309893,
+      "loss": 2.127,
       "step": 9290
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.3356587182894353e-05,
-      "loss": 2.0922,
+      "epoch": 0.86,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019200712539042987,
+      "loss": 2.1003,
       "step": 9295
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.23046875,
-      "learning_rate": 2.3205322028154687e-05,
-      "loss": 2.0962,
+      "epoch": 0.86,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019199127886745013,
+      "loss": 2.1931,
       "step": 9300
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.3054518116243773e-05,
-      "loss": 2.0909,
+      "epoch": 0.86,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019197541730675015,
+      "loss": 2.2037,
       "step": 9305
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.2904175840596742e-05,
-      "loss": 2.0702,
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001919595407109228,
+      "loss": 2.2092,
       "step": 9310
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 2.2754295593444262e-05,
-      "loss": 2.1061,
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019194364908256342,
+      "loss": 2.2276,
       "step": 9315
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.2604877765811573e-05,
-      "loss": 2.1015,
+      "epoch": 0.86,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019192774242426976,
+      "loss": 2.1501,
       "step": 9320
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 2.24559227475176e-05,
-      "loss": 2.1243,
+      "epoch": 0.86,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001919118207386421,
+      "loss": 2.1766,
       "step": 9325
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 2.2307430927173755e-05,
-      "loss": 2.1422,
+      "epoch": 0.86,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019189588402828317,
+      "loss": 2.1452,
       "step": 9330
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.21484375,
-      "learning_rate": 2.2159402692182974e-05,
-      "loss": 2.0982,
+      "epoch": 0.86,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001918799322957981,
+      "loss": 2.1436,
       "step": 9335
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.2011838428738908e-05,
-      "loss": 2.1523,
+      "epoch": 0.87,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019186396554379443,
+      "loss": 2.2139,
       "step": 9340
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.186473852182451e-05,
-      "loss": 2.0777,
+      "epoch": 0.87,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019184798377488234,
+      "loss": 2.1739,
       "step": 9345
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.228515625,
-      "learning_rate": 2.1718103355211496e-05,
-      "loss": 2.068,
+      "epoch": 0.87,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001918319869916743,
+      "loss": 2.1876,
       "step": 9350
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 2.1571933311458924e-05,
-      "loss": 2.1138,
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019181597519678527,
+      "loss": 2.1545,
       "step": 9355
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.142622877191247e-05,
-      "loss": 2.1396,
+      "epoch": 0.87,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019179994839283273,
+      "loss": 2.1768,
       "step": 9360
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.128099011670339e-05,
-      "loss": 2.1507,
+      "epoch": 0.87,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019178390658243654,
+      "loss": 2.1725,
       "step": 9365
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.1136217724747386e-05,
-      "loss": 2.0835,
+      "epoch": 0.87,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019176784976821903,
+      "loss": 2.0755,
       "step": 9370
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 2.0991911973743748e-05,
-      "loss": 2.0951,
+      "epoch": 0.87,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019175177795280503,
+      "loss": 2.1444,
       "step": 9375
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.0848073240174438e-05,
-      "loss": 2.0804,
+      "epoch": 0.87,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019173569113882172,
+      "loss": 2.1718,
       "step": 9380
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.0704701899302827e-05,
-      "loss": 2.111,
+      "epoch": 0.87,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019171958932889886,
+      "loss": 2.1661,
       "step": 9385
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 2.0561798325173067e-05,
-      "loss": 2.1172,
+      "epoch": 0.87,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019170347252566857,
+      "loss": 2.1792,
       "step": 9390
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 2.041936289060886e-05,
-      "loss": 2.1233,
+      "epoch": 0.87,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001916873407317654,
+      "loss": 2.2363,
       "step": 9395
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.0277395967212543e-05,
-      "loss": 2.155,
+      "epoch": 0.87,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019167119394982647,
+      "loss": 2.1836,
       "step": 9400
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.0135897925364277e-05,
-      "loss": 2.0959,
+      "epoch": 0.87,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019165503218249126,
+      "loss": 2.1413,
       "step": 9405
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.9994869134220772e-05,
-      "loss": 2.1398,
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019163885543240167,
+      "loss": 2.1403,
       "step": 9410
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.98543099617146e-05,
-      "loss": 2.0972,
+      "epoch": 0.87,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019162266370220215,
+      "loss": 2.1949,
       "step": 9415
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.971422077455316e-05,
-      "loss": 2.1109,
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019160645699453955,
+      "loss": 2.1768,
       "step": 9420
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.957460193821763e-05,
-      "loss": 2.0976,
+      "epoch": 0.87,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019159023531206313,
+      "loss": 2.1624,
       "step": 9425
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.23046875,
-      "learning_rate": 1.943545381696219e-05,
-      "loss": 2.1277,
+      "epoch": 0.87,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019157399865742464,
+      "loss": 2.1827,
       "step": 9430
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.9296776773812762e-05,
-      "loss": 2.0844,
+      "epoch": 0.87,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019155774703327828,
+      "loss": 2.1837,
       "step": 9435
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.9158571170566453e-05,
-      "loss": 2.0883,
+      "epoch": 0.87,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019154148044228067,
+      "loss": 2.1746,
       "step": 9440
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2265625,
-      "learning_rate": 1.902083736779039e-05,
-      "loss": 2.149,
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019152519888709094,
+      "loss": 2.1667,
       "step": 9445
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.888357572482076e-05,
-      "loss": 2.1262,
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019150890237037055,
+      "loss": 2.118,
       "step": 9450
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.8746786599761878e-05,
-      "loss": 2.1061,
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019149259089478352,
+      "loss": 2.1682,
       "step": 9455
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.22265625,
-      "learning_rate": 1.8610470349485444e-05,
-      "loss": 2.0833,
+      "epoch": 0.88,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019147626446299623,
+      "loss": 2.1778,
       "step": 9460
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.8474627329629348e-05,
-      "loss": 2.0643,
+      "epoch": 0.88,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001914599230776776,
+      "loss": 2.1192,
       "step": 9465
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.8339257894596962e-05,
-      "loss": 2.1025,
+      "epoch": 0.88,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019144356674149888,
+      "loss": 2.1963,
       "step": 9470
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.22265625,
-      "learning_rate": 1.8204362397555917e-05,
-      "loss": 2.0692,
+      "epoch": 0.88,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019142719545713389,
+      "loss": 2.1483,
       "step": 9475
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.806994119043761e-05,
-      "loss": 2.14,
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001914108092272588,
+      "loss": 2.1461,
       "step": 9480
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.7935994623935957e-05,
-      "loss": 2.1017,
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019139440805455223,
+      "loss": 2.1304,
       "step": 9485
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.7802523047506582e-05,
-      "loss": 2.0842,
+      "epoch": 0.88,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019137799194169527,
+      "loss": 2.1942,
       "step": 9490
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.7669526809365845e-05,
-      "loss": 2.0907,
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019136156089137146,
+      "loss": 2.1631,
       "step": 9495
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.7537006256489997e-05,
-      "loss": 2.096,
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019134511490626673,
+      "loss": 2.1382,
       "step": 9500
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.740496173461439e-05,
-      "loss": 2.1358,
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019132865398906957,
+      "loss": 2.0967,
       "step": 9505
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.7273393588232344e-05,
-      "loss": 2.0871,
+      "epoch": 0.88,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019131217814247073,
+      "loss": 2.1868,
       "step": 9510
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.7142302160594383e-05,
-      "loss": 2.1203,
+      "epoch": 0.88,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019129568736916356,
+      "loss": 2.1975,
       "step": 9515
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.23046875,
-      "learning_rate": 1.7011687793707252e-05,
-      "loss": 2.0944,
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019127918167184378,
+      "loss": 2.2046,
       "step": 9520
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.22265625,
-      "learning_rate": 1.6881550828333137e-05,
-      "loss": 2.1036,
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019126266105320955,
+      "loss": 2.169,
       "step": 9525
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.6751891603988735e-05,
-      "loss": 2.1204,
+      "epoch": 0.88,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019124612551596148,
+      "loss": 2.2014,
       "step": 9530
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.6622710458944413e-05,
-      "loss": 2.16,
+      "epoch": 0.88,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019122957506280264,
+      "loss": 2.2044,
       "step": 9535
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.649400773022305e-05,
-      "loss": 2.1025,
+      "epoch": 0.88,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001912130096964385,
+      "loss": 2.1845,
       "step": 9540
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.6365783753599608e-05,
-      "loss": 2.129,
+      "epoch": 0.88,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019119642941957697,
+      "loss": 2.1972,
       "step": 9545
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.6238038863599912e-05,
-      "loss": 2.0574,
+      "epoch": 0.88,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001911798342349284,
+      "loss": 2.1524,
       "step": 9550
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.6110773393499934e-05,
-      "loss": 2.0829,
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019116322414520562,
+      "loss": 2.1327,
       "step": 9555
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.5983987675324853e-05,
-      "loss": 2.0842,
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019114659915312387,
+      "loss": 2.141,
       "step": 9560
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.5857682039848187e-05,
-      "loss": 2.094,
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019112995926140081,
+      "loss": 2.1788,
       "step": 9565
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.5731856816591017e-05,
-      "loss": 2.1373,
+      "epoch": 0.89,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019111330447275652,
+      "loss": 2.1791,
       "step": 9570
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.5606512333821065e-05,
-      "loss": 2.0888,
+      "epoch": 0.89,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019109663478991356,
+      "loss": 2.0959,
       "step": 9575
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2265625,
-      "learning_rate": 1.548164891855175e-05,
-      "loss": 2.1476,
+      "epoch": 0.89,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001910799502155969,
+      "loss": 2.21,
       "step": 9580
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 1.5357266896541555e-05,
-      "loss": 2.0712,
+      "epoch": 0.89,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019106325075253397,
+      "loss": 2.1963,
       "step": 9585
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.5233366592292909e-05,
-      "loss": 2.1163,
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001910465364034546,
+      "loss": 2.1294,
       "step": 9590
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.510994832905166e-05,
-      "loss": 2.1408,
+      "epoch": 0.89,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001910298071710911,
+      "loss": 2.0931,
       "step": 9595
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2109375,
-      "learning_rate": 1.4987012428805891e-05,
-      "loss": 2.1092,
+      "epoch": 0.89,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019101306305817811,
+      "loss": 2.1308,
       "step": 9600
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.4864559212285245e-05,
-      "loss": 2.0868,
+      "epoch": 0.89,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001909963040674528,
+      "loss": 2.2289,
       "step": 9605
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.4742588998960194e-05,
-      "loss": 2.1332,
+      "epoch": 0.89,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019097953020165483,
+      "loss": 2.1809,
       "step": 9610
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.4621102107041041e-05,
-      "loss": 2.0777,
+      "epoch": 0.89,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019096274146352605,
+      "loss": 2.2123,
       "step": 9615
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 1.4500098853477052e-05,
-      "loss": 2.1007,
+      "epoch": 0.89,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019094593785581103,
+      "loss": 2.1872,
       "step": 9620
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.437957955395588e-05,
-      "loss": 2.1006,
+      "epoch": 0.89,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00019092911938125656,
+      "loss": 2.1291,
       "step": 9625
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.4259544522902458e-05,
-      "loss": 2.0954,
+      "epoch": 0.89,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019091228604261202,
+      "loss": 2.1188,
       "step": 9630
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.4139994073478369e-05,
-      "loss": 2.0669,
+      "epoch": 0.89,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019089543784262905,
+      "loss": 2.1665,
       "step": 9635
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.4020928517580944e-05,
-      "loss": 2.1218,
+      "epoch": 0.89,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019087857478406186,
+      "loss": 2.1641,
       "step": 9640
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.3902348165842438e-05,
-      "loss": 2.1473,
+      "epoch": 0.89,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019086169686966702,
+      "loss": 2.1635,
       "step": 9645
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.3784253327629316e-05,
-      "loss": 2.1085,
+      "epoch": 0.89,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019084480410220357,
+      "loss": 2.2147,
       "step": 9650
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.3666644311041299e-05,
-      "loss": 2.0782,
+      "epoch": 0.89,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019082789648443288,
+      "loss": 2.1416,
       "step": 9655
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.3549521422910639e-05,
-      "loss": 2.1336,
+      "epoch": 0.9,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019081097401911888,
+      "loss": 2.1772,
       "step": 9660
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.343288496880144e-05,
-      "loss": 2.1178,
+      "epoch": 0.9,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019079403670902792,
+      "loss": 2.1817,
       "step": 9665
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.22265625,
-      "learning_rate": 1.3316735253008562e-05,
-      "loss": 2.0642,
+      "epoch": 0.9,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001907770845569286,
+      "loss": 2.1894,
       "step": 9670
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.3201072578557206e-05,
-      "loss": 2.0848,
+      "epoch": 0.9,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019076011756559219,
+      "loss": 2.2462,
       "step": 9675
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.3085897247201729e-05,
-      "loss": 2.0419,
+      "epoch": 0.9,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019074313573779218,
+      "loss": 2.1401,
       "step": 9680
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.2971209559425124e-05,
-      "loss": 2.1245,
+      "epoch": 0.9,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00019072613907630462,
+      "loss": 2.1864,
       "step": 9685
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.2857009814438247e-05,
-      "loss": 2.0752,
+      "epoch": 0.9,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001907091275839079,
+      "loss": 2.136,
       "step": 9690
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.2743298310178842e-05,
-      "loss": 2.1045,
+      "epoch": 0.9,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019069210126338292,
+      "loss": 2.1403,
       "step": 9695
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.236328125,
-      "learning_rate": 1.2630075343310844e-05,
-      "loss": 2.1198,
+      "epoch": 0.9,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019067506011751292,
+      "loss": 2.1471,
       "step": 9700
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.2517341209223787e-05,
-      "loss": 2.1022,
+      "epoch": 0.9,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001906580041490836,
+      "loss": 2.1651,
       "step": 9705
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.2405096202031718e-05,
-      "loss": 2.1245,
+      "epoch": 0.9,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001906409333608831,
+      "loss": 2.1544,
       "step": 9710
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.2293340614572769e-05,
-      "loss": 2.0759,
+      "epoch": 0.9,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019062384775570196,
+      "loss": 2.1741,
       "step": 9715
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.2182074738407955e-05,
-      "loss": 2.1398,
+      "epoch": 0.9,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019060674733633316,
+      "loss": 2.1751,
       "step": 9720
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.2071298863820946e-05,
-      "loss": 2.1084,
+      "epoch": 0.9,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000190589632105572,
+      "loss": 2.1281,
       "step": 9725
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.196101327981689e-05,
-      "loss": 2.0783,
+      "epoch": 0.9,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001905725020662164,
+      "loss": 2.1283,
       "step": 9730
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 1.1851218274121856e-05,
-      "loss": 2.1567,
+      "epoch": 0.9,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019055535722106655,
+      "loss": 2.1431,
       "step": 9735
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.1741914133181997e-05,
-      "loss": 2.1273,
+      "epoch": 0.9,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001905381975729251,
+      "loss": 2.148,
       "step": 9740
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.1633101142162828e-05,
-      "loss": 2.1442,
+      "epoch": 0.9,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019052102312459713,
+      "loss": 2.1477,
       "step": 9745
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.1524779584948576e-05,
-      "loss": 2.1221,
+      "epoch": 0.9,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019050383387889008,
+      "loss": 2.1483,
       "step": 9750
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.1416949744141358e-05,
-      "loss": 2.0778,
+      "epoch": 0.9,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019048662983861393,
+      "loss": 2.212,
       "step": 9755
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.1309611901060257e-05,
-      "loss": 2.1166,
+      "epoch": 0.9,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019046941100658092,
+      "loss": 2.1645,
       "step": 9760
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.1202766335741022e-05,
-      "loss": 2.0887,
+      "epoch": 0.9,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001904521773856059,
+      "loss": 2.2022,
       "step": 9765
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.1096413326934917e-05,
-      "loss": 2.0831,
+      "epoch": 0.91,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019043492897850595,
+      "loss": 2.1782,
       "step": 9770
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.0990553152108263e-05,
-      "loss": 2.0461,
+      "epoch": 0.91,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001904176657881007,
+      "loss": 2.1439,
       "step": 9775
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.0885186087441602e-05,
-      "loss": 2.1048,
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019040038781721212,
+      "loss": 2.1924,
       "step": 9780
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.0780312407828885e-05,
-      "loss": 2.1238,
+      "epoch": 0.91,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019038309506866463,
+      "loss": 2.1593,
       "step": 9785
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.0675932386877008e-05,
-      "loss": 2.0902,
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019036578754528505,
+      "loss": 2.1904,
       "step": 9790
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.0572046296904959e-05,
-      "loss": 2.0718,
+      "epoch": 0.91,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001903484652499026,
+      "loss": 2.1236,
       "step": 9795
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.0468654408942912e-05,
-      "loss": 2.1241,
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019033112818534904,
+      "loss": 2.1677,
       "step": 9800
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 1.03657569927319e-05,
-      "loss": 2.1015,
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019031377635445832,
+      "loss": 2.1202,
       "step": 9805
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2265625,
-      "learning_rate": 1.0263354316722807e-05,
-      "loss": 2.1432,
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019029640976006698,
+      "loss": 2.1753,
       "step": 9810
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.0161446648075857e-05,
-      "loss": 2.1003,
+      "epoch": 0.91,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001902790284050139,
+      "loss": 2.1879,
       "step": 9815
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.0060034252659867e-05,
-      "loss": 2.0906,
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001902616322921404,
+      "loss": 2.1813,
       "step": 9820
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.21484375,
-      "learning_rate": 9.959117395051355e-06,
-      "loss": 2.128,
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019024422142429026,
+      "loss": 2.2223,
       "step": 9825
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.85869633853418e-06,
-      "loss": 2.0762,
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019022679580430954,
+      "loss": 2.1969,
       "step": 9830
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.22265625,
-      "learning_rate": 9.758771345098661e-06,
-      "loss": 2.0651,
+      "epoch": 0.91,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019020935543504681,
+      "loss": 2.2205,
       "step": 9835
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.659342675440909e-06,
-      "loss": 2.148,
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019019190031935304,
+      "loss": 2.2124,
       "step": 9840
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 9.560410588962198e-06,
-      "loss": 2.1074,
+      "epoch": 0.91,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001901744304600816,
+      "loss": 2.2054,
       "step": 9845
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 9.461975343768181e-06,
-      "loss": 2.1155,
+      "epoch": 0.91,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019015694586008825,
+      "loss": 2.165,
       "step": 9850
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.364037196668386e-06,
-      "loss": 2.0541,
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019013944652223122,
+      "loss": 2.1835,
       "step": 9855
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.266596403175443e-06,
-      "loss": 2.0918,
+      "epoch": 0.91,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001901219324493711,
+      "loss": 2.1787,
       "step": 9860
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 9.169653217504315e-06,
-      "loss": 2.0792,
+      "epoch": 0.91,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019010440364437082,
+      "loss": 2.2045,
       "step": 9865
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.21875,
-      "learning_rate": 9.07320789257191e-06,
-      "loss": 2.1151,
+      "epoch": 0.91,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001900868601100959,
+      "loss": 2.1849,
       "step": 9870
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.22265625,
-      "learning_rate": 8.977260679996136e-06,
-      "loss": 2.1125,
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019006930184941413,
+      "loss": 2.2381,
       "step": 9875
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 8.881811830095421e-06,
-      "loss": 2.0979,
+      "epoch": 0.92,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019005172886519568,
+      "loss": 2.1224,
       "step": 9880
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 8.786861591888018e-06,
-      "loss": 2.1435,
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019003414116031331,
+      "loss": 2.159,
       "step": 9885
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.2265625,
-      "learning_rate": 8.692410213091262e-06,
-      "loss": 2.1159,
+      "epoch": 0.92,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00019001653873764195,
+      "loss": 2.1741,
       "step": 9890
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 8.598457940121107e-06,
-      "loss": 2.0526,
+      "epoch": 0.92,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001899989216000591,
+      "loss": 2.1948,
       "step": 9895
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.505005018091327e-06,
-      "loss": 2.1262,
+      "epoch": 0.92,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018998128975044462,
+      "loss": 2.2075,
       "step": 9900
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.412051690812917e-06,
-      "loss": 2.1046,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018996364319168074,
+      "loss": 2.181,
       "step": 9905
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.220703125,
-      "learning_rate": 8.31959820079351e-06,
-      "loss": 2.0946,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018994598192665217,
+      "loss": 2.1989,
       "step": 9910
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 8.227644789236677e-06,
-      "loss": 2.0598,
+      "epoch": 0.92,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018992830595824592,
+      "loss": 2.129,
       "step": 9915
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.136191696041363e-06,
-      "loss": 2.1327,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018991061528935154,
+      "loss": 2.1501,
       "step": 9920
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.21875,
-      "learning_rate": 8.045239159801177e-06,
-      "loss": 2.104,
+      "epoch": 0.92,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018989290992286084,
+      "loss": 2.1867,
       "step": 9925
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 7.954787417803799e-06,
-      "loss": 2.0897,
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001898751898616681,
+      "loss": 2.1629,
       "step": 9930
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.220703125,
-      "learning_rate": 7.864836706030466e-06,
-      "loss": 2.0884,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018985745510867002,
+      "loss": 2.2214,
       "step": 9935
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 7.775387259155186e-06,
-      "loss": 2.1002,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001898397056667657,
+      "loss": 2.1166,
       "step": 9940
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 7.686439310544203e-06,
-      "loss": 2.0566,
+      "epoch": 0.92,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018982194153885663,
+      "loss": 2.1608,
       "step": 9945
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 7.5979930922555285e-06,
-      "loss": 2.0904,
+      "epoch": 0.92,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018980416272784662,
+      "loss": 2.1616,
       "step": 9950
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 7.510048835037986e-06,
-      "loss": 2.1242,
+      "epoch": 0.92,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018978636923664202,
+      "loss": 2.143,
       "step": 9955
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.2265625,
-      "learning_rate": 7.4226067683310195e-06,
-      "loss": 2.0756,
+      "epoch": 0.92,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018976856106815154,
+      "loss": 2.1926,
       "step": 9960
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.21484375,
-      "learning_rate": 7.335667120263767e-06,
-      "loss": 2.1141,
+      "epoch": 0.92,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018975073822528616,
+      "loss": 2.1287,
       "step": 9965
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 7.24923011765466e-06,
-      "loss": 2.1056,
+      "epoch": 0.92,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018973290071095945,
+      "loss": 2.1138,
       "step": 9970
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 7.163295986010776e-06,
-      "loss": 2.0881,
+      "epoch": 0.92,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018971504852808728,
+      "loss": 2.0951,
       "step": 9975
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.216796875,
-      "learning_rate": 7.077864949527202e-06,
-      "loss": 2.0661,
+      "epoch": 0.92,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001896971816795879,
+      "loss": 2.2018,
       "step": 9980
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 6.9929372310864936e-06,
-      "loss": 2.0908,
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018967930016838196,
+      "loss": 2.1758,
       "step": 9985
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 6.90851305225817e-06,
-      "loss": 2.0908,
+      "epoch": 0.93,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001896614039973926,
+      "loss": 2.1175,
       "step": 9990
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.8245926332979105e-06,
-      "loss": 2.1092,
+      "epoch": 0.93,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018964349316954524,
+      "loss": 2.1044,
       "step": 9995
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.741176193147336e-06,
-      "loss": 2.108,
+      "epoch": 0.93,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018962556768776777,
+      "loss": 2.1498,
       "step": 10000
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.21484375,
-      "learning_rate": 6.658263949432963e-06,
-      "loss": 2.0959,
+      "epoch": 0.93,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018960762755499045,
+      "loss": 2.1776,
       "step": 10005
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.5758561184661174e-06,
-      "loss": 2.0973,
+      "epoch": 0.93,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018958967277414593,
+      "loss": 2.1907,
       "step": 10010
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 6.493952915242085e-06,
-      "loss": 2.0725,
+      "epoch": 0.93,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001895717033481692,
+      "loss": 2.1778,
       "step": 10015
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.21484375,
-      "learning_rate": 6.412554553439587e-06,
-      "loss": 2.1116,
+      "epoch": 0.93,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018955371927999783,
+      "loss": 2.1326,
       "step": 10020
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 6.331661245420306e-06,
-      "loss": 2.1142,
+      "epoch": 0.93,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018953572057257153,
+      "loss": 2.1427,
       "step": 10025
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.251273202228203e-06,
-      "loss": 2.1287,
+      "epoch": 0.93,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018951770722883265,
+      "loss": 2.1292,
       "step": 10030
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 6.171390633589113e-06,
-      "loss": 2.0923,
+      "epoch": 0.93,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001894996792517257,
+      "loss": 2.1777,
       "step": 10035
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 6.0920137479101965e-06,
-      "loss": 2.0858,
+      "epoch": 0.93,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018948163664419778,
+      "loss": 2.1403,
       "step": 10040
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.232421875,
-      "learning_rate": 6.013142752279133e-06,
-      "loss": 2.0913,
+      "epoch": 0.93,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018946357940919826,
+      "loss": 2.1356,
       "step": 10045
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.934777852463946e-06,
-      "loss": 2.1084,
+      "epoch": 0.93,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001894455075496789,
+      "loss": 2.1914,
       "step": 10050
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.856919252912186e-06,
-      "loss": 2.1087,
+      "epoch": 0.93,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018942742106859396,
+      "loss": 2.1993,
       "step": 10055
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.779567156750632e-06,
-      "loss": 2.1063,
+      "epoch": 0.93,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018940931996889998,
+      "loss": 2.1493,
       "step": 10060
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.21484375,
-      "learning_rate": 5.702721765784524e-06,
-      "loss": 2.1162,
+      "epoch": 0.93,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018939120425355595,
+      "loss": 2.1767,
       "step": 10065
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.6263832804971825e-06,
-      "loss": 2.1194,
+      "epoch": 0.93,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001893730739255232,
+      "loss": 2.1229,
       "step": 10070
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.550551900049494e-06,
-      "loss": 2.122,
+      "epoch": 0.93,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018935492898776546,
+      "loss": 2.1528,
       "step": 10075
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 5.475227822279294e-06,
-      "loss": 2.1018,
+      "epoch": 0.93,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018933676944324893,
+      "loss": 2.2357,
       "step": 10080
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21484375,
-      "learning_rate": 5.40041124370092e-06,
-      "loss": 2.1249,
+      "epoch": 0.93,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018931859529494206,
+      "loss": 2.191,
       "step": 10085
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.326102359504725e-06,
-      "loss": 2.1241,
+      "epoch": 0.93,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018930040654581583,
+      "loss": 2.1128,
       "step": 10090
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.220703125,
-      "learning_rate": 5.252301363556455e-06,
-      "loss": 2.1329,
+      "epoch": 0.94,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001892822031988435,
+      "loss": 2.1683,
       "step": 10095
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21484375,
-      "learning_rate": 5.179008448396916e-06,
-      "loss": 2.0565,
+      "epoch": 0.94,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018926398525700075,
+      "loss": 2.1613,
       "step": 10100
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 5.1062238052412835e-06,
-      "loss": 2.1075,
+      "epoch": 0.94,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018924575272326563,
+      "loss": 2.1943,
       "step": 10105
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 5.0339476239787074e-06,
-      "loss": 2.1011,
+      "epoch": 0.94,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018922750560061865,
+      "loss": 2.1928,
       "step": 10110
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.962180093171886e-06,
-      "loss": 2.1014,
+      "epoch": 0.94,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018920924389204257,
+      "loss": 2.1789,
       "step": 10115
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21484375,
-      "learning_rate": 4.890921400056359e-06,
-      "loss": 2.0966,
+      "epoch": 0.94,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018919096760052268,
+      "loss": 2.1427,
       "step": 10120
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 4.82017173054028e-06,
-      "loss": 2.1117,
+      "epoch": 0.94,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018917267672904656,
+      "loss": 2.1471,
       "step": 10125
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.749931269203711e-06,
-      "loss": 2.072,
+      "epoch": 0.94,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018915437128060421,
+      "loss": 2.1646,
       "step": 10130
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.6802001992982234e-06,
-      "loss": 2.0992,
+      "epoch": 0.94,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018913605125818797,
+      "loss": 2.1837,
       "step": 10135
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.610978702746538e-06,
-      "loss": 2.0872,
+      "epoch": 0.94,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018911771666479265,
+      "loss": 2.1659,
       "step": 10140
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.542266960141816e-06,
-      "loss": 2.122,
+      "epoch": 0.94,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018909936750341537,
+      "loss": 2.1576,
       "step": 10145
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.21875,
-      "learning_rate": 4.474065150747375e-06,
-      "loss": 2.1117,
+      "epoch": 0.94,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018908100377705561,
+      "loss": 2.1553,
       "step": 10150
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.406373452496148e-06,
-      "loss": 2.1459,
+      "epoch": 0.94,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001890626254887153,
+      "loss": 2.1423,
       "step": 10155
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 4.339192041990226e-06,
-      "loss": 2.0726,
+      "epoch": 0.94,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018904423264139874,
+      "loss": 2.1431,
       "step": 10160
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.272521094500426e-06,
-      "loss": 2.1388,
+      "epoch": 0.94,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018902582523811257,
+      "loss": 2.179,
       "step": 10165
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.224609375,
-      "learning_rate": 4.206360783965768e-06,
-      "loss": 2.1223,
+      "epoch": 0.94,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001890074032818658,
+      "loss": 2.1906,
       "step": 10170
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.140711282993071e-06,
-      "loss": 2.1004,
+      "epoch": 0.94,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018898896677566987,
+      "loss": 2.1886,
       "step": 10175
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 4.075572762856528e-06,
-      "loss": 2.1162,
+      "epoch": 0.94,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001889705157225386,
+      "loss": 2.201,
       "step": 10180
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.0109453934972005e-06,
-      "loss": 2.0616,
+      "epoch": 0.94,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018895205012548814,
+      "loss": 2.1711,
       "step": 10185
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 3.946829343522551e-06,
-      "loss": 2.0862,
+      "epoch": 0.94,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018893356998753703,
+      "loss": 2.1483,
       "step": 10190
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.883224780206174e-06,
-      "loss": 2.0968,
+      "epoch": 0.94,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018891507531170626,
+      "loss": 2.1506,
       "step": 10195
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.820131869487109e-06,
-      "loss": 2.1148,
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018889656610101904,
+      "loss": 2.2364,
       "step": 10200
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.7575507759696427e-06,
-      "loss": 2.133,
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018887804235850114,
+      "loss": 2.1769,
       "step": 10205
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.6954816629226618e-06,
-      "loss": 2.0871,
+      "epoch": 0.95,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018885950408718057,
+      "loss": 2.2,
       "step": 10210
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 3.6339246922794114e-06,
-      "loss": 2.16,
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018884095129008776,
+      "loss": 2.1838,
       "step": 10215
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.5728800246370265e-06,
-      "loss": 2.1124,
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018882238397025555,
+      "loss": 2.1368,
       "step": 10220
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.512347819256001e-06,
-      "loss": 2.0771,
+      "epoch": 0.95,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001888038021307191,
+      "loss": 2.1381,
       "step": 10225
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 3.452328234059876e-06,
-      "loss": 2.1171,
+      "epoch": 0.95,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000188785205774516,
+      "loss": 2.1322,
       "step": 10230
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.3928214256349067e-06,
-      "loss": 2.1057,
+      "epoch": 0.95,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001887665949046861,
+      "loss": 2.1256,
       "step": 10235
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 3.3338275492293956e-06,
-      "loss": 2.1071,
+      "epoch": 0.95,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018874796952427175,
+      "loss": 2.1061,
       "step": 10240
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.275346758753628e-06,
-      "loss": 2.0921,
+      "epoch": 0.95,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018872932963631765,
+      "loss": 2.1842,
       "step": 10245
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.217379206779114e-06,
-      "loss": 2.1025,
+      "epoch": 0.95,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018871067524387075,
+      "loss": 2.1537,
       "step": 10250
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.224609375,
-      "learning_rate": 3.1599250445384587e-06,
-      "loss": 2.1286,
+      "epoch": 0.95,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018869200634998053,
+      "loss": 2.1058,
       "step": 10255
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.216796875,
-      "learning_rate": 3.102984421924915e-06,
-      "loss": 2.0973,
+      "epoch": 0.95,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018867332295769882,
+      "loss": 2.197,
       "step": 10260
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.21875,
-      "learning_rate": 3.0465574874918524e-06,
-      "loss": 2.1132,
+      "epoch": 0.95,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001886546250700797,
+      "loss": 2.2156,
       "step": 10265
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 2.990644388452535e-06,
-      "loss": 2.1013,
+      "epoch": 0.95,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001886359126901797,
+      "loss": 2.1835,
       "step": 10270
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.935245270679654e-06,
-      "loss": 2.0976,
+      "epoch": 0.95,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018861718582105773,
+      "loss": 2.2087,
       "step": 10275
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.880360278704952e-06,
-      "loss": 2.1008,
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018859844446577508,
+      "loss": 2.1611,
       "step": 10280
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.224609375,
-      "learning_rate": 2.8259895557189332e-06,
-      "loss": 2.1045,
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018857968862739534,
+      "loss": 2.1578,
       "step": 10285
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.22265625,
-      "learning_rate": 2.772133243570285e-06,
-      "loss": 2.0994,
+      "epoch": 0.95,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018856091830898452,
+      "loss": 2.1668,
       "step": 10290
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.7187914827657702e-06,
-      "loss": 2.0822,
+      "epoch": 0.95,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000188542133513611,
+      "loss": 2.169,
       "step": 10295
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.6659644124696235e-06,
-      "loss": 2.1504,
+      "epoch": 0.95,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018852333424434548,
+      "loss": 2.1701,
       "step": 10300
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.6136521705034e-06,
-      "loss": 2.0762,
+      "epoch": 0.95,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001885045205042611,
+      "loss": 2.167,
       "step": 10305
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.220703125,
-      "learning_rate": 2.5618548933453944e-06,
-      "loss": 2.1154,
+      "epoch": 0.96,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018848569229643327,
+      "loss": 2.1178,
       "step": 10310
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 2.5105727161304883e-06,
-      "loss": 2.1195,
+      "epoch": 0.96,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018846684962393987,
+      "loss": 2.2225,
       "step": 10315
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.4598057726496814e-06,
-      "loss": 2.0759,
+      "epoch": 0.96,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018844799248986105,
+      "loss": 2.1181,
       "step": 10320
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 2.4095541953498057e-06,
-      "loss": 2.1001,
+      "epoch": 0.96,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018842912089727938,
+      "loss": 2.1523,
       "step": 10325
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.232421875,
-      "learning_rate": 2.3598181153330123e-06,
-      "loss": 2.1505,
+      "epoch": 0.96,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018841023484927977,
+      "loss": 2.1677,
       "step": 10330
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.22265625,
-      "learning_rate": 2.3105976623567505e-06,
-      "loss": 2.1319,
+      "epoch": 0.96,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001883913343489496,
+      "loss": 2.195,
       "step": 10335
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.261892964833101e-06,
-      "loss": 2.1112,
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018837241939937834,
+      "loss": 2.1617,
       "step": 10340
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.2137041498286436e-06,
-      "loss": 2.1086,
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001883534900036581,
+      "loss": 2.0657,
       "step": 10345
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.166031343064057e-06,
-      "loss": 2.0987,
+      "epoch": 0.96,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018833454616488327,
+      "loss": 2.2112,
       "step": 10350
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.1188746689137395e-06,
-      "loss": 2.0885,
+      "epoch": 0.96,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018831558788615053,
+      "loss": 2.1563,
       "step": 10355
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 2.072234250405658e-06,
-      "loss": 2.0984,
-      "step": 10360
+      "epoch": 0.96,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.000188296615170559,
+      "loss": 2.1722,
+      "step": 10360
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.026110209220833e-06,
-      "loss": 2.1191,
+      "epoch": 0.96,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018827762802121014,
+      "loss": 2.1478,
       "step": 10365
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.9805026656930514e-06,
-      "loss": 2.1586,
+      "epoch": 0.96,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001882586264412077,
+      "loss": 2.1554,
       "step": 10370
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.9354117388087125e-06,
-      "loss": 2.0988,
+      "epoch": 0.96,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018823961043365792,
+      "loss": 2.1808,
       "step": 10375
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.8908375462063144e-06,
-      "loss": 2.0778,
+      "epoch": 0.96,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001882205800016693,
+      "loss": 2.1378,
       "step": 10380
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.846780204176324e-06,
-      "loss": 2.1076,
+      "epoch": 0.96,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018820153514835267,
+      "loss": 2.1922,
       "step": 10385
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.8032398276606854e-06,
-      "loss": 2.1024,
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018818247587682138,
+      "loss": 2.1885,
       "step": 10390
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.7602165302526673e-06,
-      "loss": 2.1098,
+      "epoch": 0.96,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018816340219019096,
+      "loss": 2.1749,
       "step": 10395
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.7177104241965946e-06,
-      "loss": 2.0772,
+      "epoch": 0.96,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018814431409157938,
+      "loss": 2.16,
       "step": 10400
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.6757216203873382e-06,
-      "loss": 2.0987,
+      "epoch": 0.96,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018812521158410693,
+      "loss": 2.1257,
       "step": 10405
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.6342502283702932e-06,
-      "loss": 2.0965,
+      "epoch": 0.96,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018810609467089632,
+      "loss": 2.1181,
       "step": 10410
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.5932963563409341e-06,
-      "loss": 2.0916,
+      "epoch": 0.97,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018808696335507254,
+      "loss": 2.1689,
       "step": 10415
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.5528601111445051e-06,
-      "loss": 2.1307,
+      "epoch": 0.97,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018806781763976298,
+      "loss": 2.135,
       "step": 10420
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.22265625,
-      "learning_rate": 1.5129415982758855e-06,
-      "loss": 2.1692,
+      "epoch": 0.97,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018804865752809739,
+      "loss": 2.214,
       "step": 10425
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.473540921879235e-06,
-      "loss": 2.1008,
+      "epoch": 0.97,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001880294830232078,
+      "loss": 2.1882,
       "step": 10430
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.434658184747617e-06,
-      "loss": 2.1081,
+      "epoch": 0.97,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018801029412822867,
+      "loss": 2.1506,
       "step": 10435
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.3962934883229751e-06,
-      "loss": 2.1117,
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018799109084629683,
+      "loss": 2.2038,
       "step": 10440
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.3584469326956228e-06,
-      "loss": 2.1159,
+      "epoch": 0.97,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018797187318055135,
+      "loss": 2.1683,
       "step": 10445
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.3211186166041557e-06,
-      "loss": 2.0878,
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001879526411341338,
+      "loss": 2.1636,
       "step": 10450
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.28430863743505e-06,
-      "loss": 2.1206,
+      "epoch": 0.97,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018793339471018795,
+      "loss": 2.2002,
       "step": 10455
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 1.248017091222553e-06,
-      "loss": 2.1235,
+      "epoch": 0.97,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018791413391186005,
+      "loss": 2.2274,
       "step": 10460
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.2122440726483275e-06,
-      "loss": 2.128,
+      "epoch": 0.97,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018789485874229865,
+      "loss": 2.1956,
       "step": 10465
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.1769896750412512e-06,
-      "loss": 2.1555,
+      "epoch": 0.97,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001878755692046546,
+      "loss": 2.1602,
       "step": 10470
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.142253990377129e-06,
-      "loss": 2.1062,
+      "epoch": 0.97,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018785626530208108,
+      "loss": 2.1358,
       "step": 10475
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.1080371092785591e-06,
-      "loss": 2.0768,
+      "epoch": 0.97,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018783694703773383,
+      "loss": 2.1721,
       "step": 10480
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.074339121014556e-06,
-      "loss": 2.1237,
+      "epoch": 0.97,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018781761441477067,
+      "loss": 2.1877,
       "step": 10485
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.220703125,
-      "learning_rate": 1.041160113500439e-06,
-      "loss": 2.1123,
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018779826743635196,
+      "loss": 2.1652,
       "step": 10490
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.216796875,
-      "learning_rate": 1.0085001732974774e-06,
-      "loss": 2.115,
+      "epoch": 0.97,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018777890610564027,
+      "loss": 2.1497,
       "step": 10495
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.763593856128017e-07,
-      "loss": 2.1598,
+      "epoch": 0.97,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001877595304258006,
+      "loss": 2.1524,
       "step": 10500
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 9.447378342991142e-07,
-      "loss": 2.1414,
+      "epoch": 0.97,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018774014040000026,
+      "loss": 2.1653,
       "step": 10505
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 9.136356018544012e-07,
-      "loss": 2.1251,
+      "epoch": 0.97,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018772073603140897,
+      "loss": 2.1582,
       "step": 10510
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 8.830527694218882e-07,
-      "loss": 2.0766,
+      "epoch": 0.97,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001877013173231987,
+      "loss": 2.1726,
       "step": 10515
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.216796875,
-      "learning_rate": 8.529894167896402e-07,
-      "loss": 2.0971,
+      "epoch": 0.97,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018768188427854376,
+      "loss": 2.1212,
       "step": 10520
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 8.234456223904951e-07,
-      "loss": 2.0995,
+      "epoch": 0.98,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018766243690062094,
+      "loss": 2.1657,
       "step": 10525
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 7.944214633018199e-07,
-      "loss": 2.1257,
+      "epoch": 0.98,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018764297519260925,
+      "loss": 2.1701,
       "step": 10530
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 7.659170152452211e-07,
-      "loss": 2.1156,
+      "epoch": 0.98,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018762349915769005,
+      "loss": 2.1244,
       "step": 10535
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 7.379323525865012e-07,
-      "loss": 2.1039,
+      "epoch": 0.98,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018760400879904706,
+      "loss": 2.1138,
       "step": 10540
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 7.104675483353474e-07,
-      "loss": 2.1651,
+      "epoch": 0.98,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001875845041198664,
+      "loss": 2.1972,
       "step": 10545
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.835226741451539e-07,
-      "loss": 2.0815,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018756498512333648,
+      "loss": 2.1373,
       "step": 10550
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.220703125,
-      "learning_rate": 6.570978003129336e-07,
-      "loss": 2.0955,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000187545451812648,
+      "loss": 2.1944,
       "step": 10555
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.216796875,
-      "learning_rate": 6.311929957789842e-07,
-      "loss": 2.1086,
+      "epoch": 0.98,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018752590419099407,
+      "loss": 2.2138,
       "step": 10560
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 6.058083281268223e-07,
-      "loss": 2.1192,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018750634226157015,
+      "loss": 2.1879,
       "step": 10565
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.21875,
-      "learning_rate": 5.809438635829833e-07,
-      "loss": 2.0822,
+      "epoch": 0.98,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018748676602757395,
+      "loss": 2.1863,
       "step": 10570
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.216796875,
-      "learning_rate": 5.565996670168216e-07,
-      "loss": 2.1074,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018746717549220565,
+      "loss": 2.1586,
       "step": 10575
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2255859375,
-      "learning_rate": 5.327758019403106e-07,
-      "loss": 2.0805,
+      "epoch": 0.98,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018744757065866763,
+      "loss": 2.1461,
       "step": 10580
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.212890625,
-      "learning_rate": 5.094723305079984e-07,
-      "loss": 2.0862,
+      "epoch": 0.98,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001874279515301647,
+      "loss": 2.1519,
       "step": 10585
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 4.866893135167416e-07,
-      "loss": 2.1187,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000187408318109904,
+      "loss": 2.2064,
       "step": 10590
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.216796875,
-      "learning_rate": 4.6442681040552714e-07,
-      "loss": 2.0898,
+      "epoch": 0.98,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018738867040109494,
+      "loss": 2.1868,
       "step": 10595
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 4.4268487925542835e-07,
-      "loss": 2.0712,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018736900840694932,
+      "loss": 2.1771,
       "step": 10600
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 4.2146357678936045e-07,
-      "loss": 2.0944,
+      "epoch": 0.98,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018734933213068129,
+      "loss": 2.1875,
       "step": 10605
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.212890625,
-      "learning_rate": 4.0076295837194745e-07,
-      "loss": 2.0915,
+      "epoch": 0.98,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001873296415755073,
+      "loss": 2.1804,
       "step": 10610
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.805830780094333e-07,
-      "loss": 2.102,
+      "epoch": 0.98,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018730993674464613,
+      "loss": 2.1724,
       "step": 10615
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.220703125,
-      "learning_rate": 3.609239883494375e-07,
-      "loss": 2.163,
+      "epoch": 0.98,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018729021764131895,
+      "loss": 2.1705,
       "step": 10620
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 3.417857406809333e-07,
-      "loss": 2.1239,
+      "epoch": 0.98,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001872704842687492,
+      "loss": 2.1661,
       "step": 10625
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 3.2316838493404723e-07,
-      "loss": 2.1066,
+      "epoch": 0.98,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018725073663016263,
+      "loss": 2.1387,
       "step": 10630
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.22265625,
-      "learning_rate": 3.0507196967992644e-07,
-      "loss": 2.1214,
+      "epoch": 0.99,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018723097472878744,
+      "loss": 2.186,
       "step": 10635
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 2.8749654213062747e-07,
-      "loss": 2.0529,
-      "step": 10640
-    },
+      "epoch": 0.99,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018721119856785403,
+      "loss": 2.1647,
+      "step": 10640
+    },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.22265625,
-      "learning_rate": 2.7044214813902737e-07,
-      "loss": 2.1039,
+      "epoch": 0.99,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001871914081505952,
+      "loss": 2.1654,
       "step": 10645
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.216796875,
-      "learning_rate": 2.539088321986016e-07,
-      "loss": 2.1145,
+      "epoch": 0.99,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001871716034802461,
+      "loss": 2.185,
       "step": 10650
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.21875,
-      "learning_rate": 2.3789663744344659e-07,
-      "loss": 2.1126,
+      "epoch": 0.99,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001871517845600442,
+      "loss": 2.1553,
       "step": 10655
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 2.2240560564807945e-07,
-      "loss": 2.1045,
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001871319513932292,
+      "loss": 2.0996,
       "step": 10660
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 2.0743577722730501e-07,
-      "loss": 2.0901,
+      "epoch": 0.99,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018711210398304325,
+      "loss": 2.1564,
       "step": 10665
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 1.9298719123617138e-07,
-      "loss": 2.1305,
+      "epoch": 0.99,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018709224233273076,
+      "loss": 2.1999,
       "step": 10670
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.7905988536988105e-07,
-      "loss": 2.0512,
+      "epoch": 0.99,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018707236644553853,
+      "loss": 2.1594,
       "step": 10675
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 1.6565389596363555e-07,
-      "loss": 2.1194,
+      "epoch": 0.99,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018705247632471564,
+      "loss": 2.1539,
       "step": 10680
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 1.527692579925466e-07,
-      "loss": 2.0744,
+      "epoch": 0.99,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018703257197351348,
+      "loss": 2.2113,
       "step": 10685
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 1.4040600507156943e-07,
-      "loss": 2.052,
+      "epoch": 0.99,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018701265339518588,
+      "loss": 2.1364,
       "step": 10690
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.212890625,
-      "learning_rate": 1.2856416945545847e-07,
-      "loss": 2.1111,
+      "epoch": 0.99,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001869927205929888,
+      "loss": 2.1848,
       "step": 10695
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.232421875,
-      "learning_rate": 1.1724378203854524e-07,
-      "loss": 2.1183,
+      "epoch": 0.99,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001869727735701807,
+      "loss": 2.0954,
       "step": 10700
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.21875,
-      "learning_rate": 1.0644487235482725e-07,
-      "loss": 2.1012,
+      "epoch": 0.99,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018695281233002225,
+      "loss": 2.1327,
       "step": 10705
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 9.616746857774583e-08,
-      "loss": 2.1225,
+      "epoch": 0.99,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018693283687577656,
+      "loss": 2.1825,
       "step": 10710
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.22265625,
-      "learning_rate": 8.641159752023064e-08,
-      "loss": 2.1564,
+      "epoch": 0.99,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018691284721070897,
+      "loss": 2.1999,
       "step": 10715
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.228515625,
-      "learning_rate": 7.717728463454421e-08,
-      "loss": 2.0847,
+      "epoch": 0.99,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001868928433380871,
+      "loss": 2.1318,
       "step": 10720
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.220703125,
-      "learning_rate": 6.846455401225971e-08,
-      "loss": 2.0842,
+      "epoch": 0.99,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001868728252611811,
+      "loss": 2.1794,
       "step": 10725
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.21875,
-      "learning_rate": 6.027342838414996e-08,
-      "loss": 2.1097,
+      "epoch": 0.99,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018685279298326322,
+      "loss": 2.1712,
       "step": 10730
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 5.260392912020962e-08,
-      "loss": 2.0956,
+      "epoch": 0.99,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018683274650760807,
+      "loss": 2.0467,
       "step": 10735
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.21484375,
-      "learning_rate": 4.545607622954417e-08,
-      "loss": 2.0895,
+      "epoch": 1.0,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018681268583749274,
+      "loss": 2.1904,
       "step": 10740
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 3.882988836028112e-08,
-      "loss": 2.0985,
+      "epoch": 1.0,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018679261097619644,
+      "loss": 2.1162,
       "step": 10745
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 3.272538279963655e-08,
-      "loss": 2.0959,
+      "epoch": 1.0,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018677252192700084,
+      "loss": 2.1819,
       "step": 10750
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.212890625,
-      "learning_rate": 2.714257547375976e-08,
-      "loss": 2.1242,
+      "epoch": 1.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018675241869318984,
+      "loss": 2.2116,
       "step": 10755
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.228515625,
-      "learning_rate": 2.2081480947711007e-08,
-      "loss": 2.0885,
+      "epoch": 1.0,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001867323012780497,
+      "loss": 2.1122,
       "step": 10760
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.21484375,
-      "learning_rate": 1.7542112425505964e-08,
-      "loss": 2.0554,
+      "epoch": 1.0,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000186712169684869,
+      "loss": 2.1676,
       "step": 10765
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.3524481749982442e-08,
-      "loss": 2.1695,
+      "epoch": 1.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018669202391693865,
+      "loss": 2.1844,
       "step": 10770
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 1.0028599402778226e-08,
-      "loss": 2.1069,
+      "epoch": 1.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018667186397755182,
+      "loss": 2.1499,
       "step": 10775
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 7.054474504375463e-09,
-      "loss": 2.1023,
+      "epoch": 1.0,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018665168987000406,
+      "loss": 2.2178,
       "step": 10780
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 4.602114814056258e-09,
-      "loss": 2.0682,
+      "epoch": 1.0,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018663150159759325,
+      "loss": 2.1679,
       "step": 10785
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.21484375,
-      "learning_rate": 2.671526729769447e-09,
-      "loss": 2.0856,
+      "epoch": 1.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018661129916361943,
+      "loss": 2.1066,
       "step": 10790
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.224609375,
-      "learning_rate": 1.262715288308236e-09,
-      "loss": 2.1324,
+      "epoch": 1.0,
+      "eval_loss": 2.1613714694976807,
+      "eval_runtime": 188.9611,
+      "eval_samples_per_second": 25.693,
+      "eval_steps_per_second": 3.212,
+      "step": 10792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018659108257138517,
+      "loss": 2.1337,
       "step": 10795
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.2138671875,
-      "learning_rate": 3.7568416513256157e-10,
-      "loss": 2.0969,
+      "epoch": 1.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018657085182419524,
+      "loss": 2.1421,
       "step": 10800
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.0435674435704812e-11,
-      "loss": 2.1084,
+      "epoch": 1.0,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018655060692535674,
+      "loss": 2.1293,
       "step": 10805
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 2.1439597606658936,
-      "eval_runtime": 190.2812,
-      "eval_samples_per_second": 25.557,
-      "eval_steps_per_second": 3.195,
-      "step": 10806
+      "epoch": 1.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018653034787817905,
+      "loss": 2.1485,
+      "step": 10810
     },
     {
-      "epoch": 2.0,
-      "step": 10806,
-      "total_flos": 6.137045068818678e+17,
-      "train_loss": 2.1470370752908954,
-      "train_runtime": 23954.8129,
-      "train_samples_per_second": 7.217,
-      "train_steps_per_second": 0.451
+      "epoch": 1.0,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018651007468597393,
+      "loss": 2.1972,
+      "step": 10815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018648978735205538,
+      "loss": 2.1058,
+      "step": 10820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001864694858797398,
+      "loss": 2.0854,
+      "step": 10825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018644917027234583,
+      "loss": 2.1178,
+      "step": 10830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018642884053319444,
+      "loss": 2.1343,
+      "step": 10835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018640849666560886,
+      "loss": 2.1163,
+      "step": 10840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001863881386729148,
+      "loss": 2.1463,
+      "step": 10845
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018636776655844004,
+      "loss": 2.1539,
+      "step": 10850
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00018634738032551485,
+      "loss": 2.1502,
+      "step": 10855
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018632697997747176,
+      "loss": 2.1207,
+      "step": 10860
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018630656551764553,
+      "loss": 2.1534,
+      "step": 10865
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018628613694937336,
+      "loss": 2.1675,
+      "step": 10870
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018626569427599465,
+      "loss": 2.1854,
+      "step": 10875
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018624523750085116,
+      "loss": 2.2078,
+      "step": 10880
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001862247666272869,
+      "loss": 2.1227,
+      "step": 10885
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001862042816586483,
+      "loss": 2.1245,
+      "step": 10890
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018618378259828392,
+      "loss": 2.1823,
+      "step": 10895
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001861632694495448,
+      "loss": 2.1337,
+      "step": 10900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018614274221578415,
+      "loss": 2.1825,
+      "step": 10905
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0001861222009003576,
+      "loss": 2.1499,
+      "step": 10910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001861016455066229,
+      "loss": 2.0932,
+      "step": 10915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018608107603794037,
+      "loss": 2.1477,
+      "step": 10920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018606049249767235,
+      "loss": 2.1347,
+      "step": 10925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001860398948891837,
+      "loss": 2.0788,
+      "step": 10930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018601928321584148,
+      "loss": 2.1276,
+      "step": 10935
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000185998657481015,
+      "loss": 2.1491,
+      "step": 10940
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018597801768807597,
+      "loss": 2.0936,
+      "step": 10945
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018595736384039832,
+      "loss": 2.1411,
+      "step": 10950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018593669594135838,
+      "loss": 2.1152,
+      "step": 10955
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018591601399433464,
+      "loss": 2.1108,
+      "step": 10960
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000185895318002708,
+      "loss": 2.2015,
+      "step": 10965
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018587460796986162,
+      "loss": 2.0554,
+      "step": 10970
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018585388389918094,
+      "loss": 2.1145,
+      "step": 10975
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018583314579405364,
+      "loss": 2.149,
+      "step": 10980
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018581239365786985,
+      "loss": 2.1613,
+      "step": 10985
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018579162749402184,
+      "loss": 2.2019,
+      "step": 10990
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018577084730590426,
+      "loss": 2.1336,
+      "step": 10995
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018575005309691403,
+      "loss": 2.1028,
+      "step": 11000
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018572924487045032,
+      "loss": 2.1211,
+      "step": 11005
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018570842262991466,
+      "loss": 2.1112,
+      "step": 11010
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018568758637871086,
+      "loss": 2.1078,
+      "step": 11015
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018566673612024497,
+      "loss": 2.1176,
+      "step": 11020
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018564587185792534,
+      "loss": 2.2093,
+      "step": 11025
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001856249935951627,
+      "loss": 2.1134,
+      "step": 11030
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018560410133536994,
+      "loss": 2.1057,
+      "step": 11035
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001855831950819623,
+      "loss": 2.1191,
+      "step": 11040
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018556227483835732,
+      "loss": 2.1232,
+      "step": 11045
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001855413406079748,
+      "loss": 2.1591,
+      "step": 11050
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018552039239423687,
+      "loss": 2.1054,
+      "step": 11055
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018549943020056787,
+      "loss": 2.1155,
+      "step": 11060
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001854784540303945,
+      "loss": 2.1343,
+      "step": 11065
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018545746388714567,
+      "loss": 2.1252,
+      "step": 11070
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018543645977425266,
+      "loss": 2.0791,
+      "step": 11075
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018541544169514896,
+      "loss": 2.0978,
+      "step": 11080
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018539440965327042,
+      "loss": 2.1392,
+      "step": 11085
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018537336365205507,
+      "loss": 2.1371,
+      "step": 11090
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001853523036949433,
+      "loss": 2.1723,
+      "step": 11095
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001853312297853778,
+      "loss": 2.1724,
+      "step": 11100
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018531014192680345,
+      "loss": 2.1597,
+      "step": 11105
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018528904012266747,
+      "loss": 2.1507,
+      "step": 11110
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018526792437641935,
+      "loss": 2.1577,
+      "step": 11115
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001852467946915109,
+      "loss": 2.1869,
+      "step": 11120
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001852256510713961,
+      "loss": 2.1202,
+      "step": 11125
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001852044935195313,
+      "loss": 2.1508,
+      "step": 11130
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018518332203937515,
+      "loss": 2.1125,
+      "step": 11135
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018516213663438848,
+      "loss": 2.1217,
+      "step": 11140
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018514093730803444,
+      "loss": 2.167,
+      "step": 11145
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001851197240637785,
+      "loss": 2.1412,
+      "step": 11150
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018509849690508836,
+      "loss": 2.1219,
+      "step": 11155
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018507725583543396,
+      "loss": 2.1057,
+      "step": 11160
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001850560008582876,
+      "loss": 2.1151,
+      "step": 11165
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001850347319771238,
+      "loss": 2.1533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018501344919541937,
+      "loss": 2.1219,
+      "step": 11175
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018499215251665334,
+      "loss": 2.1142,
+      "step": 11180
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00018497084194430712,
+      "loss": 2.1045,
+      "step": 11185
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018494951748186428,
+      "loss": 2.148,
+      "step": 11190
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018492817913281072,
+      "loss": 2.1301,
+      "step": 11195
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018490682690063462,
+      "loss": 2.1449,
+      "step": 11200
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001848854607888264,
+      "loss": 2.1128,
+      "step": 11205
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001848640808008787,
+      "loss": 2.0981,
+      "step": 11210
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018484268694028656,
+      "loss": 2.1248,
+      "step": 11215
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001848212792105472,
+      "loss": 2.1444,
+      "step": 11220
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018479985761516007,
+      "loss": 2.0962,
+      "step": 11225
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018477842215762702,
+      "loss": 2.1355,
+      "step": 11230
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018475697284145199,
+      "loss": 2.1396,
+      "step": 11235
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018473550967014135,
+      "loss": 2.1447,
+      "step": 11240
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018471403264720362,
+      "loss": 2.1467,
+      "step": 11245
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018469254177614963,
+      "loss": 2.1423,
+      "step": 11250
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018467103706049247,
+      "loss": 2.1484,
+      "step": 11255
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018464951850374753,
+      "loss": 2.1268,
+      "step": 11260
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001846279861094324,
+      "loss": 2.1442,
+      "step": 11265
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018460643988106693,
+      "loss": 2.1114,
+      "step": 11270
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018458487982217332,
+      "loss": 2.0875,
+      "step": 11275
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001845633059362759,
+      "loss": 2.1328,
+      "step": 11280
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018454171822690144,
+      "loss": 2.147,
+      "step": 11285
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018452011669757874,
+      "loss": 2.1435,
+      "step": 11290
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018449850135183904,
+      "loss": 2.2069,
+      "step": 11295
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018447687219321578,
+      "loss": 2.1452,
+      "step": 11300
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018445522922524464,
+      "loss": 2.1895,
+      "step": 11305
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001844335724514636,
+      "loss": 2.2201,
+      "step": 11310
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001844119018754128,
+      "loss": 2.1766,
+      "step": 11315
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018439021750063481,
+      "loss": 2.0766,
+      "step": 11320
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018436851933067427,
+      "loss": 2.1345,
+      "step": 11325
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018434680736907822,
+      "loss": 2.1827,
+      "step": 11330
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001843250816193958,
+      "loss": 2.1244,
+      "step": 11335
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001843033420851786,
+      "loss": 2.164,
+      "step": 11340
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018428158876998034,
+      "loss": 2.1396,
+      "step": 11345
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018425982167735695,
+      "loss": 2.1913,
+      "step": 11350
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018423804081086674,
+      "loss": 2.1024,
+      "step": 11355
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018421624617407013,
+      "loss": 2.1285,
+      "step": 11360
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018419443777052997,
+      "loss": 2.1737,
+      "step": 11365
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018417261560381117,
+      "loss": 2.0931,
+      "step": 11370
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018415077967748103,
+      "loss": 2.084,
+      "step": 11375
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000184128929995109,
+      "loss": 2.1322,
+      "step": 11380
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018410706656026688,
+      "loss": 2.0973,
+      "step": 11385
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001840851893765286,
+      "loss": 2.1353,
+      "step": 11390
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018406329844747048,
+      "loss": 2.1368,
+      "step": 11395
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018404139377667093,
+      "loss": 2.1345,
+      "step": 11400
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018401947536771075,
+      "loss": 2.1681,
+      "step": 11405
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018399754322417287,
+      "loss": 2.133,
+      "step": 11410
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001839755973496425,
+      "loss": 2.1621,
+      "step": 11415
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018395363774770717,
+      "loss": 2.1612,
+      "step": 11420
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001839316644219566,
+      "loss": 2.1268,
+      "step": 11425
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018390967737598265,
+      "loss": 2.1417,
+      "step": 11430
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018388767661337965,
+      "loss": 2.1388,
+      "step": 11435
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018386566213774393,
+      "loss": 2.1582,
+      "step": 11440
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018384363395267423,
+      "loss": 2.126,
+      "step": 11445
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018382159206177146,
+      "loss": 2.1258,
+      "step": 11450
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001837995364686388,
+      "loss": 2.144,
+      "step": 11455
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001837774671768816,
+      "loss": 2.156,
+      "step": 11460
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018375538419010762,
+      "loss": 2.1353,
+      "step": 11465
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018373328751192664,
+      "loss": 2.1234,
+      "step": 11470
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001837111771459508,
+      "loss": 2.1882,
+      "step": 11475
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001836890530957945,
+      "loss": 2.1204,
+      "step": 11480
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001836669153650743,
+      "loss": 2.141,
+      "step": 11485
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018364476395740903,
+      "loss": 2.0787,
+      "step": 11490
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018362259887641977,
+      "loss": 2.1118,
+      "step": 11495
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018360042012572982,
+      "loss": 2.1432,
+      "step": 11500
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018357822770896474,
+      "loss": 2.0937,
+      "step": 11505
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018355602162975225,
+      "loss": 2.1346,
+      "step": 11510
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001835338018917224,
+      "loss": 2.2138,
+      "step": 11515
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001835115684985074,
+      "loss": 2.1245,
+      "step": 11520
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018348932145374178,
+      "loss": 2.148,
+      "step": 11525
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018346706076106215,
+      "loss": 2.0758,
+      "step": 11530
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018344478642410753,
+      "loss": 2.1439,
+      "step": 11535
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000183422498446519,
+      "loss": 2.105,
+      "step": 11540
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018340019683194,
+      "loss": 2.1229,
+      "step": 11545
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018337788158401616,
+      "loss": 2.121,
+      "step": 11550
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001833555527063953,
+      "loss": 2.1125,
+      "step": 11555
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018333321020272755,
+      "loss": 2.1121,
+      "step": 11560
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018331085407666516,
+      "loss": 2.1396,
+      "step": 11565
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018328848433186269,
+      "loss": 2.1775,
+      "step": 11570
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001832661009719769,
+      "loss": 2.1575,
+      "step": 11575
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018324370400066674,
+      "loss": 2.2056,
+      "step": 11580
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018322129342159346,
+      "loss": 2.1579,
+      "step": 11585
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001831988692384205,
+      "loss": 2.1561,
+      "step": 11590
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018317643145481348,
+      "loss": 2.1538,
+      "step": 11595
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018315398007444033,
+      "loss": 2.1226,
+      "step": 11600
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018313151510097113,
+      "loss": 2.136,
+      "step": 11605
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018310903653807818,
+      "loss": 2.1365,
+      "step": 11610
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018308654438943607,
+      "loss": 2.1229,
+      "step": 11615
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018306403865872155,
+      "loss": 2.164,
+      "step": 11620
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001830415193496136,
+      "loss": 2.1372,
+      "step": 11625
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018301898646579345,
+      "loss": 2.1012,
+      "step": 11630
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001829964400109445,
+      "loss": 2.185,
+      "step": 11635
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018297387998875242,
+      "loss": 2.1577,
+      "step": 11640
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018295130640290505,
+      "loss": 2.1756,
+      "step": 11645
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001829287192570925,
+      "loss": 2.1197,
+      "step": 11650
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018290611855500705,
+      "loss": 2.1684,
+      "step": 11655
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018288350430034322,
+      "loss": 2.1695,
+      "step": 11660
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018286087649679772,
+      "loss": 2.1145,
+      "step": 11665
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018283823514806953,
+      "loss": 2.1792,
+      "step": 11670
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001828155802578598,
+      "loss": 2.1446,
+      "step": 11675
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018279291182987187,
+      "loss": 2.0979,
+      "step": 11680
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018277022986781135,
+      "loss": 2.1098,
+      "step": 11685
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000182747534375386,
+      "loss": 2.1955,
+      "step": 11690
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018272482535630592,
+      "loss": 2.1735,
+      "step": 11695
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001827021028142832,
+      "loss": 2.1248,
+      "step": 11700
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001826793667530324,
+      "loss": 2.1537,
+      "step": 11705
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018265661717627009,
+      "loss": 2.1406,
+      "step": 11710
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001826338540877151,
+      "loss": 2.1414,
+      "step": 11715
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018261107749108856,
+      "loss": 2.1715,
+      "step": 11720
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018258828739011365,
+      "loss": 2.1087,
+      "step": 11725
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018256548378851593,
+      "loss": 2.141,
+      "step": 11730
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018254266669002304,
+      "loss": 2.1186,
+      "step": 11735
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018251983609836488,
+      "loss": 2.186,
+      "step": 11740
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018249699201727354,
+      "loss": 2.078,
+      "step": 11745
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018247413445048326,
+      "loss": 2.161,
+      "step": 11750
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018245126340173066,
+      "loss": 2.1544,
+      "step": 11755
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018242837887475435,
+      "loss": 2.1363,
+      "step": 11760
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018240548087329527,
+      "loss": 2.1776,
+      "step": 11765
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018238256940109656,
+      "loss": 2.1611,
+      "step": 11770
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001823596444619035,
+      "loss": 2.1019,
+      "step": 11775
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018233670605946362,
+      "loss": 2.134,
+      "step": 11780
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018231375419752662,
+      "loss": 2.1057,
+      "step": 11785
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018229078887984444,
+      "loss": 2.1672,
+      "step": 11790
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001822678101101712,
+      "loss": 2.1061,
+      "step": 11795
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001822448178922632,
+      "loss": 2.1281,
+      "step": 11800
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018222181222987897,
+      "loss": 2.1984,
+      "step": 11805
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018219879312677918,
+      "loss": 2.1438,
+      "step": 11810
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001821757605867268,
+      "loss": 2.1819,
+      "step": 11815
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001821527146134869,
+      "loss": 2.1518,
+      "step": 11820
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001821296552108268,
+      "loss": 2.1334,
+      "step": 11825
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018210658238251596,
+      "loss": 2.1109,
+      "step": 11830
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018208349613232612,
+      "loss": 2.1639,
+      "step": 11835
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001820603964640311,
+      "loss": 2.1893,
+      "step": 11840
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018203728338140706,
+      "loss": 2.0978,
+      "step": 11845
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018201415688823223,
+      "loss": 2.1263,
+      "step": 11850
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018199101698828708,
+      "loss": 2.1173,
+      "step": 11855
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018196786368535425,
+      "loss": 2.1246,
+      "step": 11860
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001819446969832186,
+      "loss": 2.154,
+      "step": 11865
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001819215168856672,
+      "loss": 2.1115,
+      "step": 11870
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018189832339648922,
+      "loss": 2.1348,
+      "step": 11875
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018187511651947608,
+      "loss": 2.1724,
+      "step": 11880
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018185189625842145,
+      "loss": 2.1017,
+      "step": 11885
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018182866261712107,
+      "loss": 2.1784,
+      "step": 11890
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001818054155993729,
+      "loss": 2.1172,
+      "step": 11895
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018178215520897712,
+      "loss": 2.0791,
+      "step": 11900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018175888144973613,
+      "loss": 2.1466,
+      "step": 11905
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018173559432545444,
+      "loss": 2.1439,
+      "step": 11910
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018171229383993873,
+      "loss": 2.1362,
+      "step": 11915
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018168897999699797,
+      "loss": 2.1405,
+      "step": 11920
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018166565280044322,
+      "loss": 2.0947,
+      "step": 11925
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018164231225408775,
+      "loss": 2.126,
+      "step": 11930
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000181618958361747,
+      "loss": 2.1491,
+      "step": 11935
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018159559112723863,
+      "loss": 2.1182,
+      "step": 11940
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018157221055438248,
+      "loss": 2.1087,
+      "step": 11945
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018154881664700053,
+      "loss": 2.1287,
+      "step": 11950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001815254094089169,
+      "loss": 2.1345,
+      "step": 11955
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018150198884395803,
+      "loss": 2.1409,
+      "step": 11960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018147855495595242,
+      "loss": 2.1995,
+      "step": 11965
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001814551077487308,
+      "loss": 2.1404,
+      "step": 11970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018143164722612604,
+      "loss": 2.116,
+      "step": 11975
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018140817339197323,
+      "loss": 2.1229,
+      "step": 11980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001813846862501096,
+      "loss": 2.1141,
+      "step": 11985
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018136118580437455,
+      "loss": 2.1465,
+      "step": 11990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001813376720586097,
+      "loss": 2.2145,
+      "step": 11995
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001813141450166588,
+      "loss": 2.1534,
+      "step": 12000
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001812906046823678,
+      "loss": 2.1351,
+      "step": 12005
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018126705105958485,
+      "loss": 2.0887,
+      "step": 12010
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001812434841521602,
+      "loss": 2.1887,
+      "step": 12015
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018121990396394627,
+      "loss": 2.1065,
+      "step": 12020
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00018119631049879774,
+      "loss": 2.1853,
+      "step": 12025
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018117270376057145,
+      "loss": 2.1502,
+      "step": 12030
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018114908375312626,
+      "loss": 2.1338,
+      "step": 12035
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001811254504803234,
+      "loss": 2.1046,
+      "step": 12040
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018110180394602615,
+      "loss": 2.154,
+      "step": 12045
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018107814415409997,
+      "loss": 2.1085,
+      "step": 12050
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018105447110841253,
+      "loss": 2.1688,
+      "step": 12055
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001810307848128336,
+      "loss": 2.1498,
+      "step": 12060
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018100708527123518,
+      "loss": 2.0944,
+      "step": 12065
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001809833724874914,
+      "loss": 2.1339,
+      "step": 12070
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018095964646547857,
+      "loss": 2.1424,
+      "step": 12075
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00018093590720907516,
+      "loss": 2.1253,
+      "step": 12080
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018091215472216182,
+      "loss": 2.1035,
+      "step": 12085
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018088838900862128,
+      "loss": 2.1017,
+      "step": 12090
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018086461007233858,
+      "loss": 2.1564,
+      "step": 12095
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018084081791720078,
+      "loss": 2.1608,
+      "step": 12100
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001808170125470972,
+      "loss": 2.1624,
+      "step": 12105
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018079319396591923,
+      "loss": 2.0992,
+      "step": 12110
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018076936217756054,
+      "loss": 2.1672,
+      "step": 12115
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001807455171859168,
+      "loss": 2.1534,
+      "step": 12120
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000180721658994886,
+      "loss": 2.1749,
+      "step": 12125
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018069778760836818,
+      "loss": 2.1389,
+      "step": 12130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018067390303026556,
+      "loss": 2.1326,
+      "step": 12135
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018065000526448256,
+      "loss": 2.1168,
+      "step": 12140
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018062609431492572,
+      "loss": 2.1365,
+      "step": 12145
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001806021701855037,
+      "loss": 2.0948,
+      "step": 12150
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001805782328801274,
+      "loss": 2.1316,
+      "step": 12155
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001805542824027098,
+      "loss": 2.1357,
+      "step": 12160
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001805303187571661,
+      "loss": 2.1145,
+      "step": 12165
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018050634194741354,
+      "loss": 2.1272,
+      "step": 12170
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018048235197737166,
+      "loss": 2.1553,
+      "step": 12175
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018045834885096196,
+      "loss": 2.164,
+      "step": 12180
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018043433257210835,
+      "loss": 2.1107,
+      "step": 12185
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018041030314473668,
+      "loss": 2.1367,
+      "step": 12190
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018038626057277502,
+      "loss": 2.1025,
+      "step": 12195
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00018036220486015357,
+      "loss": 2.0917,
+      "step": 12200
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001803381360108047,
+      "loss": 2.1308,
+      "step": 12205
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001803140540286629,
+      "loss": 2.1602,
+      "step": 12210
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00018028995891766493,
+      "loss": 2.0635,
+      "step": 12215
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018026585068174947,
+      "loss": 2.1716,
+      "step": 12220
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018024172932485752,
+      "loss": 2.1264,
+      "step": 12225
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018021759485093217,
+      "loss": 2.1512,
+      "step": 12230
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018019344726391868,
+      "loss": 2.1132,
+      "step": 12235
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018016928656776435,
+      "loss": 2.1505,
+      "step": 12240
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001801451127664188,
+      "loss": 2.1064,
+      "step": 12245
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018012092586383366,
+      "loss": 2.172,
+      "step": 12250
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018009672586396272,
+      "loss": 2.0944,
+      "step": 12255
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018007251277076195,
+      "loss": 2.1585,
+      "step": 12260
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018004828658818944,
+      "loss": 2.1119,
+      "step": 12265
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001800240473202054,
+      "loss": 2.1376,
+      "step": 12270
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017999979497077226,
+      "loss": 2.1129,
+      "step": 12275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017997552954385446,
+      "loss": 2.1868,
+      "step": 12280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017995125104341866,
+      "loss": 2.0972,
+      "step": 12285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017992695947343363,
+      "loss": 2.1676,
+      "step": 12290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001799026548378703,
+      "loss": 2.1526,
+      "step": 12295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017987833714070174,
+      "loss": 2.0914,
+      "step": 12300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017985400638590316,
+      "loss": 2.1285,
+      "step": 12305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001798296625774518,
+      "loss": 2.1524,
+      "step": 12310
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001798053057193272,
+      "loss": 2.1054,
+      "step": 12315
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017978093581551087,
+      "loss": 2.0697,
+      "step": 12320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017975655286998663,
+      "loss": 2.1255,
+      "step": 12325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017973215688674027,
+      "loss": 2.148,
+      "step": 12330
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001797077478697598,
+      "loss": 2.1583,
+      "step": 12335
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017968332582303536,
+      "loss": 2.1842,
+      "step": 12340
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001796588907505591,
+      "loss": 2.1137,
+      "step": 12345
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017963444265632552,
+      "loss": 2.1079,
+      "step": 12350
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017960998154433106,
+      "loss": 2.1324,
+      "step": 12355
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017958550741857434,
+      "loss": 2.0708,
+      "step": 12360
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017956102028305617,
+      "loss": 2.1093,
+      "step": 12365
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001795365201417794,
+      "loss": 2.1288,
+      "step": 12370
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017951200699874905,
+      "loss": 2.1224,
+      "step": 12375
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017948748085797224,
+      "loss": 2.1337,
+      "step": 12380
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017946294172345827,
+      "loss": 2.115,
+      "step": 12385
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017943838959921853,
+      "loss": 2.0979,
+      "step": 12390
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017941382448926648,
+      "loss": 2.1378,
+      "step": 12395
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017938924639761782,
+      "loss": 2.1304,
+      "step": 12400
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017936465532829022,
+      "loss": 2.1363,
+      "step": 12405
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017934005128530364,
+      "loss": 2.1463,
+      "step": 12410
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017931543427268006,
+      "loss": 2.0988,
+      "step": 12415
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017929080429444357,
+      "loss": 2.1494,
+      "step": 12420
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001792661613546204,
+      "loss": 2.1744,
+      "step": 12425
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017924150545723896,
+      "loss": 2.076,
+      "step": 12430
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017921683660632965,
+      "loss": 2.106,
+      "step": 12435
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017919215480592513,
+      "loss": 2.1046,
+      "step": 12440
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017916746006006008,
+      "loss": 2.1502,
+      "step": 12445
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00017914275237277132,
+      "loss": 2.1273,
+      "step": 12450
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001791180317480978,
+      "loss": 2.0918,
+      "step": 12455
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017909329819008058,
+      "loss": 2.1553,
+      "step": 12460
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001790685517027628,
+      "loss": 2.1662,
+      "step": 12465
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001790437922901898,
+      "loss": 2.111,
+      "step": 12470
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017901901995640888,
+      "loss": 2.1815,
+      "step": 12475
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017899423470546966,
+      "loss": 2.125,
+      "step": 12480
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001789694365414237,
+      "loss": 2.1277,
+      "step": 12485
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017894462546832473,
+      "loss": 2.091,
+      "step": 12490
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001789198014902286,
+      "loss": 2.1386,
+      "step": 12495
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017889496461119325,
+      "loss": 2.1747,
+      "step": 12500
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017887011483527876,
+      "loss": 2.1152,
+      "step": 12505
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001788452521665473,
+      "loss": 2.1152,
+      "step": 12510
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00017882037660906313,
+      "loss": 2.1265,
+      "step": 12515
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001787954881668926,
+      "loss": 2.1895,
+      "step": 12520
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017877058684410423,
+      "loss": 2.1536,
+      "step": 12525
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017874567264476864,
+      "loss": 2.1725,
+      "step": 12530
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001787207455729585,
+      "loss": 2.119,
+      "step": 12535
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017869580563274862,
+      "loss": 2.1676,
+      "step": 12540
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001786708528282159,
+      "loss": 2.1064,
+      "step": 12545
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017864588716343932,
+      "loss": 2.1723,
+      "step": 12550
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017862090864250008,
+      "loss": 2.1933,
+      "step": 12555
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001785959172694813,
+      "loss": 2.0967,
+      "step": 12560
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017857091304846834,
+      "loss": 2.1185,
+      "step": 12565
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017854589598354862,
+      "loss": 2.1223,
+      "step": 12570
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017852086607881163,
+      "loss": 2.1541,
+      "step": 12575
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017849582333834897,
+      "loss": 2.174,
+      "step": 12580
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001784707677662544,
+      "loss": 2.1569,
+      "step": 12585
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001784456993666237,
+      "loss": 2.1694,
+      "step": 12590
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017842061814355478,
+      "loss": 2.1138,
+      "step": 12595
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017839552410114763,
+      "loss": 2.1607,
+      "step": 12600
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00017837041724350438,
+      "loss": 2.0624,
+      "step": 12605
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017834529757472918,
+      "loss": 2.1493,
+      "step": 12610
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017832016509892833,
+      "loss": 2.0913,
+      "step": 12615
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017829501982021025,
+      "loss": 2.156,
+      "step": 12620
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017826986174268532,
+      "loss": 2.13,
+      "step": 12625
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001782446908704662,
+      "loss": 2.1069,
+      "step": 12630
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017821950720766752,
+      "loss": 2.1197,
+      "step": 12635
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000178194310758406,
+      "loss": 2.1843,
+      "step": 12640
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017816910152680052,
+      "loss": 2.1492,
+      "step": 12645
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000178143879516972,
+      "loss": 2.1924,
+      "step": 12650
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001781186447330434,
+      "loss": 2.0997,
+      "step": 12655
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017809339717913986,
+      "loss": 2.1295,
+      "step": 12660
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017806813685938863,
+      "loss": 2.16,
+      "step": 12665
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017804286377791888,
+      "loss": 2.1831,
+      "step": 12670
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017801757793886206,
+      "loss": 2.0899,
+      "step": 12675
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017799227934635158,
+      "loss": 2.1152,
+      "step": 12680
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017796696800452297,
+      "loss": 2.114,
+      "step": 12685
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017794164391751388,
+      "loss": 2.142,
+      "step": 12690
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017791630708946401,
+      "loss": 2.1533,
+      "step": 12695
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017789095752451513,
+      "loss": 2.1613,
+      "step": 12700
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017786559522681107,
+      "loss": 2.0517,
+      "step": 12705
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017784022020049786,
+      "loss": 2.146,
+      "step": 12710
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017781483244972348,
+      "loss": 2.1149,
+      "step": 12715
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017778943197863806,
+      "loss": 2.1512,
+      "step": 12720
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017776401879139372,
+      "loss": 2.1351,
+      "step": 12725
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017773859289214484,
+      "loss": 2.1525,
+      "step": 12730
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017771315428504768,
+      "loss": 2.1131,
+      "step": 12735
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017768770297426075,
+      "loss": 2.1593,
+      "step": 12740
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017766223896394445,
+      "loss": 2.1105,
+      "step": 12745
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001776367622582614,
+      "loss": 2.1746,
+      "step": 12750
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017761127286137625,
+      "loss": 2.0963,
+      "step": 12755
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001775857707774557,
+      "loss": 2.1012,
+      "step": 12760
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001775602560106686,
+      "loss": 2.145,
+      "step": 12765
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017753472856518583,
+      "loss": 2.0636,
+      "step": 12770
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017750918844518026,
+      "loss": 2.184,
+      "step": 12775
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000177483635654827,
+      "loss": 2.1957,
+      "step": 12780
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017745807019830303,
+      "loss": 2.1316,
+      "step": 12785
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017743249207978758,
+      "loss": 2.1561,
+      "step": 12790
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017740690130346187,
+      "loss": 2.1264,
+      "step": 12795
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001773812978735092,
+      "loss": 2.14,
+      "step": 12800
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017735568179411495,
+      "loss": 2.1817,
+      "step": 12805
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001773300530694665,
+      "loss": 2.16,
+      "step": 12810
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017730441170375343,
+      "loss": 2.2036,
+      "step": 12815
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017727875770116725,
+      "loss": 2.1592,
+      "step": 12820
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001772530910659016,
+      "loss": 2.169,
+      "step": 12825
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017722741180215218,
+      "loss": 2.1034,
+      "step": 12830
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017720171991411675,
+      "loss": 2.1283,
+      "step": 12835
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001771760154059952,
+      "loss": 2.1504,
+      "step": 12840
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001771502982819893,
+      "loss": 2.1222,
+      "step": 12845
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017712456854630313,
+      "loss": 2.1717,
+      "step": 12850
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001770988262031426,
+      "loss": 2.2035,
+      "step": 12855
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001770730712567158,
+      "loss": 2.1163,
+      "step": 12860
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017704730371123298,
+      "loss": 2.1639,
+      "step": 12865
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017702152357090616,
+      "loss": 2.193,
+      "step": 12870
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017699573083994972,
+      "loss": 2.1232,
+      "step": 12875
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017696992552257993,
+      "loss": 2.0808,
+      "step": 12880
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017694410762301516,
+      "loss": 2.2211,
+      "step": 12885
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017691827714547582,
+      "loss": 2.1205,
+      "step": 12890
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00017689243409418443,
+      "loss": 2.1879,
+      "step": 12895
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001768665784733655,
+      "loss": 2.1324,
+      "step": 12900
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017684071028724557,
+      "loss": 2.1154,
+      "step": 12905
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001768148295400534,
+      "loss": 2.1655,
+      "step": 12910
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001767889362360196,
+      "loss": 2.119,
+      "step": 12915
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000176763030379377,
+      "loss": 2.0916,
+      "step": 12920
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017673711197436034,
+      "loss": 2.0698,
+      "step": 12925
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017671118102520646,
+      "loss": 2.1646,
+      "step": 12930
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017668523753615429,
+      "loss": 2.081,
+      "step": 12935
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017665928151144481,
+      "loss": 2.1821,
+      "step": 12940
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017663331295532104,
+      "loss": 2.113,
+      "step": 12945
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017660733187202797,
+      "loss": 2.12,
+      "step": 12950
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017658133826581272,
+      "loss": 2.1196,
+      "step": 12955
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017655533214092447,
+      "loss": 2.1562,
+      "step": 12960
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001765293135016144,
+      "loss": 2.1113,
+      "step": 12965
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001765032823521357,
+      "loss": 2.1097,
+      "step": 12970
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017647723869674376,
+      "loss": 2.0713,
+      "step": 12975
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001764511825396958,
+      "loss": 2.1383,
+      "step": 12980
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001764251138852513,
+      "loss": 2.1576,
+      "step": 12985
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017639903273767159,
+      "loss": 2.1683,
+      "step": 12990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017637293910122015,
+      "loss": 2.1412,
+      "step": 12995
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001763468329801625,
+      "loss": 2.1125,
+      "step": 13000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017632071437876618,
+      "loss": 2.1936,
+      "step": 13005
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00017629458330130072,
+      "loss": 2.1455,
+      "step": 13010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017626843975203783,
+      "loss": 2.0965,
+      "step": 13015
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001762422837352511,
+      "loss": 2.1826,
+      "step": 13020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017621611525521624,
+      "loss": 2.1249,
+      "step": 13025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000176189934316211,
+      "loss": 2.0584,
+      "step": 13030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017616374092251514,
+      "loss": 2.1002,
+      "step": 13035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017613753507841046,
+      "loss": 2.1658,
+      "step": 13040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017611131678818082,
+      "loss": 2.1375,
+      "step": 13045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001760850860561121,
+      "loss": 2.159,
+      "step": 13050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001760588428864922,
+      "loss": 2.0945,
+      "step": 13055
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017603258728361105,
+      "loss": 2.134,
+      "step": 13060
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017600631925176064,
+      "loss": 2.1002,
+      "step": 13065
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017598003879523502,
+      "loss": 2.1629,
+      "step": 13070
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017595374591833016,
+      "loss": 2.142,
+      "step": 13075
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017592744062534417,
+      "loss": 2.1222,
+      "step": 13080
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001759011229205771,
+      "loss": 2.1112,
+      "step": 13085
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017587479280833115,
+      "loss": 2.1517,
+      "step": 13090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017584845029291045,
+      "loss": 2.1299,
+      "step": 13095
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017582209537862116,
+      "loss": 2.1603,
+      "step": 13100
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001757957280697715,
+      "loss": 2.0936,
+      "step": 13105
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017576934837067176,
+      "loss": 2.1278,
+      "step": 13110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017574295628563412,
+      "loss": 2.1387,
+      "step": 13115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017571655181897292,
+      "loss": 2.1729,
+      "step": 13120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017569013497500447,
+      "loss": 2.1653,
+      "step": 13125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017566370575804706,
+      "loss": 2.1532,
+      "step": 13130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017563726417242108,
+      "loss": 2.1612,
+      "step": 13135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017561081022244893,
+      "loss": 2.2138,
+      "step": 13140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017558434391245497,
+      "loss": 2.1079,
+      "step": 13145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017555786524676563,
+      "loss": 2.1125,
+      "step": 13150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017553137422970935,
+      "loss": 2.1886,
+      "step": 13155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017550487086561662,
+      "loss": 2.1627,
+      "step": 13160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001754783551588199,
+      "loss": 2.096,
+      "step": 13165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017545182711365363,
+      "loss": 2.1002,
+      "step": 13170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017542528673445445,
+      "loss": 2.1364,
+      "step": 13175
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017539873402556074,
+      "loss": 2.1073,
+      "step": 13180
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017537216899131316,
+      "loss": 2.1782,
+      "step": 13185
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001753455916360542,
+      "loss": 2.1899,
+      "step": 13190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017531900196412846,
+      "loss": 2.1759,
+      "step": 13195
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017529239997988256,
+      "loss": 2.1621,
+      "step": 13200
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017526578568766502,
+      "loss": 2.1493,
+      "step": 13205
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017523915909182654,
+      "loss": 2.0908,
+      "step": 13210
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017521252019671963,
+      "loss": 2.1523,
+      "step": 13215
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017518586900669902,
+      "loss": 2.1843,
+      "step": 13220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017515920552612136,
+      "loss": 2.1415,
+      "step": 13225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001751325297593452,
+      "loss": 2.142,
+      "step": 13230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017510584171073133,
+      "loss": 2.1537,
+      "step": 13235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001750791413846423,
+      "loss": 2.1499,
+      "step": 13240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001750524287854429,
+      "loss": 2.1769,
+      "step": 13245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017502570391749966,
+      "loss": 2.1345,
+      "step": 13250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017499896678518142,
+      "loss": 2.125,
+      "step": 13255
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017497221739285878,
+      "loss": 2.11,
+      "step": 13260
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001749454557449045,
+      "loss": 2.0746,
+      "step": 13265
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001749186818456932,
+      "loss": 2.1652,
+      "step": 13270
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017489189569960164,
+      "loss": 2.1341,
+      "step": 13275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001748650973110085,
+      "loss": 2.1375,
+      "step": 13280
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001748382866842945,
+      "loss": 2.0905,
+      "step": 13285
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017481146382384235,
+      "loss": 2.1329,
+      "step": 13290
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017478462873403672,
+      "loss": 2.0823,
+      "step": 13295
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017475778141926439,
+      "loss": 2.1015,
+      "step": 13300
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017473092188391394,
+      "loss": 2.1371,
+      "step": 13305
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001747040501323762,
+      "loss": 2.1634,
+      "step": 13310
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017467716616904378,
+      "loss": 2.1324,
+      "step": 13315
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017465026999831143,
+      "loss": 2.1801,
+      "step": 13320
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017462336162457577,
+      "loss": 2.1468,
+      "step": 13325
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017459644105223557,
+      "loss": 2.1756,
+      "step": 13330
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017456950828569144,
+      "loss": 2.1591,
+      "step": 13335
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017454256332934609,
+      "loss": 2.1905,
+      "step": 13340
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001745156061876042,
+      "loss": 2.1407,
+      "step": 13345
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017448863686487235,
+      "loss": 2.1246,
+      "step": 13350
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017446165536555926,
+      "loss": 2.1122,
+      "step": 13355
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017443466169407557,
+      "loss": 2.1176,
+      "step": 13360
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017440765585483386,
+      "loss": 2.1544,
+      "step": 13365
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017438063785224877,
+      "loss": 2.1109,
+      "step": 13370
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017435360769073693,
+      "loss": 2.1688,
+      "step": 13375
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017432656537471686,
+      "loss": 2.1443,
+      "step": 13380
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017429951090860922,
+      "loss": 2.1402,
+      "step": 13385
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017427244429683655,
+      "loss": 2.1651,
+      "step": 13390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017424536554382338,
+      "loss": 2.172,
+      "step": 13395
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017421827465399626,
+      "loss": 2.1627,
+      "step": 13400
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001741911716317837,
+      "loss": 2.1361,
+      "step": 13405
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001741640564816162,
+      "loss": 2.1574,
+      "step": 13410
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017413692920792632,
+      "loss": 2.1413,
+      "step": 13415
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017410978981514846,
+      "loss": 2.1289,
+      "step": 13420
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017408263830771904,
+      "loss": 2.1413,
+      "step": 13425
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017405547469007655,
+      "loss": 2.1403,
+      "step": 13430
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017402829896666136,
+      "loss": 2.1116,
+      "step": 13435
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001740011111419159,
+      "loss": 2.141,
+      "step": 13440
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017397391122028446,
+      "loss": 2.1471,
+      "step": 13445
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017394669920621345,
+      "loss": 2.1073,
+      "step": 13450
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001739194751041512,
+      "loss": 2.1196,
+      "step": 13455
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017389223891854796,
+      "loss": 2.1571,
+      "step": 13460
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000173864990653856,
+      "loss": 2.1272,
+      "step": 13465
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017383773031452962,
+      "loss": 2.1741,
+      "step": 13470
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000173810457905025,
+      "loss": 2.1568,
+      "step": 13475
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001737831734298003,
+      "loss": 2.1463,
+      "step": 13480
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017375587689331576,
+      "loss": 2.122,
+      "step": 13485
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017372856830003344,
+      "loss": 2.1244,
+      "step": 13490
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017370124765441754,
+      "loss": 2.1372,
+      "step": 13495
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00017367391496093406,
+      "loss": 2.0958,
+      "step": 13500
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001736465702240511,
+      "loss": 2.1674,
+      "step": 13505
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017361921344823862,
+      "loss": 2.1095,
+      "step": 13510
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017359184463796866,
+      "loss": 2.1354,
+      "step": 13515
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001735644637977151,
+      "loss": 2.1187,
+      "step": 13520
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017353707093195392,
+      "loss": 2.08,
+      "step": 13525
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000173509666045163,
+      "loss": 2.121,
+      "step": 13530
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001734822491418222,
+      "loss": 2.1593,
+      "step": 13535
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017345482022641326,
+      "loss": 2.1017,
+      "step": 13540
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017342737930342003,
+      "loss": 2.1293,
+      "step": 13545
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017339992637732823,
+      "loss": 2.1261,
+      "step": 13550
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017337246145262555,
+      "loss": 2.1635,
+      "step": 13555
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017334498453380162,
+      "loss": 2.1759,
+      "step": 13560
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017331749562534817,
+      "loss": 2.1446,
+      "step": 13565
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001732899947317587,
+      "loss": 2.1355,
+      "step": 13570
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017326248185752877,
+      "loss": 2.185,
+      "step": 13575
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017323495700715587,
+      "loss": 2.1506,
+      "step": 13580
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017320742018513943,
+      "loss": 2.1622,
+      "step": 13585
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017317987139598098,
+      "loss": 2.1792,
+      "step": 13590
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017315231064418378,
+      "loss": 2.1657,
+      "step": 13595
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001731247379342532,
+      "loss": 2.1274,
+      "step": 13600
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017309715327069655,
+      "loss": 2.2496,
+      "step": 13605
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017306955665802301,
+      "loss": 2.1888,
+      "step": 13610
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001730419481007438,
+      "loss": 2.2019,
+      "step": 13615
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017301432760337204,
+      "loss": 2.0917,
+      "step": 13620
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017298669517042288,
+      "loss": 2.0501,
+      "step": 13625
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001729590508064133,
+      "loss": 2.1161,
+      "step": 13630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017293139451586236,
+      "loss": 2.0955,
+      "step": 13635
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017290372630329094,
+      "loss": 2.1491,
+      "step": 13640
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000172876046173222,
+      "loss": 2.1536,
+      "step": 13645
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017284835413018034,
+      "loss": 2.1672,
+      "step": 13650
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017282065017869272,
+      "loss": 2.0829,
+      "step": 13655
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001727929343232879,
+      "loss": 2.1674,
+      "step": 13660
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017276520656849665,
+      "loss": 2.1637,
+      "step": 13665
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017273746691885147,
+      "loss": 2.1637,
+      "step": 13670
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017270971537888702,
+      "loss": 2.1667,
+      "step": 13675
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017268195195313974,
+      "loss": 2.1373,
+      "step": 13680
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017265417664614815,
+      "loss": 2.1542,
+      "step": 13685
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017262638946245265,
+      "loss": 2.1312,
+      "step": 13690
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017259859040659556,
+      "loss": 2.211,
+      "step": 13695
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017257077948312118,
+      "loss": 2.1279,
+      "step": 13700
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001725429566965757,
+      "loss": 2.0782,
+      "step": 13705
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001725151220515073,
+      "loss": 2.1594,
+      "step": 13710
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001724872755524661,
+      "loss": 2.1418,
+      "step": 13715
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017245941720400412,
+      "loss": 2.1376,
+      "step": 13720
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017243154701067535,
+      "loss": 2.08,
+      "step": 13725
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017240366497703566,
+      "loss": 2.1783,
+      "step": 13730
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017237577110764296,
+      "loss": 2.1366,
+      "step": 13735
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000172347865407057,
+      "loss": 2.1119,
+      "step": 13740
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017231994787983947,
+      "loss": 2.1824,
+      "step": 13745
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017229201853055407,
+      "loss": 2.1874,
+      "step": 13750
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00017226407736376634,
+      "loss": 2.1064,
+      "step": 13755
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017223612438404385,
+      "loss": 2.0941,
+      "step": 13760
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017220815959595595,
+      "loss": 2.1649,
+      "step": 13765
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017218018300407413,
+      "loss": 2.139,
+      "step": 13770
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001721521946129716,
+      "loss": 2.11,
+      "step": 13775
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017212419442722365,
+      "loss": 2.1015,
+      "step": 13780
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017209618245140742,
+      "loss": 2.1839,
+      "step": 13785
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017206815869010199,
+      "loss": 2.1367,
+      "step": 13790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001720401231478884,
+      "loss": 2.1215,
+      "step": 13795
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001720120758293496,
+      "loss": 2.1855,
+      "step": 13800
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017198401673907041,
+      "loss": 2.1966,
+      "step": 13805
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017195594588163764,
+      "loss": 2.1653,
+      "step": 13810
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017192786326164004,
+      "loss": 2.1731,
+      "step": 13815
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017189976888366818,
+      "loss": 2.0981,
+      "step": 13820
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001718716627523147,
+      "loss": 2.1394,
+      "step": 13825
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017184354487217402,
+      "loss": 2.1664,
+      "step": 13830
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017181541524784256,
+      "loss": 2.1061,
+      "step": 13835
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017178727388391862,
+      "loss": 2.1661,
+      "step": 13840
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001717591207850025,
+      "loss": 2.173,
+      "step": 13845
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001717309559556963,
+      "loss": 2.0666,
+      "step": 13850
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001717027794006041,
+      "loss": 2.1514,
+      "step": 13855
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001716745911243319,
+      "loss": 2.1435,
+      "step": 13860
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017164639113148765,
+      "loss": 2.1114,
+      "step": 13865
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017161817942668112,
+      "loss": 2.1286,
+      "step": 13870
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017158995601452407,
+      "loss": 2.0965,
+      "step": 13875
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017156172089963017,
+      "loss": 2.1687,
+      "step": 13880
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017153347408661494,
+      "loss": 2.1164,
+      "step": 13885
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001715052155800959,
+      "loss": 2.1535,
+      "step": 13890
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001714769453846924,
+      "loss": 2.1429,
+      "step": 13895
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017144866350502579,
+      "loss": 2.2438,
+      "step": 13900
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017142036994571922,
+      "loss": 2.1586,
+      "step": 13905
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017139206471139786,
+      "loss": 2.1567,
+      "step": 13910
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001713637478066887,
+      "loss": 2.1321,
+      "step": 13915
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00017133541923622074,
+      "loss": 2.0636,
+      "step": 13920
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017130707900462472,
+      "loss": 2.1163,
+      "step": 13925
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001712787271165335,
+      "loss": 2.1053,
+      "step": 13930
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017125036357658164,
+      "loss": 2.1677,
+      "step": 13935
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00017122198838940577,
+      "loss": 2.1405,
+      "step": 13940
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017119360155964431,
+      "loss": 2.1164,
+      "step": 13945
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017116520309193764,
+      "loss": 2.1667,
+      "step": 13950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000171136792990928,
+      "loss": 2.1365,
+      "step": 13955
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001711083712612596,
+      "loss": 2.1195,
+      "step": 13960
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001710799379075785,
+      "loss": 2.1051,
+      "step": 13965
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00017105149293453268,
+      "loss": 2.1071,
+      "step": 13970
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017102303634677202,
+      "loss": 2.1238,
+      "step": 13975
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017099456814894824,
+      "loss": 2.1569,
+      "step": 13980
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000170966088345715,
+      "loss": 2.1658,
+      "step": 13985
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017093759694172795,
+      "loss": 2.0974,
+      "step": 13990
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001709090939416445,
+      "loss": 2.1186,
+      "step": 13995
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000170880579350124,
+      "loss": 2.1289,
+      "step": 14000
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017085205317182768,
+      "loss": 2.1778,
+      "step": 14005
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017082351541141875,
+      "loss": 2.116,
+      "step": 14010
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001707949660735622,
+      "loss": 2.1846,
+      "step": 14015
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017076640516292498,
+      "loss": 2.1616,
+      "step": 14020
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017073783268417589,
+      "loss": 2.138,
+      "step": 14025
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017070924864198564,
+      "loss": 2.1617,
+      "step": 14030
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001706806530410269,
+      "loss": 2.1304,
+      "step": 14035
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001706520458859741,
+      "loss": 2.1726,
+      "step": 14040
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001706234271815036,
+      "loss": 2.1189,
+      "step": 14045
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00017059479693229377,
+      "loss": 2.1869,
+      "step": 14050
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017056615514302468,
+      "loss": 2.1758,
+      "step": 14055
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001705375018183784,
+      "loss": 2.1467,
+      "step": 14060
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017050883696303885,
+      "loss": 2.1322,
+      "step": 14065
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017048016058169189,
+      "loss": 2.1232,
+      "step": 14070
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00017045147267902514,
+      "loss": 2.1828,
+      "step": 14075
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001704227732597283,
+      "loss": 2.1404,
+      "step": 14080
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001703940623284927,
+      "loss": 2.1463,
+      "step": 14085
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017036533989001176,
+      "loss": 2.1487,
+      "step": 14090
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017033660594898068,
+      "loss": 2.1506,
+      "step": 14095
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001703078605100966,
+      "loss": 2.1134,
+      "step": 14100
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017027910357805847,
+      "loss": 2.1438,
+      "step": 14105
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00017025033515756722,
+      "loss": 2.1105,
+      "step": 14110
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001702215552533255,
+      "loss": 2.1563,
+      "step": 14115
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00017019276387003802,
+      "loss": 2.1593,
+      "step": 14120
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00017016396101241123,
+      "loss": 2.1471,
+      "step": 14125
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017013514668515352,
+      "loss": 2.0918,
+      "step": 14130
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00017010632089297514,
+      "loss": 2.1319,
+      "step": 14135
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001700774836405882,
+      "loss": 2.1594,
+      "step": 14140
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017004863493270668,
+      "loss": 2.1139,
+      "step": 14145
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001700197747740465,
+      "loss": 2.1148,
+      "step": 14150
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016999090316932534,
+      "loss": 2.1717,
+      "step": 14155
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016996202012326288,
+      "loss": 2.1258,
+      "step": 14160
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016993312564058055,
+      "loss": 2.1338,
+      "step": 14165
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016990421972600173,
+      "loss": 2.1403,
+      "step": 14170
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001698753023842516,
+      "loss": 2.1021,
+      "step": 14175
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016984637362005728,
+      "loss": 2.1626,
+      "step": 14180
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001698174334381477,
+      "loss": 2.0984,
+      "step": 14185
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016978848184325375,
+      "loss": 2.0895,
+      "step": 14190
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00016975951884010807,
+      "loss": 2.1369,
+      "step": 14195
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016973054443344517,
+      "loss": 2.1509,
+      "step": 14200
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016970155862800152,
+      "loss": 2.1882,
+      "step": 14205
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001696725614285154,
+      "loss": 2.1573,
+      "step": 14210
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016964355283972696,
+      "loss": 2.1446,
+      "step": 14215
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016961453286637817,
+      "loss": 2.0779,
+      "step": 14220
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001695855015132129,
+      "loss": 2.1785,
+      "step": 14225
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001695564587849769,
+      "loss": 2.1321,
+      "step": 14230
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016952740468641775,
+      "loss": 2.1385,
+      "step": 14235
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00016949833922228486,
+      "loss": 2.1163,
+      "step": 14240
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016946926239732959,
+      "loss": 2.1778,
+      "step": 14245
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016944017421630501,
+      "loss": 2.1259,
+      "step": 14250
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00016941107468396627,
+      "loss": 2.1369,
+      "step": 14255
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016938196380507012,
+      "loss": 2.1715,
+      "step": 14260
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016935284158437535,
+      "loss": 2.1283,
+      "step": 14265
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001693237080266425,
+      "loss": 2.1457,
+      "step": 14270
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000169294563136634,
+      "loss": 2.1047,
+      "step": 14275
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001692654069191142,
+      "loss": 2.1419,
+      "step": 14280
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016923623937884914,
+      "loss": 2.104,
+      "step": 14285
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001692070605206069,
+      "loss": 2.1397,
+      "step": 14290
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016917787034915726,
+      "loss": 2.1768,
+      "step": 14295
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016914866886927194,
+      "loss": 2.2167,
+      "step": 14300
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016911945608572446,
+      "loss": 2.1712,
+      "step": 14305
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0001690902320032902,
+      "loss": 2.1785,
+      "step": 14310
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001690609966267464,
+      "loss": 2.1341,
+      "step": 14315
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016903174996087212,
+      "loss": 2.1845,
+      "step": 14320
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001690024920104483,
+      "loss": 2.1375,
+      "step": 14325
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016897322278025773,
+      "loss": 2.0983,
+      "step": 14330
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016894394227508496,
+      "loss": 2.167,
+      "step": 14335
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001689146504997165,
+      "loss": 2.1286,
+      "step": 14340
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016888534745894065,
+      "loss": 2.1266,
+      "step": 14345
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001688560331575475,
+      "loss": 2.1808,
+      "step": 14350
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016882670760032907,
+      "loss": 2.1283,
+      "step": 14355
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016879737079207913,
+      "loss": 2.1208,
+      "step": 14360
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001687680227375934,
+      "loss": 2.2056,
+      "step": 14365
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016873866344166931,
+      "loss": 2.1903,
+      "step": 14370
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016870929290910627,
+      "loss": 2.1432,
+      "step": 14375
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001686799111447054,
+      "loss": 2.1384,
+      "step": 14380
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001686505181532697,
+      "loss": 2.1255,
+      "step": 14385
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016862111393960402,
+      "loss": 2.1505,
+      "step": 14390
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016859169850851508,
+      "loss": 2.1788,
+      "step": 14395
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016856227186481135,
+      "loss": 2.1317,
+      "step": 14400
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016853283401330318,
+      "loss": 2.1225,
+      "step": 14405
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016850338495880273,
+      "loss": 2.1473,
+      "step": 14410
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016847392470612405,
+      "loss": 2.1722,
+      "step": 14415
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016844445326008294,
+      "loss": 2.1161,
+      "step": 14420
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016841497062549705,
+      "loss": 2.139,
+      "step": 14425
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016838547680718592,
+      "loss": 2.1284,
+      "step": 14430
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016835597180997087,
+      "loss": 2.1554,
+      "step": 14435
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016832645563867504,
+      "loss": 2.0747,
+      "step": 14440
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001682969282981234,
+      "loss": 2.0626,
+      "step": 14445
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016826738979314277,
+      "loss": 2.1004,
+      "step": 14450
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00016823784012856175,
+      "loss": 2.1047,
+      "step": 14455
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016820827930921084,
+      "loss": 2.1407,
+      "step": 14460
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016817870733992229,
+      "loss": 2.1148,
+      "step": 14465
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001681491242255302,
+      "loss": 2.186,
+      "step": 14470
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001681195299708705,
+      "loss": 2.1364,
+      "step": 14475
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001680899245807809,
+      "loss": 2.0979,
+      "step": 14480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000168060308060101,
+      "loss": 2.1223,
+      "step": 14485
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016803068041367216,
+      "loss": 2.1101,
+      "step": 14490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016800104164633762,
+      "loss": 2.155,
+      "step": 14495
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016797139176294234,
+      "loss": 2.1349,
+      "step": 14500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016794173076833323,
+      "loss": 2.1087,
+      "step": 14505
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016791205866735885,
+      "loss": 2.214,
+      "step": 14510
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016788237546486973,
+      "loss": 2.1551,
+      "step": 14515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00016785268116571814,
+      "loss": 2.1072,
+      "step": 14520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001678229757747582,
+      "loss": 2.1849,
+      "step": 14525
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016779325929684578,
+      "loss": 2.0973,
+      "step": 14530
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016776353173683864,
+      "loss": 2.117,
+      "step": 14535
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016773379309959624,
+      "loss": 2.1053,
+      "step": 14540
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016770404338998,
+      "loss": 2.1216,
+      "step": 14545
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016767428261285308,
+      "loss": 2.1372,
+      "step": 14550
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016764451077308038,
+      "loss": 2.1493,
+      "step": 14555
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016761472787552871,
+      "loss": 2.1012,
+      "step": 14560
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016758493392506668,
+      "loss": 2.0982,
+      "step": 14565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016755512892656463,
+      "loss": 2.1435,
+      "step": 14570
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016752531288489474,
+      "loss": 2.1026,
+      "step": 14575
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016749548580493107,
+      "loss": 2.1371,
+      "step": 14580
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016746564769154934,
+      "loss": 2.1314,
+      "step": 14585
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016743579854962726,
+      "loss": 2.1671,
+      "step": 14590
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016740593838404414,
+      "loss": 2.1533,
+      "step": 14595
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016737606719968126,
+      "loss": 2.107,
+      "step": 14600
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001673461850014216,
+      "loss": 2.1216,
+      "step": 14605
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016731629179414994,
+      "loss": 2.1822,
+      "step": 14610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016728638758275297,
+      "loss": 2.104,
+      "step": 14615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016725647237211905,
+      "loss": 2.1131,
+      "step": 14620
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001672265461671384,
+      "loss": 2.1599,
+      "step": 14625
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000167196608972703,
+      "loss": 2.1985,
+      "step": 14630
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001671666607937067,
+      "loss": 2.1669,
+      "step": 14635
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001671367016350451,
+      "loss": 2.1256,
+      "step": 14640
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016710673150161553,
+      "loss": 2.1884,
+      "step": 14645
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001670767503983172,
+      "loss": 2.117,
+      "step": 14650
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016704675833005114,
+      "loss": 2.1397,
+      "step": 14655
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016701675530172007,
+      "loss": 2.0927,
+      "step": 14660
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016698674131822858,
+      "loss": 2.1224,
+      "step": 14665
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000166956716384483,
+      "loss": 2.155,
+      "step": 14670
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001669266805053915,
+      "loss": 2.0944,
+      "step": 14675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016689663368586405,
+      "loss": 2.161,
+      "step": 14680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016686657593081226,
+      "loss": 2.1774,
+      "step": 14685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016683650724514973,
+      "loss": 2.1657,
+      "step": 14690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016680642763379177,
+      "loss": 2.1569,
+      "step": 14695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016677633710165542,
+      "loss": 2.1899,
+      "step": 14700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016674623565365954,
+      "loss": 2.1539,
+      "step": 14705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016671612329472485,
+      "loss": 2.1593,
+      "step": 14710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001666860000297737,
+      "loss": 2.1393,
+      "step": 14715
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016665586586373033,
+      "loss": 2.1288,
+      "step": 14720
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001666257208015208,
+      "loss": 2.1607,
+      "step": 14725
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016659556484807282,
+      "loss": 2.1304,
+      "step": 14730
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000166565398008316,
+      "loss": 2.1635,
+      "step": 14735
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016653522028718166,
+      "loss": 2.1378,
+      "step": 14740
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016650503168960293,
+      "loss": 2.1395,
+      "step": 14745
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016647483222051473,
+      "loss": 2.1899,
+      "step": 14750
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001664446218848537,
+      "loss": 2.1421,
+      "step": 14755
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016641440068755828,
+      "loss": 2.1541,
+      "step": 14760
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016638416863356875,
+      "loss": 2.1431,
+      "step": 14765
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016635392572782707,
+      "loss": 2.1674,
+      "step": 14770
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016632367197527708,
+      "loss": 2.1435,
+      "step": 14775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016629340738086426,
+      "loss": 2.1009,
+      "step": 14780
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016626313194953592,
+      "loss": 2.1428,
+      "step": 14785
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016623284568624124,
+      "loss": 2.1577,
+      "step": 14790
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.000166202548595931,
+      "loss": 2.1142,
+      "step": 14795
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001661722406835579,
+      "loss": 2.1492,
+      "step": 14800
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001661419219540763,
+      "loss": 2.1582,
+      "step": 14805
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016611159241244238,
+      "loss": 2.1816,
+      "step": 14810
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001660812520636141,
+      "loss": 2.1346,
+      "step": 14815
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016605090091255113,
+      "loss": 2.1768,
+      "step": 14820
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016602053896421496,
+      "loss": 2.1033,
+      "step": 14825
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016599016622356887,
+      "loss": 2.1664,
+      "step": 14830
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016595978269557776,
+      "loss": 2.1246,
+      "step": 14835
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016592938838520848,
+      "loss": 2.1518,
+      "step": 14840
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016589898329742955,
+      "loss": 2.1464,
+      "step": 14845
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001658685674372112,
+      "loss": 2.1468,
+      "step": 14850
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016583814080952555,
+      "loss": 2.1007,
+      "step": 14855
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016580770341934634,
+      "loss": 2.0453,
+      "step": 14860
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001657772552716492,
+      "loss": 2.125,
+      "step": 14865
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016574679637141145,
+      "loss": 2.1002,
+      "step": 14870
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016571632672361214,
+      "loss": 2.1206,
+      "step": 14875
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016568584633323212,
+      "loss": 2.1084,
+      "step": 14880
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000165655355205254,
+      "loss": 2.1423,
+      "step": 14885
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016562485334466214,
+      "loss": 2.1552,
+      "step": 14890
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001655943407564427,
+      "loss": 2.1041,
+      "step": 14895
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001655638174455834,
+      "loss": 2.1917,
+      "step": 14900
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016553328341707398,
+      "loss": 2.1561,
+      "step": 14905
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016550273867590572,
+      "loss": 2.1012,
+      "step": 14910
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016547218322707187,
+      "loss": 2.1402,
+      "step": 14915
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001654416170755671,
+      "loss": 2.1344,
+      "step": 14920
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001654110402263882,
+      "loss": 2.1499,
+      "step": 14925
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016538045268453346,
+      "loss": 2.1788,
+      "step": 14930
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016534985445500303,
+      "loss": 2.1781,
+      "step": 14935
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016531924554279872,
+      "loss": 2.1377,
+      "step": 14940
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016528862595292416,
+      "loss": 2.1279,
+      "step": 14945
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016525799569038472,
+      "loss": 2.1802,
+      "step": 14950
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001652273547601875,
+      "loss": 2.1502,
+      "step": 14955
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001651967031673413,
+      "loss": 2.1105,
+      "step": 14960
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016516604091685677,
+      "loss": 2.1366,
+      "step": 14965
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016513536801374617,
+      "loss": 2.1642,
+      "step": 14970
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001651046844630236,
+      "loss": 2.1302,
+      "step": 14975
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016507399026970487,
+      "loss": 2.1429,
+      "step": 14980
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016504328543880754,
+      "loss": 2.1167,
+      "step": 14985
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016501256997535085,
+      "loss": 2.1922,
+      "step": 14990
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016498184388435592,
+      "loss": 2.1209,
+      "step": 14995
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001649511071708454,
+      "loss": 2.1548,
+      "step": 15000
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016492035983984383,
+      "loss": 2.1324,
+      "step": 15005
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016488960189637748,
+      "loss": 2.1299,
+      "step": 15010
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016485883334547426,
+      "loss": 2.1074,
+      "step": 15015
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016482805419216394,
+      "loss": 2.1578,
+      "step": 15020
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001647972644414779,
+      "loss": 2.0989,
+      "step": 15025
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016476646409844937,
+      "loss": 2.0962,
+      "step": 15030
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016473565316811316,
+      "loss": 2.1503,
+      "step": 15035
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000164704831655506,
+      "loss": 2.0978,
+      "step": 15040
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016467399956566619,
+      "loss": 2.0911,
+      "step": 15045
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016464315690363385,
+      "loss": 2.0765,
+      "step": 15050
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016461230367445073,
+      "loss": 2.1471,
+      "step": 15055
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016458143988316048,
+      "loss": 2.1608,
+      "step": 15060
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001645505655348083,
+      "loss": 2.0666,
+      "step": 15065
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001645196806344412,
+      "loss": 2.106,
+      "step": 15070
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016448878518710797,
+      "loss": 2.1435,
+      "step": 15075
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016445787919785895,
+      "loss": 2.1497,
+      "step": 15080
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016442696267174637,
+      "loss": 2.0981,
+      "step": 15085
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016439603561382415,
+      "loss": 2.1299,
+      "step": 15090
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016436509802914784,
+      "loss": 2.1248,
+      "step": 15095
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001643341499227748,
+      "loss": 2.12,
+      "step": 15100
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001643031912997641,
+      "loss": 2.1829,
+      "step": 15105
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016427222216517648,
+      "loss": 2.1274,
+      "step": 15110
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016424124252407447,
+      "loss": 2.1759,
+      "step": 15115
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016421025238152229,
+      "loss": 2.1142,
+      "step": 15120
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001641792517425858,
+      "loss": 2.1343,
+      "step": 15125
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016414824061233269,
+      "loss": 2.1427,
+      "step": 15130
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016411721899583235,
+      "loss": 2.1859,
+      "step": 15135
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001640861868981558,
+      "loss": 2.1322,
+      "step": 15140
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016405514432437583,
+      "loss": 2.146,
+      "step": 15145
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016402409127956697,
+      "loss": 2.1681,
+      "step": 15150
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016399302776880538,
+      "loss": 2.0804,
+      "step": 15155
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016396195379716902,
+      "loss": 2.082,
+      "step": 15160
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00016393086936973753,
+      "loss": 2.1626,
+      "step": 15165
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001638997744915922,
+      "loss": 2.1749,
+      "step": 15170
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016386866916781614,
+      "loss": 2.1333,
+      "step": 15175
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016383755340349402,
+      "loss": 2.175,
+      "step": 15180
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001638064272037124,
+      "loss": 2.1788,
+      "step": 15185
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001637752905735594,
+      "loss": 2.1215,
+      "step": 15190
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016374414351812488,
+      "loss": 2.1513,
+      "step": 15195
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016371298604250044,
+      "loss": 2.1492,
+      "step": 15200
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001636818181517793,
+      "loss": 2.1662,
+      "step": 15205
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016365063985105657,
+      "loss": 2.158,
+      "step": 15210
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016361945114542884,
+      "loss": 2.152,
+      "step": 15215
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001635882520399945,
+      "loss": 2.11,
+      "step": 15220
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016355704253985369,
+      "loss": 2.1624,
+      "step": 15225
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001635258226501081,
+      "loss": 2.1425,
+      "step": 15230
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016349459237586134,
+      "loss": 2.1243,
+      "step": 15235
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016346335172221847,
+      "loss": 2.1562,
+      "step": 15240
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016343210069428646,
+      "loss": 2.1058,
+      "step": 15245
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001634008392971738,
+      "loss": 2.2019,
+      "step": 15250
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016336956753599085,
+      "loss": 2.1088,
+      "step": 15255
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016333828541584947,
+      "loss": 2.1508,
+      "step": 15260
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016330699294186341,
+      "loss": 2.1259,
+      "step": 15265
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000163275690119148,
+      "loss": 2.1313,
+      "step": 15270
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001632443769528202,
+      "loss": 2.1102,
+      "step": 15275
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016321305344799885,
+      "loss": 2.1749,
+      "step": 15280
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001631817196098043,
+      "loss": 2.1164,
+      "step": 15285
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00016315037544335868,
+      "loss": 2.1171,
+      "step": 15290
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001631190209537858,
+      "loss": 2.1283,
+      "step": 15295
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016308765614621114,
+      "loss": 2.1179,
+      "step": 15300
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016305628102576187,
+      "loss": 2.1268,
+      "step": 15305
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016302489559756684,
+      "loss": 2.1904,
+      "step": 15310
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001629934998667566,
+      "loss": 2.1286,
+      "step": 15315
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016296209383846342,
+      "loss": 2.1689,
+      "step": 15320
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016293067751782116,
+      "loss": 2.1147,
+      "step": 15325
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001628992509099654,
+      "loss": 2.1403,
+      "step": 15330
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016286781402003348,
+      "loss": 2.1287,
+      "step": 15335
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016283636685316427,
+      "loss": 2.122,
+      "step": 15340
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001628049094144985,
+      "loss": 2.1815,
+      "step": 15345
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016277344170917847,
+      "loss": 2.124,
+      "step": 15350
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001627419637423481,
+      "loss": 2.1943,
+      "step": 15355
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016271047551915307,
+      "loss": 2.1616,
+      "step": 15360
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016267897704474082,
+      "loss": 2.1817,
+      "step": 15365
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001626474683242603,
+      "loss": 2.1526,
+      "step": 15370
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001626159493628622,
+      "loss": 2.1317,
+      "step": 15375
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016258442016569893,
+      "loss": 2.0953,
+      "step": 15380
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016255288073792454,
+      "loss": 2.1183,
+      "step": 15385
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001625213310846947,
+      "loss": 2.11,
+      "step": 15390
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016248977121116688,
+      "loss": 2.1431,
+      "step": 15395
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016245820112250004,
+      "loss": 2.1108,
+      "step": 15400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000162426620823855,
+      "loss": 2.1426,
+      "step": 15405
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00016239503032039408,
+      "loss": 2.1507,
+      "step": 15410
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001623634296172814,
+      "loss": 2.0867,
+      "step": 15415
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001623318187196827,
+      "loss": 2.151,
+      "step": 15420
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016230019763276534,
+      "loss": 2.1122,
+      "step": 15425
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016226856636169842,
+      "loss": 2.1781,
+      "step": 15430
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016223692491165265,
+      "loss": 2.1693,
+      "step": 15435
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016220527328780047,
+      "loss": 2.1278,
+      "step": 15440
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016217361149531588,
+      "loss": 2.068,
+      "step": 15445
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001621419395393746,
+      "loss": 2.0845,
+      "step": 15450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001621102574251541,
+      "loss": 2.1593,
+      "step": 15455
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001620785651578333,
+      "loss": 2.1073,
+      "step": 15460
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000162046862742593,
+      "loss": 2.169,
+      "step": 15465
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001620151501846155,
+      "loss": 2.1374,
+      "step": 15470
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016198342748908486,
+      "loss": 2.2181,
+      "step": 15475
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016195169466118675,
+      "loss": 2.1608,
+      "step": 15480
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016191995170610849,
+      "loss": 2.1378,
+      "step": 15485
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016188819862903903,
+      "loss": 2.139,
+      "step": 15490
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016185643543516909,
+      "loss": 2.1121,
+      "step": 15495
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016182466212969092,
+      "loss": 2.1114,
+      "step": 15500
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016179287871779852,
+      "loss": 2.1556,
+      "step": 15505
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001617610852046874,
+      "loss": 2.1579,
+      "step": 15510
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001617292815955549,
+      "loss": 2.1286,
+      "step": 15515
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016169746789559988,
+      "loss": 2.1404,
+      "step": 15520
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001616656441100229,
+      "loss": 2.1949,
+      "step": 15525
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016163381024402614,
+      "loss": 2.1848,
+      "step": 15530
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016160196630281352,
+      "loss": 2.1315,
+      "step": 15535
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016157011229159048,
+      "loss": 2.1804,
+      "step": 15540
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016153824821556415,
+      "loss": 2.1415,
+      "step": 15545
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00016150637407994335,
+      "loss": 2.1587,
+      "step": 15550
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016147448988993853,
+      "loss": 2.1941,
+      "step": 15555
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001614425956507617,
+      "loss": 2.136,
+      "step": 15560
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016141069136762663,
+      "loss": 2.0866,
+      "step": 15565
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016137877704574865,
+      "loss": 2.1581,
+      "step": 15570
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016134685269034483,
+      "loss": 2.1576,
+      "step": 15575
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016131491830663372,
+      "loss": 2.21,
+      "step": 15580
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016128297389983567,
+      "loss": 2.158,
+      "step": 15585
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016125101947517256,
+      "loss": 2.0873,
+      "step": 15590
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016121905503786797,
+      "loss": 2.1333,
+      "step": 15595
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016118708059314707,
+      "loss": 2.1238,
+      "step": 15600
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001611550961462367,
+      "loss": 2.1604,
+      "step": 15605
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016112310170236533,
+      "loss": 2.1924,
+      "step": 15610
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016109109726676306,
+      "loss": 2.127,
+      "step": 15615
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016105908284466162,
+      "loss": 2.0979,
+      "step": 15620
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016102705844129437,
+      "loss": 2.0953,
+      "step": 15625
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016099502406189631,
+      "loss": 2.1669,
+      "step": 15630
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016096297971170408,
+      "loss": 2.1175,
+      "step": 15635
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001609309253959559,
+      "loss": 2.1382,
+      "step": 15640
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001608988611198917,
+      "loss": 2.1415,
+      "step": 15645
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016086678688875296,
+      "loss": 2.1488,
+      "step": 15650
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016083470270778286,
+      "loss": 2.1894,
+      "step": 15655
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001608026085822261,
+      "loss": 2.1818,
+      "step": 15660
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016077050451732917,
+      "loss": 2.1265,
+      "step": 15665
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016073839051834,
+      "loss": 2.1049,
+      "step": 15670
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001607062665905083,
+      "loss": 2.1406,
+      "step": 15675
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001606741327390853,
+      "loss": 2.1805,
+      "step": 15680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001606419889693239,
+      "loss": 2.0967,
+      "step": 15685
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016060983528647858,
+      "loss": 2.1145,
+      "step": 15690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001605776716958055,
+      "loss": 2.144,
+      "step": 15695
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001605454982025624,
+      "loss": 2.1034,
+      "step": 15700
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016051331481200867,
+      "loss": 2.1056,
+      "step": 15705
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016048112152940527,
+      "loss": 2.1578,
+      "step": 15710
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016044891836001482,
+      "loss": 2.1447,
+      "step": 15715
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001604167053091015,
+      "loss": 2.1666,
+      "step": 15720
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016038448238193115,
+      "loss": 2.1395,
+      "step": 15725
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016035224958377132,
+      "loss": 2.1343,
+      "step": 15730
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001603200069198909,
+      "loss": 2.1661,
+      "step": 15735
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00016028775439556073,
+      "loss": 2.1537,
+      "step": 15740
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000160255492016053,
+      "loss": 2.162,
+      "step": 15745
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016022321978664163,
+      "loss": 2.2456,
+      "step": 15750
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016019093771260214,
+      "loss": 2.073,
+      "step": 15755
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016015864579921164,
+      "loss": 2.188,
+      "step": 15760
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016012634405174886,
+      "loss": 2.1576,
+      "step": 15765
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016009403247549414,
+      "loss": 2.1576,
+      "step": 15770
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016006171107572938,
+      "loss": 2.1703,
+      "step": 15775
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016002937985773818,
+      "loss": 2.1393,
+      "step": 15780
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015999703882680564,
+      "loss": 2.0992,
+      "step": 15785
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00015996468798821858,
+      "loss": 2.1285,
+      "step": 15790
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015993232734726531,
+      "loss": 2.1099,
+      "step": 15795
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001598999569092358,
+      "loss": 2.1471,
+      "step": 15800
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015986757667942162,
+      "loss": 2.158,
+      "step": 15805
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001598351866631159,
+      "loss": 2.0817,
+      "step": 15810
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015980278686561345,
+      "loss": 2.1249,
+      "step": 15815
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001597703772922106,
+      "loss": 2.1407,
+      "step": 15820
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015973795794820537,
+      "loss": 2.1276,
+      "step": 15825
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001597055288388972,
+      "loss": 2.1982,
+      "step": 15830
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015967308996958738,
+      "loss": 2.1694,
+      "step": 15835
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015964064134557854,
+      "loss": 2.1464,
+      "step": 15840
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015960818297217508,
+      "loss": 2.1489,
+      "step": 15845
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015957571485468295,
+      "loss": 2.1575,
+      "step": 15850
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015954323699840964,
+      "loss": 2.1837,
+      "step": 15855
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015951074940866432,
+      "loss": 2.1552,
+      "step": 15860
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015947825209075767,
+      "loss": 2.1557,
+      "step": 15865
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000159445745050002,
+      "loss": 2.1086,
+      "step": 15870
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001594132282917112,
+      "loss": 2.1711,
+      "step": 15875
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015938070182120075,
+      "loss": 2.163,
+      "step": 15880
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015934816564378771,
+      "loss": 2.0641,
+      "step": 15885
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001593156197647908,
+      "loss": 2.1141,
+      "step": 15890
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015928306418953016,
+      "loss": 2.1101,
+      "step": 15895
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015925049892332767,
+      "loss": 2.1176,
+      "step": 15900
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015921792397150677,
+      "loss": 2.1796,
+      "step": 15905
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015918533933939236,
+      "loss": 2.153,
+      "step": 15910
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015915274503231115,
+      "loss": 2.1045,
+      "step": 15915
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015912014105559118,
+      "loss": 2.1141,
+      "step": 15920
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015908752741456222,
+      "loss": 2.1433,
+      "step": 15925
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015905490411455565,
+      "loss": 2.1276,
+      "step": 15930
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015902227116090431,
+      "loss": 2.1555,
+      "step": 15935
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001589896285589427,
+      "loss": 2.1621,
+      "step": 15940
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001589569763140068,
+      "loss": 2.1083,
+      "step": 15945
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015892431443143435,
+      "loss": 2.1943,
+      "step": 15950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015889164291656448,
+      "loss": 2.1252,
+      "step": 15955
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015885896177473798,
+      "loss": 2.1277,
+      "step": 15960
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015882627101129724,
+      "loss": 2.1099,
+      "step": 15965
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015879357063158613,
+      "loss": 2.1237,
+      "step": 15970
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015876086064095022,
+      "loss": 2.0854,
+      "step": 15975
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001587281410447365,
+      "loss": 2.1176,
+      "step": 15980
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015869541184829368,
+      "loss": 2.1292,
+      "step": 15985
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001586626730569719,
+      "loss": 2.1352,
+      "step": 15990
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000158629924676123,
+      "loss": 2.1671,
+      "step": 15995
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001585971667111003,
+      "loss": 2.1205,
+      "step": 16000
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015856439916725864,
+      "loss": 2.1724,
+      "step": 16005
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015853162204995463,
+      "loss": 2.1188,
+      "step": 16010
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001584988353645462,
+      "loss": 2.1041,
+      "step": 16015
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015846603911639302,
+      "loss": 2.0687,
+      "step": 16020
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015843323331085624,
+      "loss": 2.116,
+      "step": 16025
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001584004179532986,
+      "loss": 2.1146,
+      "step": 16030
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015836759304908434,
+      "loss": 2.1715,
+      "step": 16035
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015833475860357941,
+      "loss": 2.1178,
+      "step": 16040
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015830191462215117,
+      "loss": 2.1518,
+      "step": 16045
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00015826906111016857,
+      "loss": 2.1588,
+      "step": 16050
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015823619807300214,
+      "loss": 2.037,
+      "step": 16055
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000158203325516024,
+      "loss": 2.1309,
+      "step": 16060
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015817044344460784,
+      "loss": 2.1318,
+      "step": 16065
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015813755186412876,
+      "loss": 2.1138,
+      "step": 16070
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015810465077996356,
+      "loss": 2.1219,
+      "step": 16075
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015807174019749054,
+      "loss": 2.1696,
+      "step": 16080
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015803882012208954,
+      "loss": 2.1113,
+      "step": 16085
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015800589055914204,
+      "loss": 2.1372,
+      "step": 16090
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001579729515140309,
+      "loss": 2.1694,
+      "step": 16095
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00015794000299214068,
+      "loss": 2.1287,
+      "step": 16100
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001579070449988575,
+      "loss": 2.1728,
+      "step": 16105
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015787407753956888,
+      "loss": 2.1172,
+      "step": 16110
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000157841100619664,
+      "loss": 2.1316,
+      "step": 16115
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015780811424453362,
+      "loss": 2.1107,
+      "step": 16120
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015777511841956994,
+      "loss": 2.1208,
+      "step": 16125
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015774211315016672,
+      "loss": 2.126,
+      "step": 16130
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015770909844171934,
+      "loss": 2.1447,
+      "step": 16135
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001576760742996247,
+      "loss": 2.1356,
+      "step": 16140
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001576430407292812,
+      "loss": 2.1085,
+      "step": 16145
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001576099977360888,
+      "loss": 2.1701,
+      "step": 16150
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000157576945325449,
+      "loss": 2.109,
+      "step": 16155
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015754388350276485,
+      "loss": 2.0931,
+      "step": 16160
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015751081227344095,
+      "loss": 2.1601,
+      "step": 16165
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001574777316428834,
+      "loss": 2.1114,
+      "step": 16170
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015744464161649986,
+      "loss": 2.1681,
+      "step": 16175
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015741154219969951,
+      "loss": 2.1311,
+      "step": 16180
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001573784333978931,
+      "loss": 2.1507,
+      "step": 16185
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001573453152164929,
+      "loss": 2.0949,
+      "step": 16190
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015731218766091266,
+      "loss": 2.1355,
+      "step": 16195
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00015727905073656775,
+      "loss": 2.1222,
+      "step": 16200
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000157245904448875,
+      "loss": 2.1849,
+      "step": 16205
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015721274880325283,
+      "loss": 2.1149,
+      "step": 16210
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015717958380512116,
+      "loss": 2.1617,
+      "step": 16215
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001571464094599014,
+      "loss": 2.1815,
+      "step": 16220
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015711322577301656,
+      "loss": 2.1143,
+      "step": 16225
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015708003274989113,
+      "loss": 2.1094,
+      "step": 16230
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015704683039595113,
+      "loss": 2.1288,
+      "step": 16235
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015701361871662414,
+      "loss": 2.1285,
+      "step": 16240
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015698039771733924,
+      "loss": 2.1105,
+      "step": 16245
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000156947167403527,
+      "loss": 2.0918,
+      "step": 16250
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015691392778061956,
+      "loss": 2.1209,
+      "step": 16255
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015688067885405058,
+      "loss": 2.0956,
+      "step": 16260
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015684742062925522,
+      "loss": 2.1569,
+      "step": 16265
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015681415311167015,
+      "loss": 2.1589,
+      "step": 16270
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015678087630673364,
+      "loss": 2.1173,
+      "step": 16275
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015674759021988533,
+      "loss": 2.1235,
+      "step": 16280
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015671429485656654,
+      "loss": 2.14,
+      "step": 16285
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015668099022221995,
+      "loss": 2.0921,
+      "step": 16290
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015664767632228986,
+      "loss": 2.1163,
+      "step": 16295
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015661435316222214,
+      "loss": 2.1906,
+      "step": 16300
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015658102074746398,
+      "loss": 2.0884,
+      "step": 16305
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015654767908346426,
+      "loss": 2.1026,
+      "step": 16310
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015651432817567325,
+      "loss": 2.1383,
+      "step": 16315
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015648096802954288,
+      "loss": 2.1864,
+      "step": 16320
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015644759865052644,
+      "loss": 2.1292,
+      "step": 16325
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001564142200440788,
+      "loss": 2.1109,
+      "step": 16330
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015638083221565632,
+      "loss": 2.1868,
+      "step": 16335
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015634743517071686,
+      "loss": 2.1416,
+      "step": 16340
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015631402891471986,
+      "loss": 2.1526,
+      "step": 16345
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015628061345312615,
+      "loss": 2.1247,
+      "step": 16350
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015624718879139813,
+      "loss": 2.114,
+      "step": 16355
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001562137549349997,
+      "loss": 2.1213,
+      "step": 16360
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015618031188939627,
+      "loss": 2.1099,
+      "step": 16365
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001561468596600547,
+      "loss": 2.0994,
+      "step": 16370
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015611339825244343,
+      "loss": 2.1734,
+      "step": 16375
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001560799276720324,
+      "loss": 2.0778,
+      "step": 16380
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015604644792429292,
+      "loss": 2.1192,
+      "step": 16385
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015601295901469792,
+      "loss": 2.1437,
+      "step": 16390
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015597946094872187,
+      "loss": 2.1075,
+      "step": 16395
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015594595373184057,
+      "loss": 2.1006,
+      "step": 16400
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015591243736953145,
+      "loss": 2.1347,
+      "step": 16405
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015587891186727342,
+      "loss": 2.1815,
+      "step": 16410
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015584537723054682,
+      "loss": 2.1393,
+      "step": 16415
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015581183346483353,
+      "loss": 2.1622,
+      "step": 16420
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001557782805756169,
+      "loss": 2.1555,
+      "step": 16425
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015574471856838183,
+      "loss": 2.1339,
+      "step": 16430
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015571114744861468,
+      "loss": 2.0762,
+      "step": 16435
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015567756722180325,
+      "loss": 2.1331,
+      "step": 16440
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015564397789343684,
+      "loss": 2.1551,
+      "step": 16445
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015561037946900628,
+      "loss": 2.1267,
+      "step": 16450
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015557677195400386,
+      "loss": 2.1104,
+      "step": 16455
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015554315535392342,
+      "loss": 2.1813,
+      "step": 16460
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015550952967426023,
+      "loss": 2.1494,
+      "step": 16465
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015547589492051096,
+      "loss": 2.1144,
+      "step": 16470
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015544225109817393,
+      "loss": 2.1643,
+      "step": 16475
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015540859821274881,
+      "loss": 2.1371,
+      "step": 16480
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015537493626973685,
+      "loss": 2.1575,
+      "step": 16485
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001553412652746407,
+      "loss": 2.14,
+      "step": 16490
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001553075852329645,
+      "loss": 2.1428,
+      "step": 16495
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015527389615021397,
+      "loss": 2.1617,
+      "step": 16500
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015524019803189618,
+      "loss": 2.1338,
+      "step": 16505
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001552064908835197,
+      "loss": 2.1342,
+      "step": 16510
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015517277471059466,
+      "loss": 2.1262,
+      "step": 16515
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001551390495186326,
+      "loss": 2.1411,
+      "step": 16520
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001551053153131465,
+      "loss": 2.1023,
+      "step": 16525
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015507157209965086,
+      "loss": 2.1114,
+      "step": 16530
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001550378198836617,
+      "loss": 2.0953,
+      "step": 16535
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015500405867069643,
+      "loss": 2.1293,
+      "step": 16540
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015497028846627394,
+      "loss": 2.1379,
+      "step": 16545
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015493650927591465,
+      "loss": 2.0791,
+      "step": 16550
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001549027211051404,
+      "loss": 2.131,
+      "step": 16555
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015486892395947448,
+      "loss": 2.1258,
+      "step": 16560
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001548351178444417,
+      "loss": 2.1482,
+      "step": 16565
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001548013027655683,
+      "loss": 2.1614,
+      "step": 16570
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000154767478728382,
+      "loss": 2.1653,
+      "step": 16575
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000154733645738412,
+      "loss": 2.1712,
+      "step": 16580
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015469980380118892,
+      "loss": 2.1328,
+      "step": 16585
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015466595292224487,
+      "loss": 2.1166,
+      "step": 16590
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015463209310711342,
+      "loss": 2.138,
+      "step": 16595
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015459822436132958,
+      "loss": 2.1583,
+      "step": 16600
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015456434669042988,
+      "loss": 2.1078,
+      "step": 16605
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015453046009995228,
+      "loss": 2.1474,
+      "step": 16610
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001544965645954361,
+      "loss": 2.107,
+      "step": 16615
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001544626601824223,
+      "loss": 2.1422,
+      "step": 16620
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015442874686645314,
+      "loss": 2.1229,
+      "step": 16625
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015439482465307242,
+      "loss": 2.1242,
+      "step": 16630
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015436089354782534,
+      "loss": 2.1282,
+      "step": 16635
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015432695355625862,
+      "loss": 2.1458,
+      "step": 16640
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015429300468392041,
+      "loss": 2.1653,
+      "step": 16645
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015425904693636025,
+      "loss": 2.1595,
+      "step": 16650
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015422508031912916,
+      "loss": 2.1892,
+      "step": 16655
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001541911048377797,
+      "loss": 2.1159,
+      "step": 16660
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015415712049786576,
+      "loss": 2.1537,
+      "step": 16665
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015412312730494275,
+      "loss": 2.1158,
+      "step": 16670
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015408912526456749,
+      "loss": 2.1856,
+      "step": 16675
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015405511438229826,
+      "loss": 2.081,
+      "step": 16680
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015402109466369475,
+      "loss": 2.1335,
+      "step": 16685
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015398706611431822,
+      "loss": 2.1698,
+      "step": 16690
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015395302873973118,
+      "loss": 2.0887,
+      "step": 16695
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015391898254549775,
+      "loss": 2.1266,
+      "step": 16700
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015388492753718336,
+      "loss": 2.1669,
+      "step": 16705
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015385086372035506,
+      "loss": 2.1328,
+      "step": 16710
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015381679110058113,
+      "loss": 2.1782,
+      "step": 16715
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015378270968343142,
+      "loss": 2.0816,
+      "step": 16720
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001537486194744772,
+      "loss": 2.1289,
+      "step": 16725
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001537145204792911,
+      "loss": 2.145,
+      "step": 16730
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001536804127034473,
+      "loss": 2.1598,
+      "step": 16735
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015364629615252136,
+      "loss": 2.121,
+      "step": 16740
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015361217083209027,
+      "loss": 2.1519,
+      "step": 16745
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015357803674773246,
+      "loss": 2.1389,
+      "step": 16750
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015354389390502782,
+      "loss": 2.0861,
+      "step": 16755
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001535097423095576,
+      "loss": 2.1302,
+      "step": 16760
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015347558196690452,
+      "loss": 2.1126,
+      "step": 16765
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001534414128826528,
+      "loss": 2.1716,
+      "step": 16770
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000153407235062388,
+      "loss": 2.1506,
+      "step": 16775
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001533730485116971,
+      "loss": 2.1539,
+      "step": 16780
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015333885323616856,
+      "loss": 2.1668,
+      "step": 16785
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015330464924139225,
+      "loss": 2.1263,
+      "step": 16790
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015327043653295947,
+      "loss": 2.1159,
+      "step": 16795
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015323621511646293,
+      "loss": 2.1421,
+      "step": 16800
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001532019849974968,
+      "loss": 2.2243,
+      "step": 16805
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015316774618165654,
+      "loss": 2.1423,
+      "step": 16810
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015313349867453923,
+      "loss": 2.202,
+      "step": 16815
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015309924248174328,
+      "loss": 2.1078,
+      "step": 16820
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015306497760886843,
+      "loss": 2.1024,
+      "step": 16825
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015303070406151603,
+      "loss": 2.1606,
+      "step": 16830
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015299642184528865,
+      "loss": 2.1457,
+      "step": 16835
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015296213096579044,
+      "loss": 2.086,
+      "step": 16840
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015292783142862687,
+      "loss": 2.1274,
+      "step": 16845
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015289352323940482,
+      "loss": 2.1368,
+      "step": 16850
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015285920640373266,
+      "loss": 2.0729,
+      "step": 16855
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015282488092722009,
+      "loss": 2.1709,
+      "step": 16860
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001527905468154783,
+      "loss": 2.0944,
+      "step": 16865
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001527562040741198,
+      "loss": 2.1477,
+      "step": 16870
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015272185270875861,
+      "loss": 2.0686,
+      "step": 16875
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015268749272501008,
+      "loss": 2.1559,
+      "step": 16880
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015265312412849103,
+      "loss": 2.1578,
+      "step": 16885
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015261874692481966,
+      "loss": 2.1202,
+      "step": 16890
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015258436111961554,
+      "loss": 2.221,
+      "step": 16895
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001525499667184997,
+      "loss": 2.1301,
+      "step": 16900
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015251556372709456,
+      "loss": 2.0972,
+      "step": 16905
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015248115215102398,
+      "loss": 2.1165,
+      "step": 16910
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001524467319959131,
+      "loss": 2.149,
+      "step": 16915
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015241230326738862,
+      "loss": 2.1283,
+      "step": 16920
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015237786597107854,
+      "loss": 2.1699,
+      "step": 16925
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001523434201126123,
+      "loss": 2.1569,
+      "step": 16930
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015230896569762075,
+      "loss": 2.1619,
+      "step": 16935
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015227450273173605,
+      "loss": 2.1029,
+      "step": 16940
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015224003122059186,
+      "loss": 2.1976,
+      "step": 16945
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015220555116982325,
+      "loss": 2.1298,
+      "step": 16950
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015217106258506657,
+      "loss": 2.1669,
+      "step": 16955
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015213656547195966,
+      "loss": 2.1812,
+      "step": 16960
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015210205983614175,
+      "loss": 2.1003,
+      "step": 16965
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015206754568325345,
+      "loss": 2.1596,
+      "step": 16970
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015203302301893668,
+      "loss": 2.1446,
+      "step": 16975
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001519984918488349,
+      "loss": 2.1511,
+      "step": 16980
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001519639521785928,
+      "loss": 2.0988,
+      "step": 16985
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015192940401385666,
+      "loss": 2.1537,
+      "step": 16990
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015189484736027393,
+      "loss": 2.121,
+      "step": 16995
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015186028222349363,
+      "loss": 2.2082,
+      "step": 17000
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00015182570860916602,
+      "loss": 2.1123,
+      "step": 17005
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015179112652294286,
+      "loss": 2.1473,
+      "step": 17010
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015175653597047722,
+      "loss": 2.1251,
+      "step": 17015
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001517219369574236,
+      "loss": 2.1514,
+      "step": 17020
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015168732948943786,
+      "loss": 2.1155,
+      "step": 17025
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015165271357217726,
+      "loss": 2.1307,
+      "step": 17030
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015161808921130039,
+      "loss": 2.1074,
+      "step": 17035
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015158345641246728,
+      "loss": 2.1121,
+      "step": 17040
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015154881518133933,
+      "loss": 2.1118,
+      "step": 17045
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015151416552357933,
+      "loss": 2.1242,
+      "step": 17050
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015147950744485137,
+      "loss": 2.1354,
+      "step": 17055
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000151444840950821,
+      "loss": 2.113,
+      "step": 17060
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001514101660471551,
+      "loss": 2.1725,
+      "step": 17065
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015137548273952195,
+      "loss": 2.1657,
+      "step": 17070
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015134079103359122,
+      "loss": 2.1593,
+      "step": 17075
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001513060909350339,
+      "loss": 2.1163,
+      "step": 17080
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015127138244952238,
+      "loss": 2.1351,
+      "step": 17085
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001512366655827304,
+      "loss": 2.1416,
+      "step": 17090
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00015120194034033314,
+      "loss": 2.1391,
+      "step": 17095
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015116720672800707,
+      "loss": 2.0779,
+      "step": 17100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015113246475143005,
+      "loss": 2.0871,
+      "step": 17105
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015109771441628134,
+      "loss": 2.0728,
+      "step": 17110
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001510629557282415,
+      "loss": 2.0854,
+      "step": 17115
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001510281886929926,
+      "loss": 2.1162,
+      "step": 17120
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015099341331621782,
+      "loss": 2.0857,
+      "step": 17125
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015095862960360197,
+      "loss": 2.1779,
+      "step": 17130
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015092383756083107,
+      "loss": 2.1515,
+      "step": 17135
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015088903719359252,
+      "loss": 2.1718,
+      "step": 17140
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015085422850757516,
+      "loss": 2.1575,
+      "step": 17145
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015081941150846907,
+      "loss": 2.1186,
+      "step": 17150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001507845862019658,
+      "loss": 2.1192,
+      "step": 17155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015074975259375814,
+      "loss": 2.1471,
+      "step": 17160
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001507149106895404,
+      "loss": 2.1564,
+      "step": 17165
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015068006049500802,
+      "loss": 2.0964,
+      "step": 17170
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00015064520201585804,
+      "loss": 2.144,
+      "step": 17175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001506103352577887,
+      "loss": 2.1667,
+      "step": 17180
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001505754602264996,
+      "loss": 2.1011,
+      "step": 17185
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001505405769276918,
+      "loss": 2.1688,
+      "step": 17190
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015050568536706754,
+      "loss": 2.1387,
+      "step": 17195
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00015047078555033064,
+      "loss": 2.1358,
+      "step": 17200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.000150435877483186,
+      "loss": 2.1592,
+      "step": 17205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00015040096117134008,
+      "loss": 2.1153,
+      "step": 17210
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001503660366205006,
+      "loss": 2.1487,
+      "step": 17215
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015033110383637664,
+      "loss": 2.1063,
+      "step": 17220
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001502961628246786,
+      "loss": 2.1509,
+      "step": 17225
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015026121359111832,
+      "loss": 2.1379,
+      "step": 17230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015022625614140889,
+      "loss": 2.155,
+      "step": 17235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015019129048126473,
+      "loss": 2.1392,
+      "step": 17240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015015631661640164,
+      "loss": 2.131,
+      "step": 17245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015012133455253682,
+      "loss": 2.1392,
+      "step": 17250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015008634429538873,
+      "loss": 2.0772,
+      "step": 17255
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015005134585067715,
+      "loss": 2.1609,
+      "step": 17260
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00015001633922412326,
+      "loss": 2.1262,
+      "step": 17265
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001499813244214496,
+      "loss": 2.0787,
+      "step": 17270
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001499463014483799,
+      "loss": 2.0992,
+      "step": 17275
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014991127031063948,
+      "loss": 2.1597,
+      "step": 17280
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014987623101395472,
+      "loss": 2.1823,
+      "step": 17285
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001498411835640535,
+      "loss": 2.1405,
+      "step": 17290
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000149806127966665,
+      "loss": 2.1195,
+      "step": 17295
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014977106422751972,
+      "loss": 2.1273,
+      "step": 17300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001497359923523495,
+      "loss": 2.1473,
+      "step": 17305
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014970091234688747,
+      "loss": 2.1505,
+      "step": 17310
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014966582421686817,
+      "loss": 2.1229,
+      "step": 17315
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001496307279680274,
+      "loss": 2.1518,
+      "step": 17320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014959562360610226,
+      "loss": 2.1687,
+      "step": 17325
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014956051113683132,
+      "loss": 2.1298,
+      "step": 17330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001495253905659543,
+      "loss": 2.1264,
+      "step": 17335
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014949026189921234,
+      "loss": 2.1646,
+      "step": 17340
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014945512514234793,
+      "loss": 2.1834,
+      "step": 17345
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014941998030110475,
+      "loss": 2.1215,
+      "step": 17350
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000149384827381228,
+      "loss": 2.1116,
+      "step": 17355
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014934966638846403,
+      "loss": 2.1721,
+      "step": 17360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001493144973285606,
+      "loss": 2.2014,
+      "step": 17365
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014927932020726673,
+      "loss": 2.1416,
+      "step": 17370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001492441350303328,
+      "loss": 2.1394,
+      "step": 17375
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001492089418035105,
+      "loss": 2.1054,
+      "step": 17380
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001491737405325528,
+      "loss": 2.0881,
+      "step": 17385
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014913853122321404,
+      "loss": 2.1215,
+      "step": 17390
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014910331388124986,
+      "loss": 2.1735,
+      "step": 17395
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014906808851241723,
+      "loss": 2.1124,
+      "step": 17400
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014903285512247435,
+      "loss": 2.182,
+      "step": 17405
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001489976137171808,
+      "loss": 2.1555,
+      "step": 17410
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014896236430229746,
+      "loss": 2.0696,
+      "step": 17415
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014892710688358654,
+      "loss": 2.1831,
+      "step": 17420
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001488918414668115,
+      "loss": 2.1474,
+      "step": 17425
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014885656805773713,
+      "loss": 2.1339,
+      "step": 17430
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014882128666212963,
+      "loss": 2.0911,
+      "step": 17435
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014878599728575628,
+      "loss": 2.1583,
+      "step": 17440
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014875069993438588,
+      "loss": 2.2018,
+      "step": 17445
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014871539461378845,
+      "loss": 2.1151,
+      "step": 17450
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001486800813297353,
+      "loss": 2.1418,
+      "step": 17455
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014864476008799905,
+      "loss": 2.1619,
+      "step": 17460
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014860943089435363,
+      "loss": 2.1143,
+      "step": 17465
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014857409375457429,
+      "loss": 2.194,
+      "step": 17470
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001485387486744375,
+      "loss": 2.1391,
+      "step": 17475
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014850339565972113,
+      "loss": 2.0887,
+      "step": 17480
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014846803471620432,
+      "loss": 2.1965,
+      "step": 17485
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014843266584966744,
+      "loss": 2.1282,
+      "step": 17490
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014839728906589224,
+      "loss": 2.1653,
+      "step": 17495
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001483619043706617,
+      "loss": 2.1374,
+      "step": 17500
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014832651176976014,
+      "loss": 2.1438,
+      "step": 17505
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014829111126897311,
+      "loss": 2.1271,
+      "step": 17510
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014825570287408754,
+      "loss": 2.1277,
+      "step": 17515
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001482202865908916,
+      "loss": 2.1328,
+      "step": 17520
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014818486242517475,
+      "loss": 2.131,
+      "step": 17525
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014814943038272773,
+      "loss": 2.1002,
+      "step": 17530
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014811399046934257,
+      "loss": 2.1415,
+      "step": 17535
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014807854269081262,
+      "loss": 2.0984,
+      "step": 17540
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014804308705293248,
+      "loss": 2.1023,
+      "step": 17545
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014800762356149803,
+      "loss": 2.1606,
+      "step": 17550
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014797215222230652,
+      "loss": 2.1346,
+      "step": 17555
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014793667304115632,
+      "loss": 2.1183,
+      "step": 17560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014790118602384724,
+      "loss": 2.1487,
+      "step": 17565
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001478656911761803,
+      "loss": 2.165,
+      "step": 17570
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014783018850395772,
+      "loss": 2.1233,
+      "step": 17575
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014779467801298322,
+      "loss": 2.1649,
+      "step": 17580
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001477591597090616,
+      "loss": 2.1833,
+      "step": 17585
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014772363359799899,
+      "loss": 2.0814,
+      "step": 17590
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014768809968560283,
+      "loss": 2.1567,
+      "step": 17595
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001476525579776818,
+      "loss": 2.1612,
+      "step": 17600
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014761700848004585,
+      "loss": 2.0892,
+      "step": 17605
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014758145119850627,
+      "loss": 2.1573,
+      "step": 17610
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001475458861388755,
+      "loss": 2.1598,
+      "step": 17615
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001475103133069674,
+      "loss": 2.0964,
+      "step": 17620
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014747473270859698,
+      "loss": 2.1373,
+      "step": 17625
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014743914434958057,
+      "loss": 2.0717,
+      "step": 17630
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014740354823573578,
+      "loss": 2.1386,
+      "step": 17635
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014736794437288144,
+      "loss": 2.1461,
+      "step": 17640
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014733233276683768,
+      "loss": 2.1017,
+      "step": 17645
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014729671342342593,
+      "loss": 2.1102,
+      "step": 17650
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001472610863484688,
+      "loss": 2.1308,
+      "step": 17655
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014722545154779025,
+      "loss": 2.1126,
+      "step": 17660
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014718980902721546,
+      "loss": 2.1595,
+      "step": 17665
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014715415879257084,
+      "loss": 2.1354,
+      "step": 17670
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014711850084968414,
+      "loss": 2.1141,
+      "step": 17675
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014708283520438432,
+      "loss": 2.1036,
+      "step": 17680
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014704716186250157,
+      "loss": 2.0774,
+      "step": 17685
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001470114808298674,
+      "loss": 2.1624,
+      "step": 17690
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001469757921123146,
+      "loss": 2.0873,
+      "step": 17695
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00014694009571567708,
+      "loss": 2.0946,
+      "step": 17700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014690439164579015,
+      "loss": 2.1503,
+      "step": 17705
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014686867990849033,
+      "loss": 2.1587,
+      "step": 17710
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014683296050961532,
+      "loss": 2.1649,
+      "step": 17715
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014679723345500419,
+      "loss": 2.1265,
+      "step": 17720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014676149875049717,
+      "loss": 2.1148,
+      "step": 17725
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00014672575640193577,
+      "loss": 2.1268,
+      "step": 17730
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001466900064151628,
+      "loss": 2.1062,
+      "step": 17735
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014665424879602222,
+      "loss": 2.1252,
+      "step": 17740
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014661848355035934,
+      "loss": 2.171,
+      "step": 17745
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014658271068402064,
+      "loss": 2.1173,
+      "step": 17750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014654693020285387,
+      "loss": 2.1237,
+      "step": 17755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014651114211270807,
+      "loss": 2.1405,
+      "step": 17760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014647534641943345,
+      "loss": 2.1143,
+      "step": 17765
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014643954312888147,
+      "loss": 2.17,
+      "step": 17770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001464037322469049,
+      "loss": 2.0799,
+      "step": 17775
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001463679137793577,
+      "loss": 2.1427,
+      "step": 17780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014633208773209505,
+      "loss": 2.1056,
+      "step": 17785
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014629625411097342,
+      "loss": 2.1379,
+      "step": 17790
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014626041292185052,
+      "loss": 2.1778,
+      "step": 17795
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014622456417058522,
+      "loss": 2.1297,
+      "step": 17800
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014618870786303772,
+      "loss": 2.1437,
+      "step": 17805
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001461528440050694,
+      "loss": 2.1287,
+      "step": 17810
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014611697260254286,
+      "loss": 2.1958,
+      "step": 17815
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014608109366132202,
+      "loss": 2.1225,
+      "step": 17820
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014604520718727195,
+      "loss": 2.1412,
+      "step": 17825
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014600931318625894,
+      "loss": 2.1718,
+      "step": 17830
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014597341166415058,
+      "loss": 2.1472,
+      "step": 17835
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00014593750262681564,
+      "loss": 2.1568,
+      "step": 17840
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014590158608012417,
+      "loss": 2.1441,
+      "step": 17845
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014586566202994737,
+      "loss": 2.1018,
+      "step": 17850
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014582973048215772,
+      "loss": 2.1024,
+      "step": 17855
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001457937914426289,
+      "loss": 2.1582,
+      "step": 17860
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014575784491723586,
+      "loss": 2.1096,
+      "step": 17865
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001457218909118547,
+      "loss": 2.1492,
+      "step": 17870
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014568592943236283,
+      "loss": 2.1585,
+      "step": 17875
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014564996048463882,
+      "loss": 2.1156,
+      "step": 17880
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014561398407456246,
+      "loss": 2.1076,
+      "step": 17885
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001455780002080148,
+      "loss": 2.1032,
+      "step": 17890
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014554200889087807,
+      "loss": 2.1752,
+      "step": 17895
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014550601012903577,
+      "loss": 2.144,
+      "step": 17900
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001454700039283725,
+      "loss": 2.1089,
+      "step": 17905
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001454339902947742,
+      "loss": 2.1479,
+      "step": 17910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014539796923412803,
+      "loss": 2.1595,
+      "step": 17915
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014536194075232223,
+      "loss": 2.1336,
+      "step": 17920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001453259048552464,
+      "loss": 2.18,
+      "step": 17925
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001452898615487913,
+      "loss": 2.1309,
+      "step": 17930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014525381083884884,
+      "loss": 2.1698,
+      "step": 17935
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001452177527313122,
+      "loss": 2.1038,
+      "step": 17940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001451816872320758,
+      "loss": 2.1128,
+      "step": 17945
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014514561434703525,
+      "loss": 2.1661,
+      "step": 17950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014510953408208724,
+      "loss": 2.1262,
+      "step": 17955
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001450734464431299,
+      "loss": 2.0253,
+      "step": 17960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014503735143606234,
+      "loss": 2.1417,
+      "step": 17965
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014500124906678506,
+      "loss": 2.1169,
+      "step": 17970
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001449651393411996,
+      "loss": 2.1072,
+      "step": 17975
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014492902226520882,
+      "loss": 2.1844,
+      "step": 17980
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014489289784471677,
+      "loss": 2.1351,
+      "step": 17985
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001448567660856286,
+      "loss": 2.1751,
+      "step": 17990
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014482062699385076,
+      "loss": 2.1294,
+      "step": 17995
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001447844805752909,
+      "loss": 2.1737,
+      "step": 18000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014474832683585782,
+      "loss": 2.1656,
+      "step": 18005
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001447121657814615,
+      "loss": 2.1231,
+      "step": 18010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001446759974180132,
+      "loss": 2.174,
+      "step": 18015
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014463982175142532,
+      "loss": 2.1395,
+      "step": 18020
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014460363878761144,
+      "loss": 2.1145,
+      "step": 18025
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014456744853248637,
+      "loss": 2.1589,
+      "step": 18030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014453125099196608,
+      "loss": 2.0924,
+      "step": 18035
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014449504617196772,
+      "loss": 2.2053,
+      "step": 18040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001444588340784097,
+      "loss": 2.1145,
+      "step": 18045
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014442261471721155,
+      "loss": 2.1126,
+      "step": 18050
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014438638809429404,
+      "loss": 2.1451,
+      "step": 18055
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014435015421557904,
+      "loss": 2.1758,
+      "step": 18060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014431391308698975,
+      "loss": 2.0972,
+      "step": 18065
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014427766471445042,
+      "loss": 2.1583,
+      "step": 18070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001442414091038865,
+      "loss": 2.1019,
+      "step": 18075
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001442051462612247,
+      "loss": 2.1201,
+      "step": 18080
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001441688761923929,
+      "loss": 2.1314,
+      "step": 18085
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014413259890332008,
+      "loss": 2.1151,
+      "step": 18090
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014409631439993645,
+      "loss": 2.1584,
+      "step": 18095
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014406002268817342,
+      "loss": 2.1462,
+      "step": 18100
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014402372377396355,
+      "loss": 2.1468,
+      "step": 18105
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00014398741766324062,
+      "loss": 2.1693,
+      "step": 18110
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014395110436193952,
+      "loss": 2.2293,
+      "step": 18115
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014391478387599634,
+      "loss": 2.0769,
+      "step": 18120
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014387845621134838,
+      "loss": 2.1946,
+      "step": 18125
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014384212137393408,
+      "loss": 2.0896,
+      "step": 18130
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014380577936969302,
+      "loss": 2.1447,
+      "step": 18135
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014376943020456606,
+      "loss": 2.1179,
+      "step": 18140
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001437330738844951,
+      "loss": 2.1637,
+      "step": 18145
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001436967104154233,
+      "loss": 2.1938,
+      "step": 18150
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014366033980329495,
+      "loss": 2.1588,
+      "step": 18155
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014362396205405554,
+      "loss": 2.1256,
+      "step": 18160
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014358757717365165,
+      "loss": 2.0774,
+      "step": 18165
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014355118516803112,
+      "loss": 2.1461,
+      "step": 18170
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014351478604314292,
+      "loss": 2.094,
+      "step": 18175
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014347837980493715,
+      "loss": 2.126,
+      "step": 18180
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014344196645936513,
+      "loss": 2.1581,
+      "step": 18185
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014340554601237929,
+      "loss": 2.1106,
+      "step": 18190
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001433691184699332,
+      "loss": 2.1373,
+      "step": 18195
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014333268383798174,
+      "loss": 2.1129,
+      "step": 18200
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001432962421224808,
+      "loss": 2.1128,
+      "step": 18205
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001432597933293874,
+      "loss": 2.1469,
+      "step": 18210
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014322333746465986,
+      "loss": 2.0915,
+      "step": 18215
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014318687453425757,
+      "loss": 2.1177,
+      "step": 18220
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001431504045441411,
+      "loss": 2.1116,
+      "step": 18225
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001431139275002721,
+      "loss": 2.1185,
+      "step": 18230
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001430774434086135,
+      "loss": 2.1525,
+      "step": 18235
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014304095227512932,
+      "loss": 2.1048,
+      "step": 18240
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014300445410578467,
+      "loss": 2.1696,
+      "step": 18245
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014296794890654592,
+      "loss": 2.1575,
+      "step": 18250
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014293143668338054,
+      "loss": 2.1283,
+      "step": 18255
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014289491744225708,
+      "loss": 2.1898,
+      "step": 18260
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014285839118914541,
+      "loss": 2.1295,
+      "step": 18265
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014282185793001632,
+      "loss": 2.1318,
+      "step": 18270
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00014278531767084195,
+      "loss": 2.1372,
+      "step": 18275
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00014274877041759547,
+      "loss": 2.1081,
+      "step": 18280
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014271221617625124,
+      "loss": 2.1656,
+      "step": 18285
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001426756549527847,
+      "loss": 2.0975,
+      "step": 18290
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014263908675317253,
+      "loss": 2.1607,
+      "step": 18295
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00014260251158339243,
+      "loss": 2.1198,
+      "step": 18300
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014256592944942337,
+      "loss": 2.1396,
+      "step": 18305
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014252934035724536,
+      "loss": 2.1296,
+      "step": 18310
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001424927443128396,
+      "loss": 2.175,
+      "step": 18315
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014245614132218844,
+      "loss": 2.1299,
+      "step": 18320
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014241953139127524,
+      "loss": 2.1673,
+      "step": 18325
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014238291452608464,
+      "loss": 2.1416,
+      "step": 18330
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014234629073260236,
+      "loss": 2.1891,
+      "step": 18335
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014230966001681526,
+      "loss": 2.0653,
+      "step": 18340
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014227302238471138,
+      "loss": 2.0972,
+      "step": 18345
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014223637784227972,
+      "loss": 2.0956,
+      "step": 18350
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001421997263955106,
+      "loss": 2.117,
+      "step": 18355
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001421630680503954,
+      "loss": 2.1482,
+      "step": 18360
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014212640281292658,
+      "loss": 2.0616,
+      "step": 18365
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014208973068909777,
+      "loss": 2.0873,
+      "step": 18370
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014205305168490378,
+      "loss": 2.1583,
+      "step": 18375
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014201636580634042,
+      "loss": 2.1349,
+      "step": 18380
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014197967305940474,
+      "loss": 2.1423,
+      "step": 18385
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014194297345009486,
+      "loss": 2.1405,
+      "step": 18390
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014190626698441,
+      "loss": 2.1201,
+      "step": 18395
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014186955366835057,
+      "loss": 2.174,
+      "step": 18400
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014183283350791803,
+      "loss": 2.1312,
+      "step": 18405
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000141796106509115,
+      "loss": 2.1615,
+      "step": 18410
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014175937267794518,
+      "loss": 2.1472,
+      "step": 18415
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014172263202041343,
+      "loss": 2.152,
+      "step": 18420
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014168588454252572,
+      "loss": 2.1099,
+      "step": 18425
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001416491302502891,
+      "loss": 2.1528,
+      "step": 18430
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001416123691497118,
+      "loss": 2.1429,
+      "step": 18435
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014157560124680306,
+      "loss": 2.0892,
+      "step": 18440
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014153882654757333,
+      "loss": 2.1214,
+      "step": 18445
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001415020450580341,
+      "loss": 2.0961,
+      "step": 18450
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014146525678419803,
+      "loss": 2.1167,
+      "step": 18455
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014142846173207887,
+      "loss": 2.1283,
+      "step": 18460
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014139165990769144,
+      "loss": 2.1717,
+      "step": 18465
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014135485131705174,
+      "loss": 2.1866,
+      "step": 18470
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014131803596617676,
+      "loss": 2.1195,
+      "step": 18475
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001412812138610847,
+      "loss": 2.1441,
+      "step": 18480
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014124438500779488,
+      "loss": 2.1484,
+      "step": 18485
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014120754941232764,
+      "loss": 2.1574,
+      "step": 18490
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014117070708070443,
+      "loss": 2.1315,
+      "step": 18495
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014113385801894788,
+      "loss": 2.1214,
+      "step": 18500
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014109700223308163,
+      "loss": 2.1536,
+      "step": 18505
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014106013972913046,
+      "loss": 2.0965,
+      "step": 18510
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014102327051312025,
+      "loss": 2.1613,
+      "step": 18515
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000140986394591078,
+      "loss": 2.1011,
+      "step": 18520
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014094951196903178,
+      "loss": 2.119,
+      "step": 18525
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014091262265301068,
+      "loss": 2.1043,
+      "step": 18530
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014087572664904502,
+      "loss": 2.1256,
+      "step": 18535
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00014083882396316618,
+      "loss": 2.1195,
+      "step": 18540
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00014080191460140651,
+      "loss": 2.1485,
+      "step": 18545
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014076499856979964,
+      "loss": 2.1671,
+      "step": 18550
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014072807587438011,
+      "loss": 2.225,
+      "step": 18555
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014069114652118375,
+      "loss": 2.1267,
+      "step": 18560
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014065421051624727,
+      "loss": 2.1248,
+      "step": 18565
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014061726786560857,
+      "loss": 2.1245,
+      "step": 18570
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001405803185753067,
+      "loss": 2.17,
+      "step": 18575
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001405433626513816,
+      "loss": 2.1515,
+      "step": 18580
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00014050640009987455,
+      "loss": 2.1942,
+      "step": 18585
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014046943092682767,
+      "loss": 2.1103,
+      "step": 18590
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014043245513828433,
+      "loss": 2.1156,
+      "step": 18595
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014039547274028892,
+      "loss": 2.1556,
+      "step": 18600
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014035848373888688,
+      "loss": 2.1019,
+      "step": 18605
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00014032148814012484,
+      "loss": 2.1266,
+      "step": 18610
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00014028448595005037,
+      "loss": 2.1237,
+      "step": 18615
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001402474771747122,
+      "loss": 2.1442,
+      "step": 18620
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001402104618201601,
+      "loss": 2.1854,
+      "step": 18625
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00014017343989244493,
+      "loss": 2.1286,
+      "step": 18630
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00014013641139761865,
+      "loss": 2.203,
+      "step": 18635
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00014009937634173425,
+      "loss": 2.1074,
+      "step": 18640
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014006233473084584,
+      "loss": 2.1772,
+      "step": 18645
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00014002528657100854,
+      "loss": 2.1478,
+      "step": 18650
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001399882318682786,
+      "loss": 2.1826,
+      "step": 18655
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013995117062871334,
+      "loss": 2.1307,
+      "step": 18660
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013991410285837103,
+      "loss": 2.1414,
+      "step": 18665
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001398770285633112,
+      "loss": 2.0938,
+      "step": 18670
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001398399477495943,
+      "loss": 2.1185,
+      "step": 18675
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001398028604232819,
+      "loss": 2.1683,
+      "step": 18680
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001397657665904366,
+      "loss": 2.1914,
+      "step": 18685
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013972866625712216,
+      "loss": 2.09,
+      "step": 18690
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013969155942940328,
+      "loss": 2.1402,
+      "step": 18695
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013965444611334577,
+      "loss": 2.1279,
+      "step": 18700
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013961732631501654,
+      "loss": 2.0843,
+      "step": 18705
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001395802000404835,
+      "loss": 2.1244,
+      "step": 18710
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013954306729581565,
+      "loss": 2.0847,
+      "step": 18715
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013950592808708304,
+      "loss": 2.1157,
+      "step": 18720
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001394687824203568,
+      "loss": 2.1388,
+      "step": 18725
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013943163030170905,
+      "loss": 2.1318,
+      "step": 18730
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013939447173721305,
+      "loss": 2.1291,
+      "step": 18735
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013935730673294306,
+      "loss": 2.1605,
+      "step": 18740
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013932013529497438,
+      "loss": 2.16,
+      "step": 18745
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00013928295742938343,
+      "loss": 2.0951,
+      "step": 18750
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013924577314224762,
+      "loss": 2.0896,
+      "step": 18755
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001392085824396454,
+      "loss": 2.1957,
+      "step": 18760
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013917138532765634,
+      "loss": 2.1927,
+      "step": 18765
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.000139134181812361,
+      "loss": 2.1274,
+      "step": 18770
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000139096971899841,
+      "loss": 2.155,
+      "step": 18775
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000139059755596179,
+      "loss": 2.1388,
+      "step": 18780
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013902253290745868,
+      "loss": 2.0999,
+      "step": 18785
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013898530383976487,
+      "loss": 2.1429,
+      "step": 18790
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001389480683991833,
+      "loss": 2.1252,
+      "step": 18795
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013891082659180085,
+      "loss": 2.146,
+      "step": 18800
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013887357842370536,
+      "loss": 2.1113,
+      "step": 18805
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001388363239009858,
+      "loss": 2.1795,
+      "step": 18810
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001387990630297321,
+      "loss": 2.2264,
+      "step": 18815
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013876179581603524,
+      "loss": 2.164,
+      "step": 18820
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001387245222659873,
+      "loss": 2.184,
+      "step": 18825
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013868724238568132,
+      "loss": 2.1095,
+      "step": 18830
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001386499561812114,
+      "loss": 2.0739,
+      "step": 18835
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013861266365867267,
+      "loss": 2.1039,
+      "step": 18840
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013857536482416133,
+      "loss": 2.0955,
+      "step": 18845
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013853805968377456,
+      "loss": 2.1011,
+      "step": 18850
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013850074824361058,
+      "loss": 2.1503,
+      "step": 18855
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013846343050976872,
+      "loss": 2.1283,
+      "step": 18860
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001384261064883492,
+      "loss": 2.1539,
+      "step": 18865
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013838877618545334,
+      "loss": 2.0997,
+      "step": 18870
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013835143960718352,
+      "loss": 2.0809,
+      "step": 18875
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013831409675964314,
+      "loss": 2.1481,
+      "step": 18880
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013827674764893652,
+      "loss": 2.0724,
+      "step": 18885
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013823939228116915,
+      "loss": 2.2133,
+      "step": 18890
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013820203066244746,
+      "loss": 2.0823,
+      "step": 18895
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013816466279887885,
+      "loss": 2.1469,
+      "step": 18900
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013812728869657187,
+      "loss": 2.1243,
+      "step": 18905
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013808990836163606,
+      "loss": 2.1275,
+      "step": 18910
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013805252180018184,
+      "loss": 2.0938,
+      "step": 18915
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013801512901832083,
+      "loss": 2.1906,
+      "step": 18920
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013797773002216556,
+      "loss": 2.1671,
+      "step": 18925
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013794032481782964,
+      "loss": 2.1503,
+      "step": 18930
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013790291341142763,
+      "loss": 2.1504,
+      "step": 18935
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013786549580907512,
+      "loss": 2.1334,
+      "step": 18940
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013782807201688876,
+      "loss": 2.1571,
+      "step": 18945
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001377906420409862,
+      "loss": 2.1515,
+      "step": 18950
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000137753205887486,
+      "loss": 2.1205,
+      "step": 18955
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013771576356250788,
+      "loss": 2.04,
+      "step": 18960
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013767831507217245,
+      "loss": 2.0821,
+      "step": 18965
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013764086042260143,
+      "loss": 2.1129,
+      "step": 18970
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013760339961991746,
+      "loss": 2.1195,
+      "step": 18975
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013756593267024423,
+      "loss": 2.1444,
+      "step": 18980
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013752845957970645,
+      "loss": 2.1312,
+      "step": 18985
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001374909803544297,
+      "loss": 2.1442,
+      "step": 18990
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001374534950005408,
+      "loss": 2.097,
+      "step": 18995
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013741600352416738,
+      "loss": 2.0889,
+      "step": 19000
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013737850593143814,
+      "loss": 2.147,
+      "step": 19005
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013734100222848275,
+      "loss": 2.1237,
+      "step": 19010
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013730349242143194,
+      "loss": 2.1137,
+      "step": 19015
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001372659765164174,
+      "loss": 2.1834,
+      "step": 19020
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001372284545195718,
+      "loss": 2.1277,
+      "step": 19025
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013719092643702882,
+      "loss": 2.17,
+      "step": 19030
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013715339227492318,
+      "loss": 2.1572,
+      "step": 19035
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013711585203939045,
+      "loss": 2.174,
+      "step": 19040
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001370783057365674,
+      "loss": 2.1887,
+      "step": 19045
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013704075337259164,
+      "loss": 2.2354,
+      "step": 19050
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001370031949536018,
+      "loss": 2.1228,
+      "step": 19055
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013696563048573755,
+      "loss": 2.1007,
+      "step": 19060
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00013692805997513952,
+      "loss": 2.0853,
+      "step": 19065
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00013689048342794928,
+      "loss": 2.0543,
+      "step": 19070
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001368529008503095,
+      "loss": 2.1306,
+      "step": 19075
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013681531224836365,
+      "loss": 2.1398,
+      "step": 19080
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013677771762825642,
+      "loss": 2.1512,
+      "step": 19085
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001367401169961333,
+      "loss": 2.1696,
+      "step": 19090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013670251035814083,
+      "loss": 2.1518,
+      "step": 19095
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00013666489772042658,
+      "loss": 2.128,
+      "step": 19100
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013662727908913898,
+      "loss": 2.1166,
+      "step": 19105
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013658965447042754,
+      "loss": 2.1131,
+      "step": 19110
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001365520238704427,
+      "loss": 2.1,
+      "step": 19115
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013651438729533593,
+      "loss": 2.0881,
+      "step": 19120
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013647674475125962,
+      "loss": 2.1344,
+      "step": 19125
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013643909624436716,
+      "loss": 2.186,
+      "step": 19130
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013640144178081297,
+      "loss": 2.1642,
+      "step": 19135
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013636378136675225,
+      "loss": 2.1005,
+      "step": 19140
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013632611500834143,
+      "loss": 2.1225,
+      "step": 19145
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013628844271173775,
+      "loss": 2.19,
+      "step": 19150
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013625076448309947,
+      "loss": 2.1405,
+      "step": 19155
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013621308032858584,
+      "loss": 2.1016,
+      "step": 19160
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013617539025435698,
+      "loss": 2.1394,
+      "step": 19165
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013613769426657407,
+      "loss": 2.1067,
+      "step": 19170
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013609999237139926,
+      "loss": 2.1597,
+      "step": 19175
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013606228457499565,
+      "loss": 2.1135,
+      "step": 19180
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013602457088352724,
+      "loss": 2.0879,
+      "step": 19185
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013598685130315912,
+      "loss": 2.1152,
+      "step": 19190
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001359491258400572,
+      "loss": 2.135,
+      "step": 19195
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013591139450038845,
+      "loss": 2.1172,
+      "step": 19200
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001358736572903208,
+      "loss": 2.1152,
+      "step": 19205
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013583591421602312,
+      "loss": 2.1671,
+      "step": 19210
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013579816528366516,
+      "loss": 2.1183,
+      "step": 19215
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013576041049941776,
+      "loss": 2.0937,
+      "step": 19220
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013572264986945264,
+      "loss": 2.1799,
+      "step": 19225
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001356848833999425,
+      "loss": 2.1614,
+      "step": 19230
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013564711109706096,
+      "loss": 2.1817,
+      "step": 19235
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013560933296698265,
+      "loss": 2.1365,
+      "step": 19240
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013557154901588312,
+      "loss": 2.1541,
+      "step": 19245
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013553375924993887,
+      "loss": 2.0992,
+      "step": 19250
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013549596367532732,
+      "loss": 2.1337,
+      "step": 19255
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013545816229822692,
+      "loss": 2.1494,
+      "step": 19260
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000135420355124817,
+      "loss": 2.1015,
+      "step": 19265
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013538254216127787,
+      "loss": 2.1216,
+      "step": 19270
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013534472341379075,
+      "loss": 2.2029,
+      "step": 19275
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001353068988885379,
+      "loss": 2.0979,
+      "step": 19280
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013526906859170237,
+      "loss": 2.1816,
+      "step": 19285
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001352312325294683,
+      "loss": 2.1037,
+      "step": 19290
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013519339070802063,
+      "loss": 2.202,
+      "step": 19295
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013515554313354544,
+      "loss": 2.1933,
+      "step": 19300
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013511768981222958,
+      "loss": 2.1679,
+      "step": 19305
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013507983075026087,
+      "loss": 2.0753,
+      "step": 19310
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013504196595382815,
+      "loss": 2.1353,
+      "step": 19315
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013500409542912108,
+      "loss": 2.138,
+      "step": 19320
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013496621918233033,
+      "loss": 2.1788,
+      "step": 19325
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013492833721964753,
+      "loss": 2.1243,
+      "step": 19330
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013489044954726514,
+      "loss": 2.1545,
+      "step": 19335
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013485255617137672,
+      "loss": 2.1068,
+      "step": 19340
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013481465709817657,
+      "loss": 2.1274,
+      "step": 19345
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013477675233386006,
+      "loss": 2.1071,
+      "step": 19350
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013473884188462337,
+      "loss": 2.0999,
+      "step": 19355
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013470092575666379,
+      "loss": 2.1294,
+      "step": 19360
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013466300395617933,
+      "loss": 2.149,
+      "step": 19365
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001346250764893691,
+      "loss": 2.1164,
+      "step": 19370
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013458714336243307,
+      "loss": 2.1001,
+      "step": 19375
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013454920458157206,
+      "loss": 2.1134,
+      "step": 19380
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001345112601529879,
+      "loss": 2.1605,
+      "step": 19385
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013447331008288339,
+      "loss": 2.155,
+      "step": 19390
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013443535437746216,
+      "loss": 2.1357,
+      "step": 19395
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00013439739304292876,
+      "loss": 2.1306,
+      "step": 19400
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013435942608548872,
+      "loss": 2.091,
+      "step": 19405
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013432145351134842,
+      "loss": 2.1536,
+      "step": 19410
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013428347532671523,
+      "loss": 2.1029,
+      "step": 19415
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013424549153779743,
+      "loss": 2.1519,
+      "step": 19420
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013420750215080413,
+      "loss": 2.1472,
+      "step": 19425
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013416950717194548,
+      "loss": 2.1242,
+      "step": 19430
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013413150660743244,
+      "loss": 2.1436,
+      "step": 19435
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013409350046347695,
+      "loss": 2.1165,
+      "step": 19440
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013405548874629179,
+      "loss": 2.1009,
+      "step": 19445
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013401747146209075,
+      "loss": 2.1448,
+      "step": 19450
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001339794486170884,
+      "loss": 2.1087,
+      "step": 19455
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013394142021750041,
+      "loss": 2.0922,
+      "step": 19460
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013390338626954315,
+      "loss": 2.1292,
+      "step": 19465
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013386534677943402,
+      "loss": 2.1025,
+      "step": 19470
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001338273017533913,
+      "loss": 2.1173,
+      "step": 19475
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001337892511976342,
+      "loss": 2.1372,
+      "step": 19480
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013375119511838273,
+      "loss": 2.1515,
+      "step": 19485
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013371313352185794,
+      "loss": 2.2141,
+      "step": 19490
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013367506641428172,
+      "loss": 2.1261,
+      "step": 19495
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001336369938018768,
+      "loss": 2.1249,
+      "step": 19500
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013359891569086695,
+      "loss": 2.1596,
+      "step": 19505
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001335608320874767,
+      "loss": 2.1336,
+      "step": 19510
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013352274299793159,
+      "loss": 2.1581,
+      "step": 19515
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000133484648428458,
+      "loss": 2.1553,
+      "step": 19520
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013344654838528316,
+      "loss": 2.0643,
+      "step": 19525
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013340844287463525,
+      "loss": 2.144,
+      "step": 19530
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013337033190274339,
+      "loss": 2.1049,
+      "step": 19535
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001333322154758375,
+      "loss": 2.1189,
+      "step": 19540
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013329409360014847,
+      "loss": 2.1325,
+      "step": 19545
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.000133255966281908,
+      "loss": 2.1432,
+      "step": 19550
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013321783352734874,
+      "loss": 2.1925,
+      "step": 19555
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001331796953427042,
+      "loss": 2.1479,
+      "step": 19560
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013314155173420886,
+      "loss": 2.1241,
+      "step": 19565
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013310340270809794,
+      "loss": 2.1774,
+      "step": 19570
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013306524827060763,
+      "loss": 2.1393,
+      "step": 19575
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000133027088427975,
+      "loss": 2.0714,
+      "step": 19580
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013298892318643806,
+      "loss": 2.1239,
+      "step": 19585
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013295075255223557,
+      "loss": 2.1332,
+      "step": 19590
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013291257653160726,
+      "loss": 2.0828,
+      "step": 19595
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013287439513079375,
+      "loss": 2.1069,
+      "step": 19600
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013283620835603653,
+      "loss": 2.1386,
+      "step": 19605
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013279801621357787,
+      "loss": 2.1372,
+      "step": 19610
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001327598187096611,
+      "loss": 2.1486,
+      "step": 19615
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00013272161585053027,
+      "loss": 2.0606,
+      "step": 19620
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013268340764243034,
+      "loss": 2.11,
+      "step": 19625
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013264519409160724,
+      "loss": 2.1153,
+      "step": 19630
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013260697520430762,
+      "loss": 2.1626,
+      "step": 19635
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013256875098677915,
+      "loss": 2.0846,
+      "step": 19640
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013253052144527024,
+      "loss": 2.1661,
+      "step": 19645
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001324922865860303,
+      "loss": 2.1519,
+      "step": 19650
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001324540464153095,
+      "loss": 2.1744,
+      "step": 19655
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013241580093935896,
+      "loss": 2.1451,
+      "step": 19660
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001323775501644306,
+      "loss": 2.1365,
+      "step": 19665
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001323392940967772,
+      "loss": 2.1487,
+      "step": 19670
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001323010327426525,
+      "loss": 2.1587,
+      "step": 19675
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013226276610831103,
+      "loss": 2.1653,
+      "step": 19680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001322244942000082,
+      "loss": 2.121,
+      "step": 19685
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013218621702400024,
+      "loss": 2.1667,
+      "step": 19690
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013214793458654432,
+      "loss": 2.1096,
+      "step": 19695
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013210964689389844,
+      "loss": 2.1076,
+      "step": 19700
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013207135395232144,
+      "loss": 2.0877,
+      "step": 19705
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000132033055768073,
+      "loss": 2.1231,
+      "step": 19710
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013199475234741372,
+      "loss": 2.1244,
+      "step": 19715
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013195644369660498,
+      "loss": 2.1716,
+      "step": 19720
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013191812982190913,
+      "loss": 2.1514,
+      "step": 19725
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001318798107295892,
+      "loss": 2.1812,
+      "step": 19730
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013184148642590924,
+      "loss": 2.1533,
+      "step": 19735
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013180315691713407,
+      "loss": 2.1549,
+      "step": 19740
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013176482220952936,
+      "loss": 2.093,
+      "step": 19745
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001317264823093617,
+      "loss": 2.1172,
+      "step": 19750
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001316881372228984,
+      "loss": 2.2167,
+      "step": 19755
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013164978695640774,
+      "loss": 2.126,
+      "step": 19760
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013161143151615878,
+      "loss": 2.1308,
+      "step": 19765
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013157307090842145,
+      "loss": 2.0876,
+      "step": 19770
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00013153470513946653,
+      "loss": 2.0843,
+      "step": 19775
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001314963342155656,
+      "loss": 2.088,
+      "step": 19780
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013145795814299116,
+      "loss": 2.1018,
+      "step": 19785
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00013141957692801648,
+      "loss": 2.124,
+      "step": 19790
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001313811905769157,
+      "loss": 2.0651,
+      "step": 19795
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001313427990959638,
+      "loss": 2.0702,
+      "step": 19800
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001313044024914366,
+      "loss": 2.1177,
+      "step": 19805
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013126600076961075,
+      "loss": 2.1586,
+      "step": 19810
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013122759393676377,
+      "loss": 2.1429,
+      "step": 19815
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001311891819991739,
+      "loss": 2.1125,
+      "step": 19820
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001311507649631204,
+      "loss": 2.1173,
+      "step": 19825
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013111234283488322,
+      "loss": 2.0774,
+      "step": 19830
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013107391562074315,
+      "loss": 2.154,
+      "step": 19835
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001310354833269819,
+      "loss": 2.1226,
+      "step": 19840
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013099704595988197,
+      "loss": 2.1391,
+      "step": 19845
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001309586035257266,
+      "loss": 2.1382,
+      "step": 19850
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00013092015603080002,
+      "loss": 2.1179,
+      "step": 19855
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013088170348138714,
+      "loss": 2.1287,
+      "step": 19860
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013084324588377383,
+      "loss": 2.096,
+      "step": 19865
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013080478324424663,
+      "loss": 2.1587,
+      "step": 19870
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013076631556909303,
+      "loss": 2.1756,
+      "step": 19875
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00013072784286460133,
+      "loss": 2.1242,
+      "step": 19880
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013068936513706056,
+      "loss": 2.1048,
+      "step": 19885
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001306508823927607,
+      "loss": 2.0848,
+      "step": 19890
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013061239463799243,
+      "loss": 2.1253,
+      "step": 19895
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001305739018790473,
+      "loss": 2.1657,
+      "step": 19900
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013053540412221777,
+      "loss": 2.1419,
+      "step": 19905
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001304969013737969,
+      "loss": 2.1247,
+      "step": 19910
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001304583936400788,
+      "loss": 2.2078,
+      "step": 19915
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00013041988092735827,
+      "loss": 2.1664,
+      "step": 19920
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001303813632419309,
+      "loss": 2.1679,
+      "step": 19925
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00013034284059009317,
+      "loss": 2.1655,
+      "step": 19930
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013030431297814232,
+      "loss": 2.1377,
+      "step": 19935
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00013026578041237642,
+      "loss": 2.1956,
+      "step": 19940
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00013022724289909438,
+      "loss": 2.1473,
+      "step": 19945
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013018870044459587,
+      "loss": 2.153,
+      "step": 19950
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013015015305518137,
+      "loss": 2.1606,
+      "step": 19955
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00013011160073715222,
+      "loss": 2.0908,
+      "step": 19960
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00013007304349681048,
+      "loss": 2.1431,
+      "step": 19965
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001300344813404591,
+      "loss": 2.1421,
+      "step": 19970
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012999591427440176,
+      "loss": 2.1074,
+      "step": 19975
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012995734230494303,
+      "loss": 2.0893,
+      "step": 19980
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012991876543838818,
+      "loss": 2.0928,
+      "step": 19985
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012988018368104335,
+      "loss": 2.1944,
+      "step": 19990
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001298415970392155,
+      "loss": 2.1463,
+      "step": 19995
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012980300551921226,
+      "loss": 2.1682,
+      "step": 20000
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012976440912734223,
+      "loss": 2.1495,
+      "step": 20005
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012972580786991468,
+      "loss": 2.115,
+      "step": 20010
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001296872017532397,
+      "loss": 2.1464,
+      "step": 20015
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00012964859078362827,
+      "loss": 2.1557,
+      "step": 20020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012960997496739202,
+      "loss": 2.1202,
+      "step": 20025
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012957135431084345,
+      "loss": 2.1262,
+      "step": 20030
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012953272882029583,
+      "loss": 2.1279,
+      "step": 20035
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012949409850206327,
+      "loss": 2.1489,
+      "step": 20040
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001294554633624606,
+      "loss": 2.1405,
+      "step": 20045
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012941682340780343,
+      "loss": 2.1074,
+      "step": 20050
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001293781786444083,
+      "loss": 2.1011,
+      "step": 20055
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012933952907859231,
+      "loss": 2.1791,
+      "step": 20060
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012930087471667352,
+      "loss": 2.0836,
+      "step": 20065
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001292622155649707,
+      "loss": 2.1091,
+      "step": 20070
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012922355162980347,
+      "loss": 2.1684,
+      "step": 20075
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012918488291749214,
+      "loss": 2.1353,
+      "step": 20080
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001291462094343579,
+      "loss": 2.093,
+      "step": 20085
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012910753118672256,
+      "loss": 2.1379,
+      "step": 20090
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001290688481809089,
+      "loss": 2.1185,
+      "step": 20095
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012903016042324037,
+      "loss": 2.1045,
+      "step": 20100
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012899146792004121,
+      "loss": 2.1386,
+      "step": 20105
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012895277067763648,
+      "loss": 2.1408,
+      "step": 20110
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001289140687023519,
+      "loss": 2.1734,
+      "step": 20115
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012887536200051412,
+      "loss": 2.1097,
+      "step": 20120
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00012883665057845042,
+      "loss": 2.1093,
+      "step": 20125
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012879793444248898,
+      "loss": 2.0777,
+      "step": 20130
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012875921359895866,
+      "loss": 2.0842,
+      "step": 20135
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001287204880541891,
+      "loss": 2.1662,
+      "step": 20140
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012868175781451075,
+      "loss": 2.1337,
+      "step": 20145
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012864302288625478,
+      "loss": 2.108,
+      "step": 20150
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001286042832757532,
+      "loss": 2.2093,
+      "step": 20155
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012856553898933868,
+      "loss": 2.1508,
+      "step": 20160
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012852679003334472,
+      "loss": 2.1002,
+      "step": 20165
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012848803641410557,
+      "loss": 2.1426,
+      "step": 20170
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012844927813795623,
+      "loss": 2.1393,
+      "step": 20175
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012841051521123254,
+      "loss": 2.2157,
+      "step": 20180
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000128371747640271,
+      "loss": 2.1597,
+      "step": 20185
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012833297543140884,
+      "loss": 2.1329,
+      "step": 20190
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012829419859098422,
+      "loss": 2.1681,
+      "step": 20195
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001282554171253359,
+      "loss": 2.1191,
+      "step": 20200
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001282166310408034,
+      "loss": 2.1039,
+      "step": 20205
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012817784034372716,
+      "loss": 2.1073,
+      "step": 20210
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012813904504044814,
+      "loss": 2.1218,
+      "step": 20215
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012810024513730826,
+      "loss": 2.1528,
+      "step": 20220
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012806144064065,
+      "loss": 2.1828,
+      "step": 20225
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001280226315568168,
+      "loss": 2.0558,
+      "step": 20230
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012798381789215263,
+      "loss": 2.1371,
+      "step": 20235
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012794499965300242,
+      "loss": 2.1177,
+      "step": 20240
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001279061768457117,
+      "loss": 2.134,
+      "step": 20245
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001278673494766268,
+      "loss": 2.1798,
+      "step": 20250
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001278285175520948,
+      "loss": 2.1333,
+      "step": 20255
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012778968107846346,
+      "loss": 2.1284,
+      "step": 20260
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012775084006208144,
+      "loss": 2.1047,
+      "step": 20265
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012771199450929796,
+      "loss": 2.1856,
+      "step": 20270
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001276731444264631,
+      "loss": 2.1651,
+      "step": 20275
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001276342898199276,
+      "loss": 2.185,
+      "step": 20280
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012759543069604303,
+      "loss": 2.1183,
+      "step": 20285
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012755656706116164,
+      "loss": 2.1086,
+      "step": 20290
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012751769892163642,
+      "loss": 2.1532,
+      "step": 20295
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012747882628382114,
+      "loss": 2.1515,
+      "step": 20300
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012743994915407017,
+      "loss": 2.1962,
+      "step": 20305
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012740106753873887,
+      "loss": 2.0861,
+      "step": 20310
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012736218144418303,
+      "loss": 2.1262,
+      "step": 20315
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001273232908767594,
+      "loss": 2.1493,
+      "step": 20320
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012728439584282537,
+      "loss": 2.0743,
+      "step": 20325
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012724549634873905,
+      "loss": 2.1743,
+      "step": 20330
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012720659240085934,
+      "loss": 2.1682,
+      "step": 20335
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001271676840055458,
+      "loss": 2.1183,
+      "step": 20340
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012712877116915876,
+      "loss": 2.1032,
+      "step": 20345
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012708985389805924,
+      "loss": 2.212,
+      "step": 20350
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000127050932198609,
+      "loss": 2.1763,
+      "step": 20355
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012701200607717056,
+      "loss": 2.1104,
+      "step": 20360
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001269730755401071,
+      "loss": 2.0861,
+      "step": 20365
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001269341405937826,
+      "loss": 2.0952,
+      "step": 20370
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001268952012445617,
+      "loss": 2.1364,
+      "step": 20375
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012685625749880974,
+      "loss": 2.0656,
+      "step": 20380
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012681730936289288,
+      "loss": 2.1274,
+      "step": 20385
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012677835684317783,
+      "loss": 2.0776,
+      "step": 20390
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012673939994603225,
+      "loss": 2.1119,
+      "step": 20395
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012670043867782428,
+      "loss": 2.1514,
+      "step": 20400
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012666147304492292,
+      "loss": 2.1177,
+      "step": 20405
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012662250305369785,
+      "loss": 2.1169,
+      "step": 20410
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012658352871051945,
+      "loss": 2.1325,
+      "step": 20415
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012654455002175877,
+      "loss": 2.1661,
+      "step": 20420
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012650556699378768,
+      "loss": 2.1137,
+      "step": 20425
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012646657963297868,
+      "loss": 2.1709,
+      "step": 20430
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012642758794570498,
+      "loss": 2.1517,
+      "step": 20435
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001263885919383405,
+      "loss": 2.0768,
+      "step": 20440
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001263495916172599,
+      "loss": 2.1028,
+      "step": 20445
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001263105869888385,
+      "loss": 2.1622,
+      "step": 20450
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012627157805945237,
+      "loss": 2.1604,
+      "step": 20455
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012623256483547827,
+      "loss": 2.1291,
+      "step": 20460
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012619354732329362,
+      "loss": 2.0785,
+      "step": 20465
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012615452552927658,
+      "loss": 2.0937,
+      "step": 20470
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012611549945980603,
+      "loss": 2.1103,
+      "step": 20475
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012607646912126148,
+      "loss": 2.1761,
+      "step": 20480
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012603743452002322,
+      "loss": 2.1039,
+      "step": 20485
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00012599839566247213,
+      "loss": 2.1101,
+      "step": 20490
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012595935255498993,
+      "loss": 2.1153,
+      "step": 20495
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012592030520395894,
+      "loss": 2.1324,
+      "step": 20500
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012588125361576218,
+      "loss": 2.0979,
+      "step": 20505
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012584219779678335,
+      "loss": 2.1033,
+      "step": 20510
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001258031377534069,
+      "loss": 2.1105,
+      "step": 20515
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001257640734920179,
+      "loss": 2.163,
+      "step": 20520
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012572500501900215,
+      "loss": 2.1457,
+      "step": 20525
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012568593234074617,
+      "loss": 2.0372,
+      "step": 20530
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001256468554636371,
+      "loss": 2.081,
+      "step": 20535
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001256077743940628,
+      "loss": 2.1105,
+      "step": 20540
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012556868913841184,
+      "loss": 2.1358,
+      "step": 20545
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012552959970307343,
+      "loss": 2.1386,
+      "step": 20550
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012549050609443746,
+      "loss": 2.1387,
+      "step": 20555
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001254514083188945,
+      "loss": 2.1033,
+      "step": 20560
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012541230638283592,
+      "loss": 2.175,
+      "step": 20565
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001253732002926536,
+      "loss": 2.1345,
+      "step": 20570
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012533409005474017,
+      "loss": 2.1427,
+      "step": 20575
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012529497567548897,
+      "loss": 2.1943,
+      "step": 20580
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012525585716129396,
+      "loss": 2.1059,
+      "step": 20585
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012521673451854984,
+      "loss": 2.1226,
+      "step": 20590
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001251776077536519,
+      "loss": 2.1263,
+      "step": 20595
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012513847687299618,
+      "loss": 2.0743,
+      "step": 20600
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012509934188297938,
+      "loss": 2.1139,
+      "step": 20605
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012506020278999885,
+      "loss": 2.1295,
+      "step": 20610
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001250210596004526,
+      "loss": 2.1338,
+      "step": 20615
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012498191232073933,
+      "loss": 2.1208,
+      "step": 20620
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001249427609572584,
+      "loss": 2.148,
+      "step": 20625
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012490360551640985,
+      "loss": 2.1656,
+      "step": 20630
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001248644460045944,
+      "loss": 2.0704,
+      "step": 20635
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001248252824282134,
+      "loss": 2.1184,
+      "step": 20640
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012478611479366887,
+      "loss": 2.1149,
+      "step": 20645
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012474694310736351,
+      "loss": 2.1146,
+      "step": 20650
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012470776737570066,
+      "loss": 2.2048,
+      "step": 20655
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001246685876050844,
+      "loss": 2.1585,
+      "step": 20660
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012462940380191933,
+      "loss": 2.1644,
+      "step": 20665
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012459021597261083,
+      "loss": 2.1567,
+      "step": 20670
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012455102412356486,
+      "loss": 2.1853,
+      "step": 20675
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012451182826118812,
+      "loss": 2.1286,
+      "step": 20680
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012447262839188788,
+      "loss": 2.1066,
+      "step": 20685
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001244334245220721,
+      "loss": 2.1362,
+      "step": 20690
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012439421665814946,
+      "loss": 2.0641,
+      "step": 20695
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012435500480652913,
+      "loss": 2.0703,
+      "step": 20700
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012431578897362111,
+      "loss": 2.1238,
+      "step": 20705
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012427656916583595,
+      "loss": 2.1235,
+      "step": 20710
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012423734538958482,
+      "loss": 2.1496,
+      "step": 20715
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012419811765127967,
+      "loss": 2.117,
+      "step": 20720
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012415888595733297,
+      "loss": 2.1642,
+      "step": 20725
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001241196503141579,
+      "loss": 2.1102,
+      "step": 20730
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012408041072816824,
+      "loss": 2.123,
+      "step": 20735
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012404116720577852,
+      "loss": 2.0836,
+      "step": 20740
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012400191975340378,
+      "loss": 2.0737,
+      "step": 20745
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012396266837745975,
+      "loss": 2.1124,
+      "step": 20750
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012392341308436287,
+      "loss": 2.1315,
+      "step": 20755
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012388415388053009,
+      "loss": 2.1343,
+      "step": 20760
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012384489077237916,
+      "loss": 2.1598,
+      "step": 20765
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001238056237663283,
+      "loss": 2.0921,
+      "step": 20770
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012376635286879645,
+      "loss": 2.1346,
+      "step": 20775
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012372707808620326,
+      "loss": 2.1435,
+      "step": 20780
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012368779942496886,
+      "loss": 2.1324,
+      "step": 20785
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012364851689151412,
+      "loss": 2.1386,
+      "step": 20790
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012360923049226053,
+      "loss": 2.1851,
+      "step": 20795
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012356994023363018,
+      "loss": 2.1245,
+      "step": 20800
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012353064612204582,
+      "loss": 2.1409,
+      "step": 20805
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001234913481639308,
+      "loss": 2.1202,
+      "step": 20810
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012345204636570912,
+      "loss": 2.1862,
+      "step": 20815
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012341274073380545,
+      "loss": 2.0956,
+      "step": 20820
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012337343127464498,
+      "loss": 2.0698,
+      "step": 20825
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001233341179946536,
+      "loss": 2.1185,
+      "step": 20830
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012329480090025785,
+      "loss": 2.1246,
+      "step": 20835
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001232554799978848,
+      "loss": 2.0881,
+      "step": 20840
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00012321615529396225,
+      "loss": 2.1443,
+      "step": 20845
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012317682679491854,
+      "loss": 2.1239,
+      "step": 20850
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012313749450718265,
+      "loss": 2.1422,
+      "step": 20855
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012309815843718425,
+      "loss": 2.1186,
+      "step": 20860
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012305881859135347,
+      "loss": 2.1688,
+      "step": 20865
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012301947497612124,
+      "loss": 2.1761,
+      "step": 20870
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012298012759791896,
+      "loss": 2.1507,
+      "step": 20875
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012294077646317873,
+      "loss": 2.1353,
+      "step": 20880
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012290142157833327,
+      "loss": 2.1187,
+      "step": 20885
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012286206294981583,
+      "loss": 2.0863,
+      "step": 20890
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012282270058406035,
+      "loss": 2.1312,
+      "step": 20895
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012278333448750137,
+      "loss": 2.193,
+      "step": 20900
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.000122743964666574,
+      "loss": 2.1034,
+      "step": 20905
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012270459112771398,
+      "loss": 2.1441,
+      "step": 20910
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001226652138773577,
+      "loss": 2.1615,
+      "step": 20915
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012262583292194207,
+      "loss": 2.1073,
+      "step": 20920
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001225864482679047,
+      "loss": 2.1243,
+      "step": 20925
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012254705992168372,
+      "loss": 2.1215,
+      "step": 20930
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012250766788971795,
+      "loss": 2.1268,
+      "step": 20935
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001224682721784467,
+      "loss": 2.1512,
+      "step": 20940
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012242887279431002,
+      "loss": 2.1487,
+      "step": 20945
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012238946974374844,
+      "loss": 2.1302,
+      "step": 20950
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001223500630332031,
+      "loss": 2.1021,
+      "step": 20955
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012231065266911587,
+      "loss": 2.1008,
+      "step": 20960
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012227123865792905,
+      "loss": 2.1502,
+      "step": 20965
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012223182100608567,
+      "loss": 2.1364,
+      "step": 20970
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00012219239972002924,
+      "loss": 2.1491,
+      "step": 20975
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012215297480620393,
+      "loss": 2.1199,
+      "step": 20980
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012211354627105446,
+      "loss": 2.1051,
+      "step": 20985
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012207411412102625,
+      "loss": 2.1478,
+      "step": 20990
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012203467836256517,
+      "loss": 2.1039,
+      "step": 20995
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012199523900211774,
+      "loss": 2.0574,
+      "step": 21000
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012195579604613109,
+      "loss": 2.1128,
+      "step": 21005
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012191634950105295,
+      "loss": 2.1197,
+      "step": 21010
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012187689937333158,
+      "loss": 2.0816,
+      "step": 21015
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012183744566941583,
+      "loss": 2.1075,
+      "step": 21020
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012179798839575517,
+      "loss": 2.163,
+      "step": 21025
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012175852755879969,
+      "loss": 2.1408,
+      "step": 21030
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012171906316499994,
+      "loss": 2.128,
+      "step": 21035
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012167959522080714,
+      "loss": 2.1252,
+      "step": 21040
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012164012373267311,
+      "loss": 2.1767,
+      "step": 21045
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012160064870705017,
+      "loss": 2.1636,
+      "step": 21050
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001215611701503913,
+      "loss": 2.1152,
+      "step": 21055
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012152168806915001,
+      "loss": 2.1204,
+      "step": 21060
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012148220246978042,
+      "loss": 2.1906,
+      "step": 21065
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012144271335873712,
+      "loss": 2.0752,
+      "step": 21070
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00012140322074247544,
+      "loss": 2.1697,
+      "step": 21075
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012136372462745114,
+      "loss": 2.1003,
+      "step": 21080
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012132422502012063,
+      "loss": 2.0915,
+      "step": 21085
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012128472192694088,
+      "loss": 2.0562,
+      "step": 21090
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001212452153543694,
+      "loss": 2.1092,
+      "step": 21095
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012120570530886431,
+      "loss": 2.0927,
+      "step": 21100
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012116619179688424,
+      "loss": 2.174,
+      "step": 21105
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00012112667482488849,
+      "loss": 2.1676,
+      "step": 21110
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001210871543993368,
+      "loss": 2.1157,
+      "step": 21115
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00012104763052668957,
+      "loss": 2.1385,
+      "step": 21120
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001210081032134077,
+      "loss": 2.0736,
+      "step": 21125
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012096857246595268,
+      "loss": 2.1436,
+      "step": 21130
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012092903829078656,
+      "loss": 2.1341,
+      "step": 21135
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012088950069437199,
+      "loss": 2.15,
+      "step": 21140
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012084995968317212,
+      "loss": 2.1244,
+      "step": 21145
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012081041526365065,
+      "loss": 2.0713,
+      "step": 21150
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00012077086744227189,
+      "loss": 2.1766,
+      "step": 21155
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012073131622550071,
+      "loss": 2.1314,
+      "step": 21160
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012069176161980245,
+      "loss": 2.0421,
+      "step": 21165
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012065220363164309,
+      "loss": 2.1496,
+      "step": 21170
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012061264226748911,
+      "loss": 2.1357,
+      "step": 21175
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001205730775338076,
+      "loss": 2.1571,
+      "step": 21180
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012053350943706616,
+      "loss": 2.1642,
+      "step": 21185
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012049393798373293,
+      "loss": 2.1261,
+      "step": 21190
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012045436318027663,
+      "loss": 2.1457,
+      "step": 21195
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001204147850331665,
+      "loss": 2.1329,
+      "step": 21200
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00012037520354887235,
+      "loss": 2.1939,
+      "step": 21205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00012033561873386452,
+      "loss": 2.0829,
+      "step": 21210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012029603059461392,
+      "loss": 2.1451,
+      "step": 21215
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00012025643913759193,
+      "loss": 2.1647,
+      "step": 21220
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00012021684436927056,
+      "loss": 2.1898,
+      "step": 21225
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012017724629612234,
+      "loss": 2.1465,
+      "step": 21230
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00012013764492462032,
+      "loss": 2.1699,
+      "step": 21235
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00012009804026123806,
+      "loss": 2.1141,
+      "step": 21240
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00012005843231244974,
+      "loss": 2.139,
+      "step": 21245
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00012001882108473,
+      "loss": 2.151,
+      "step": 21250
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011997920658455405,
+      "loss": 2.1489,
+      "step": 21255
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011993958881839764,
+      "loss": 2.1042,
+      "step": 21260
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011989996779273706,
+      "loss": 2.1102,
+      "step": 21265
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011986034351404911,
+      "loss": 2.1804,
+      "step": 21270
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001198207159888111,
+      "loss": 2.1262,
+      "step": 21275
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011978108522350093,
+      "loss": 2.0662,
+      "step": 21280
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011974145122459701,
+      "loss": 2.1361,
+      "step": 21285
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011970181399857824,
+      "loss": 2.1074,
+      "step": 21290
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001196621735519241,
+      "loss": 2.1505,
+      "step": 21295
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011962252989111458,
+      "loss": 2.1499,
+      "step": 21300
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011958288302263015,
+      "loss": 2.0898,
+      "step": 21305
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011954323295295187,
+      "loss": 2.1091,
+      "step": 21310
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001195035796885613,
+      "loss": 2.1696,
+      "step": 21315
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011946392323594049,
+      "loss": 2.1057,
+      "step": 21320
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011942426360157209,
+      "loss": 2.1303,
+      "step": 21325
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011938460079193915,
+      "loss": 2.1823,
+      "step": 21330
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001193449348135254,
+      "loss": 2.1659,
+      "step": 21335
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011930526567281495,
+      "loss": 2.1207,
+      "step": 21340
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011926559337629244,
+      "loss": 2.1348,
+      "step": 21345
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011922591793044311,
+      "loss": 2.083,
+      "step": 21350
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011918623934175261,
+      "loss": 2.1263,
+      "step": 21355
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011914655761670723,
+      "loss": 2.1726,
+      "step": 21360
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011910687276179364,
+      "loss": 2.1481,
+      "step": 21365
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011906718478349912,
+      "loss": 2.1172,
+      "step": 21370
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0001190274936883114,
+      "loss": 2.0849,
+      "step": 21375
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011898779948271876,
+      "loss": 2.1287,
+      "step": 21380
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011894810217320997,
+      "loss": 2.164,
+      "step": 21385
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001189084017662743,
+      "loss": 2.1797,
+      "step": 21390
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011886869826840153,
+      "loss": 2.1191,
+      "step": 21395
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00011882899168608197,
+      "loss": 2.0912,
+      "step": 21400
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011878928202580639,
+      "loss": 2.1206,
+      "step": 21405
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011874956929406611,
+      "loss": 2.164,
+      "step": 21410
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001187098534973529,
+      "loss": 2.145,
+      "step": 21415
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001186701346421591,
+      "loss": 2.1593,
+      "step": 21420
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011863041273497748,
+      "loss": 2.1104,
+      "step": 21425
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011859068778230138,
+      "loss": 2.1085,
+      "step": 21430
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00011855095979062454,
+      "loss": 2.1707,
+      "step": 21435
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00011851122876644127,
+      "loss": 2.0822,
+      "step": 21440
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00011847149471624639,
+      "loss": 2.0667,
+      "step": 21445
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011843175764653516,
+      "loss": 2.1703,
+      "step": 21450
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011839201756380336,
+      "loss": 2.2032,
+      "step": 21455
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011835227447454727,
+      "loss": 2.1068,
+      "step": 21460
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011831252838526363,
+      "loss": 2.1608,
+      "step": 21465
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011827277930244973,
+      "loss": 2.1244,
+      "step": 21470
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001182330272326033,
+      "loss": 2.138,
+      "step": 21475
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011819327218222253,
+      "loss": 2.1667,
+      "step": 21480
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001181535141578062,
+      "loss": 2.1254,
+      "step": 21485
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011811375316585347,
+      "loss": 2.0963,
+      "step": 21490
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011807398921286404,
+      "loss": 2.1379,
+      "step": 21495
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011803422230533807,
+      "loss": 2.1776,
+      "step": 21500
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011799445244977625,
+      "loss": 2.1174,
+      "step": 21505
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001179546796526797,
+      "loss": 2.0667,
+      "step": 21510
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011791490392055002,
+      "loss": 2.0996,
+      "step": 21515
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011787512525988933,
+      "loss": 2.1007,
+      "step": 21520
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001178353436772002,
+      "loss": 2.1531,
+      "step": 21525
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011779555917898569,
+      "loss": 2.1511,
+      "step": 21530
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011775577177174932,
+      "loss": 2.1499,
+      "step": 21535
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011771598146199511,
+      "loss": 2.0948,
+      "step": 21540
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011767618825622753,
+      "loss": 2.1254,
+      "step": 21545
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011763639216095155,
+      "loss": 2.0858,
+      "step": 21550
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011759659318267258,
+      "loss": 2.1356,
+      "step": 21555
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011755679132789652,
+      "loss": 2.1476,
+      "step": 21560
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011751698660312974,
+      "loss": 2.1671,
+      "step": 21565
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011747717901487909,
+      "loss": 2.0776,
+      "step": 21570
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011743736856965186,
+      "loss": 2.1151,
+      "step": 21575
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011739755527395582,
+      "loss": 2.144,
+      "step": 21580
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.1491036415100098,
+      "eval_runtime": 188.7513,
+      "eval_samples_per_second": 25.722,
+      "eval_steps_per_second": 3.216,
+      "step": 21584
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011735773913429924,
+      "loss": 2.0741,
+      "step": 21585
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00011731792015719079,
+      "loss": 2.092,
+      "step": 21590
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011727809834913965,
+      "loss": 2.1277,
+      "step": 21595
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011723827371665545,
+      "loss": 2.1293,
+      "step": 21600
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011719844626624828,
+      "loss": 2.0831,
+      "step": 21605
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011715861600442872,
+      "loss": 2.0923,
+      "step": 21610
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011711878293770772,
+      "loss": 2.1695,
+      "step": 21615
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011707894707259681,
+      "loss": 2.1554,
+      "step": 21620
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011703910841560787,
+      "loss": 2.1296,
+      "step": 21625
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011699926697325331,
+      "loss": 2.2003,
+      "step": 21630
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011695942275204599,
+      "loss": 2.1606,
+      "step": 21635
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011691957575849918,
+      "loss": 2.1305,
+      "step": 21640
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011687972599912662,
+      "loss": 2.1345,
+      "step": 21645
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011683987348044251,
+      "loss": 2.122,
+      "step": 21650
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011680001820896154,
+      "loss": 2.0852,
+      "step": 21655
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011676016019119875,
+      "loss": 2.1183,
+      "step": 21660
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011672029943366976,
+      "loss": 2.1206,
+      "step": 21665
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011668043594289048,
+      "loss": 2.1277,
+      "step": 21670
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011664056972537742,
+      "loss": 2.1494,
+      "step": 21675
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011660070078764746,
+      "loss": 2.0702,
+      "step": 21680
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001165608291362179,
+      "loss": 2.1387,
+      "step": 21685
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011652095477760659,
+      "loss": 2.085,
+      "step": 21690
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011648107771833167,
+      "loss": 2.1234,
+      "step": 21695
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00011644119796491184,
+      "loss": 2.0603,
+      "step": 21700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011640131552386622,
+      "loss": 2.1109,
+      "step": 21705
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011636143040171432,
+      "loss": 2.1062,
+      "step": 21710
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011632154260497611,
+      "loss": 2.0379,
+      "step": 21715
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011628165214017206,
+      "loss": 2.1122,
+      "step": 21720
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011624175901382297,
+      "loss": 2.0566,
+      "step": 21725
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011620186323245014,
+      "loss": 2.1424,
+      "step": 21730
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011616196480257533,
+      "loss": 2.0979,
+      "step": 21735
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011612206373072065,
+      "loss": 2.124,
+      "step": 21740
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001160821600234087,
+      "loss": 2.1437,
+      "step": 21745
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011604225368716253,
+      "loss": 2.072,
+      "step": 21750
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011600234472850555,
+      "loss": 2.0915,
+      "step": 21755
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011596243315396163,
+      "loss": 2.1192,
+      "step": 21760
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001159225189700551,
+      "loss": 2.0442,
+      "step": 21765
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001158826021833107,
+      "loss": 2.1511,
+      "step": 21770
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00011584268280025356,
+      "loss": 2.1061,
+      "step": 21775
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011580276082740927,
+      "loss": 2.1277,
+      "step": 21780
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011576283627130384,
+      "loss": 2.1009,
+      "step": 21785
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011572290913846369,
+      "loss": 2.0909,
+      "step": 21790
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011568297943541566,
+      "loss": 2.1055,
+      "step": 21795
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011564304716868704,
+      "loss": 2.1208,
+      "step": 21800
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011560311234480548,
+      "loss": 2.1185,
+      "step": 21805
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011556317497029912,
+      "loss": 2.1292,
+      "step": 21810
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011552323505169646,
+      "loss": 2.1429,
+      "step": 21815
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00011548329259552645,
+      "loss": 2.1214,
+      "step": 21820
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011544334760831847,
+      "loss": 2.1173,
+      "step": 21825
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011540340009660222,
+      "loss": 2.1099,
+      "step": 21830
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011536345006690794,
+      "loss": 2.0512,
+      "step": 21835
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011532349752576618,
+      "loss": 2.1293,
+      "step": 21840
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011528354247970797,
+      "loss": 2.1124,
+      "step": 21845
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011524358493526467,
+      "loss": 2.0705,
+      "step": 21850
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011520362489896817,
+      "loss": 2.1359,
+      "step": 21855
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011516366237735065,
+      "loss": 2.0999,
+      "step": 21860
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011512369737694477,
+      "loss": 2.1233,
+      "step": 21865
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011508372990428353,
+      "loss": 2.0968,
+      "step": 21870
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001150437599659004,
+      "loss": 2.1204,
+      "step": 21875
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011500378756832923,
+      "loss": 2.0889,
+      "step": 21880
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011496381271810429,
+      "loss": 2.0677,
+      "step": 21885
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011492383542176013,
+      "loss": 2.0809,
+      "step": 21890
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011488385568583189,
+      "loss": 2.1711,
+      "step": 21895
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011484387351685497,
+      "loss": 2.1277,
+      "step": 21900
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011480388892136526,
+      "loss": 2.0615,
+      "step": 21905
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011476390190589894,
+      "loss": 2.1076,
+      "step": 21910
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001147239124769927,
+      "loss": 2.0687,
+      "step": 21915
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00011468392064118357,
+      "loss": 2.0901,
+      "step": 21920
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011464392640500892,
+      "loss": 2.1283,
+      "step": 21925
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001146039297750066,
+      "loss": 2.1189,
+      "step": 21930
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011456393075771483,
+      "loss": 2.1823,
+      "step": 21935
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011452392935967223,
+      "loss": 2.104,
+      "step": 21940
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00011448392558741771,
+      "loss": 2.1269,
+      "step": 21945
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011444391944749071,
+      "loss": 2.0608,
+      "step": 21950
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000114403910946431,
+      "loss": 2.1272,
+      "step": 21955
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011436390009077866,
+      "loss": 2.0541,
+      "step": 21960
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011432388688707429,
+      "loss": 2.1351,
+      "step": 21965
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011428387134185879,
+      "loss": 2.149,
+      "step": 21970
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011424385346167346,
+      "loss": 2.0966,
+      "step": 21975
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011420383325305997,
+      "loss": 2.1318,
+      "step": 21980
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011416381072256037,
+      "loss": 2.0863,
+      "step": 21985
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011412378587671715,
+      "loss": 2.1376,
+      "step": 21990
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011408375872207307,
+      "loss": 2.0813,
+      "step": 21995
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011404372926517137,
+      "loss": 2.1227,
+      "step": 22000
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011400369751255563,
+      "loss": 2.0736,
+      "step": 22005
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011396366347076979,
+      "loss": 2.0809,
+      "step": 22010
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011392362714635812,
+      "loss": 2.1458,
+      "step": 22015
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011388358854586541,
+      "loss": 2.0973,
+      "step": 22020
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011384354767583665,
+      "loss": 2.0937,
+      "step": 22025
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011380350454281734,
+      "loss": 2.0845,
+      "step": 22030
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011376345915335327,
+      "loss": 2.0685,
+      "step": 22035
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011372341151399059,
+      "loss": 2.1277,
+      "step": 22040
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00011368336163127588,
+      "loss": 2.1485,
+      "step": 22045
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011364330951175605,
+      "loss": 2.1095,
+      "step": 22050
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011360325516197836,
+      "loss": 2.0798,
+      "step": 22055
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011356319858849047,
+      "loss": 2.0813,
+      "step": 22060
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011352313979784042,
+      "loss": 2.0923,
+      "step": 22065
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011348307879657652,
+      "loss": 2.1241,
+      "step": 22070
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011344301559124751,
+      "loss": 2.0743,
+      "step": 22075
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011340295018840248,
+      "loss": 2.1187,
+      "step": 22080
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011336288259459091,
+      "loss": 2.103,
+      "step": 22085
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001133228128163626,
+      "loss": 2.1624,
+      "step": 22090
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011328274086026768,
+      "loss": 2.0692,
+      "step": 22095
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011324266673285673,
+      "loss": 2.0255,
+      "step": 22100
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001132025904406806,
+      "loss": 2.0783,
+      "step": 22105
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011316251199029048,
+      "loss": 2.1175,
+      "step": 22110
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011312243138823804,
+      "loss": 2.072,
+      "step": 22115
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011308234864107514,
+      "loss": 2.1011,
+      "step": 22120
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001130422637553541,
+      "loss": 2.0573,
+      "step": 22125
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011300217673762752,
+      "loss": 2.1375,
+      "step": 22130
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011296208759444844,
+      "loss": 2.0553,
+      "step": 22135
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011292199633237015,
+      "loss": 2.1439,
+      "step": 22140
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011288190295794633,
+      "loss": 2.1253,
+      "step": 22145
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011284180747773105,
+      "loss": 2.1041,
+      "step": 22150
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011280170989827862,
+      "loss": 2.1563,
+      "step": 22155
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011276161022614377,
+      "loss": 2.1292,
+      "step": 22160
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011272150846788155,
+      "loss": 2.1071,
+      "step": 22165
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011268140463004735,
+      "loss": 2.103,
+      "step": 22170
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001126412987191969,
+      "loss": 2.0783,
+      "step": 22175
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001126011907418863,
+      "loss": 2.0757,
+      "step": 22180
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011256108070467193,
+      "loss": 2.1168,
+      "step": 22185
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011252096861411056,
+      "loss": 2.0884,
+      "step": 22190
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011248085447675924,
+      "loss": 2.0739,
+      "step": 22195
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011244073829917543,
+      "loss": 2.1175,
+      "step": 22200
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011240062008791683,
+      "loss": 2.0563,
+      "step": 22205
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011236049984954157,
+      "loss": 2.1126,
+      "step": 22210
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011232037759060805,
+      "loss": 2.1057,
+      "step": 22215
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000112280253317675,
+      "loss": 2.1085,
+      "step": 22220
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011224012703730152,
+      "loss": 2.1439,
+      "step": 22225
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011219999875604697,
+      "loss": 2.1412,
+      "step": 22230
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011215986848047111,
+      "loss": 2.1137,
+      "step": 22235
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011211973621713398,
+      "loss": 2.1405,
+      "step": 22240
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011207960197259599,
+      "loss": 2.0845,
+      "step": 22245
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011203946575341781,
+      "loss": 2.1128,
+      "step": 22250
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001119993275661605,
+      "loss": 2.1316,
+      "step": 22255
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011195918741738538,
+      "loss": 2.1931,
+      "step": 22260
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011191904531365411,
+      "loss": 2.0941,
+      "step": 22265
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011187890126152871,
+      "loss": 2.0993,
+      "step": 22270
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011183875526757149,
+      "loss": 2.0642,
+      "step": 22275
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011179860733834504,
+      "loss": 2.1136,
+      "step": 22280
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011175845748041231,
+      "loss": 2.0501,
+      "step": 22285
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011171830570033659,
+      "loss": 2.0958,
+      "step": 22290
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011167815200468141,
+      "loss": 2.1693,
+      "step": 22295
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00011163799640001067,
+      "loss": 2.1778,
+      "step": 22300
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011159783889288855,
+      "loss": 2.0458,
+      "step": 22305
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011155767948987961,
+      "loss": 2.185,
+      "step": 22310
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011151751819754859,
+      "loss": 2.1815,
+      "step": 22315
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011147735502246067,
+      "loss": 2.1002,
+      "step": 22320
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00011143718997118127,
+      "loss": 2.0826,
+      "step": 22325
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001113970230502761,
+      "loss": 2.1095,
+      "step": 22330
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011135685426631126,
+      "loss": 2.0821,
+      "step": 22335
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001113166836258531,
+      "loss": 2.1779,
+      "step": 22340
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001112765111354682,
+      "loss": 2.1033,
+      "step": 22345
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011123633680172356,
+      "loss": 2.0778,
+      "step": 22350
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011119616063118645,
+      "loss": 2.1331,
+      "step": 22355
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011115598263042442,
+      "loss": 2.1429,
+      "step": 22360
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011111580280600533,
+      "loss": 2.1599,
+      "step": 22365
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001110756211644973,
+      "loss": 2.0905,
+      "step": 22370
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011103543771246882,
+      "loss": 2.0807,
+      "step": 22375
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011099525245648863,
+      "loss": 2.1442,
+      "step": 22380
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011095506540312577,
+      "loss": 2.0881,
+      "step": 22385
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011091487655894959,
+      "loss": 2.1199,
+      "step": 22390
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00011087468593052967,
+      "loss": 2.165,
+      "step": 22395
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.000110834493524436,
+      "loss": 2.0844,
+      "step": 22400
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011079429934723875,
+      "loss": 2.0827,
+      "step": 22405
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011075410340550843,
+      "loss": 2.0795,
+      "step": 22410
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011071390570581582,
+      "loss": 2.0741,
+      "step": 22415
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011067370625473201,
+      "loss": 2.12,
+      "step": 22420
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011063350505882835,
+      "loss": 2.1172,
+      "step": 22425
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011059330212467653,
+      "loss": 2.1211,
+      "step": 22430
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011055309745884842,
+      "loss": 2.1006,
+      "step": 22435
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011051289106791627,
+      "loss": 2.0467,
+      "step": 22440
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00011047268295845256,
+      "loss": 2.097,
+      "step": 22445
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011043247313703009,
+      "loss": 2.1068,
+      "step": 22450
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00011039226161022191,
+      "loss": 2.0651,
+      "step": 22455
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011035204838460136,
+      "loss": 2.1257,
+      "step": 22460
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011031183346674206,
+      "loss": 2.1254,
+      "step": 22465
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00011027161686321786,
+      "loss": 2.1145,
+      "step": 22470
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00011023139858060298,
+      "loss": 2.0762,
+      "step": 22475
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011019117862547184,
+      "loss": 2.0694,
+      "step": 22480
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00011015095700439917,
+      "loss": 2.1224,
+      "step": 22485
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00011011073372395993,
+      "loss": 2.1023,
+      "step": 22490
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00011007050879072938,
+      "loss": 2.084,
+      "step": 22495
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00011003028221128306,
+      "loss": 2.1255,
+      "step": 22500
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010999005399219679,
+      "loss": 2.1576,
+      "step": 22505
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001099498241400466,
+      "loss": 2.1263,
+      "step": 22510
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010990959266140886,
+      "loss": 2.1157,
+      "step": 22515
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001098693595628601,
+      "loss": 2.0783,
+      "step": 22520
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010982912485097728,
+      "loss": 2.1051,
+      "step": 22525
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010978888853233744,
+      "loss": 2.1403,
+      "step": 22530
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010974865061351802,
+      "loss": 2.1081,
+      "step": 22535
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010970841110109666,
+      "loss": 2.1183,
+      "step": 22540
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010966817000165126,
+      "loss": 2.0843,
+      "step": 22545
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010962792732176002,
+      "loss": 2.0512,
+      "step": 22550
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010958768306800134,
+      "loss": 2.0489,
+      "step": 22555
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010954743724695391,
+      "loss": 2.1472,
+      "step": 22560
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010950718986519668,
+      "loss": 2.1059,
+      "step": 22565
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010946694092930887,
+      "loss": 2.0586,
+      "step": 22570
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010942669044586989,
+      "loss": 2.1361,
+      "step": 22575
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010938643842145948,
+      "loss": 2.1689,
+      "step": 22580
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010934618486265758,
+      "loss": 2.1016,
+      "step": 22585
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010930592977604439,
+      "loss": 2.0956,
+      "step": 22590
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010926567316820037,
+      "loss": 2.0934,
+      "step": 22595
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010922541504570627,
+      "loss": 2.1449,
+      "step": 22600
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000109185155415143,
+      "loss": 2.0911,
+      "step": 22605
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010914489428309178,
+      "loss": 2.1257,
+      "step": 22610
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010910463165613405,
+      "loss": 2.1458,
+      "step": 22615
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010906436754085147,
+      "loss": 2.0708,
+      "step": 22620
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010902410194382601,
+      "loss": 2.1332,
+      "step": 22625
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010898383487163986,
+      "loss": 2.1223,
+      "step": 22630
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010894356633087539,
+      "loss": 2.0891,
+      "step": 22635
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001089032963281153,
+      "loss": 2.1406,
+      "step": 22640
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010886302486994244,
+      "loss": 2.13,
+      "step": 22645
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010882275196293998,
+      "loss": 2.0739,
+      "step": 22650
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010878247761369129,
+      "loss": 2.144,
+      "step": 22655
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010874220182877996,
+      "loss": 2.131,
+      "step": 22660
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010870192461478987,
+      "loss": 2.1101,
+      "step": 22665
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010866164597830504,
+      "loss": 2.129,
+      "step": 22670
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010862136592590981,
+      "loss": 2.0363,
+      "step": 22675
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001085810844641887,
+      "loss": 2.0746,
+      "step": 22680
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001085408015997265,
+      "loss": 2.0776,
+      "step": 22685
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010850051733910822,
+      "loss": 2.1138,
+      "step": 22690
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010846023168891903,
+      "loss": 2.1203,
+      "step": 22695
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010841994465574448,
+      "loss": 2.1349,
+      "step": 22700
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010837965624617014,
+      "loss": 2.1169,
+      "step": 22705
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010833936646678198,
+      "loss": 2.1791,
+      "step": 22710
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010829907532416612,
+      "loss": 2.0724,
+      "step": 22715
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001082587828249089,
+      "loss": 2.1318,
+      "step": 22720
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001082184889755969,
+      "loss": 2.1121,
+      "step": 22725
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010817819378281693,
+      "loss": 2.1121,
+      "step": 22730
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010813789725315597,
+      "loss": 2.0626,
+      "step": 22735
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010809759939320127,
+      "loss": 2.0958,
+      "step": 22740
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010805730020954028,
+      "loss": 2.1161,
+      "step": 22745
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001080169997087607,
+      "loss": 2.1251,
+      "step": 22750
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010797669789745033,
+      "loss": 2.0852,
+      "step": 22755
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010793639478219735,
+      "loss": 2.0509,
+      "step": 22760
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010789609036959002,
+      "loss": 2.0757,
+      "step": 22765
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010785578466621688,
+      "loss": 2.1345,
+      "step": 22770
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010781547767866664,
+      "loss": 2.1197,
+      "step": 22775
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010777516941352828,
+      "loss": 2.0933,
+      "step": 22780
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010773485987739091,
+      "loss": 2.179,
+      "step": 22785
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010769454907684396,
+      "loss": 2.1247,
+      "step": 22790
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010765423701847692,
+      "loss": 2.1162,
+      "step": 22795
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010761392370887957,
+      "loss": 2.1676,
+      "step": 22800
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001075736091546419,
+      "loss": 2.0899,
+      "step": 22805
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010753329336235409,
+      "loss": 2.0735,
+      "step": 22810
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010749297633860653,
+      "loss": 2.1149,
+      "step": 22815
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010745265808998981,
+      "loss": 2.1203,
+      "step": 22820
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010741233862309467,
+      "loss": 2.088,
+      "step": 22825
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010737201794451213,
+      "loss": 2.1788,
+      "step": 22830
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010733169606083335,
+      "loss": 2.1491,
+      "step": 22835
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010729137297864973,
+      "loss": 2.1784,
+      "step": 22840
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010725104870455281,
+      "loss": 2.1579,
+      "step": 22845
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001072107232451344,
+      "loss": 2.1496,
+      "step": 22850
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010717039660698641,
+      "loss": 2.1319,
+      "step": 22855
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010713006879670103,
+      "loss": 2.0884,
+      "step": 22860
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001070897398208706,
+      "loss": 2.1352,
+      "step": 22865
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010704940968608765,
+      "loss": 2.1688,
+      "step": 22870
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010700907839894493,
+      "loss": 2.1193,
+      "step": 22875
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010696874596603532,
+      "loss": 2.0667,
+      "step": 22880
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010692841239395197,
+      "loss": 2.064,
+      "step": 22885
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001068880776892881,
+      "loss": 2.1331,
+      "step": 22890
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010684774185863721,
+      "loss": 2.0985,
+      "step": 22895
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010680740490859298,
+      "loss": 2.1176,
+      "step": 22900
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010676706684574925,
+      "loss": 2.0825,
+      "step": 22905
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001067267276767,
+      "loss": 2.1431,
+      "step": 22910
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010668638740803945,
+      "loss": 2.109,
+      "step": 22915
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010664604604636202,
+      "loss": 2.0887,
+      "step": 22920
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010660570359826221,
+      "loss": 2.1041,
+      "step": 22925
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010656536007033482,
+      "loss": 2.1174,
+      "step": 22930
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010652501546917473,
+      "loss": 2.1096,
+      "step": 22935
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000106484669801377,
+      "loss": 2.1055,
+      "step": 22940
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010644432307353698,
+      "loss": 2.1076,
+      "step": 22945
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010640397529225004,
+      "loss": 2.1626,
+      "step": 22950
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010636362646411183,
+      "loss": 2.1721,
+      "step": 22955
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010632327659571808,
+      "loss": 2.0655,
+      "step": 22960
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010628292569366478,
+      "loss": 2.119,
+      "step": 22965
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010624257376454804,
+      "loss": 2.1367,
+      "step": 22970
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001062022208149642,
+      "loss": 2.1616,
+      "step": 22975
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010616186685150964,
+      "loss": 2.0259,
+      "step": 22980
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010612151188078099,
+      "loss": 2.1695,
+      "step": 22985
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010608115590937507,
+      "loss": 2.1536,
+      "step": 22990
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010604079894388881,
+      "loss": 2.1368,
+      "step": 22995
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010600044099091932,
+      "loss": 2.0686,
+      "step": 23000
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001059600820570639,
+      "loss": 2.069,
+      "step": 23005
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010591972214891994,
+      "loss": 2.1336,
+      "step": 23010
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010587936127308509,
+      "loss": 2.1422,
+      "step": 23015
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010583899943615704,
+      "loss": 2.123,
+      "step": 23020
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010579863664473372,
+      "loss": 2.1723,
+      "step": 23025
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010575827290541322,
+      "loss": 2.1094,
+      "step": 23030
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010571790822479374,
+      "loss": 2.1404,
+      "step": 23035
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010567754260947364,
+      "loss": 2.0641,
+      "step": 23040
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010563717606605147,
+      "loss": 2.1102,
+      "step": 23045
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010559680860112589,
+      "loss": 2.1402,
+      "step": 23050
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010555644022129574,
+      "loss": 2.126,
+      "step": 23055
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010551607093315998,
+      "loss": 2.1257,
+      "step": 23060
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010547570074331776,
+      "loss": 2.0608,
+      "step": 23065
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010543532965836835,
+      "loss": 2.0933,
+      "step": 23070
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010539495768491114,
+      "loss": 2.0888,
+      "step": 23075
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010535458482954572,
+      "loss": 2.1372,
+      "step": 23080
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010531421109887181,
+      "loss": 2.0887,
+      "step": 23085
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010527383649948923,
+      "loss": 2.1318,
+      "step": 23090
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010523346103799801,
+      "loss": 2.1164,
+      "step": 23095
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010519308472099825,
+      "loss": 2.0967,
+      "step": 23100
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010515270755509024,
+      "loss": 2.1124,
+      "step": 23105
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010511232954687443,
+      "loss": 2.1478,
+      "step": 23110
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010507195070295131,
+      "loss": 2.1224,
+      "step": 23115
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010503157102992158,
+      "loss": 2.0993,
+      "step": 23120
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001049911905343861,
+      "loss": 2.0809,
+      "step": 23125
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010495080922294582,
+      "loss": 2.1211,
+      "step": 23130
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00010491042710220183,
+      "loss": 2.0628,
+      "step": 23135
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010487004417875532,
+      "loss": 2.1468,
+      "step": 23140
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001048296604592077,
+      "loss": 2.0732,
+      "step": 23145
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010478927595016039,
+      "loss": 2.1487,
+      "step": 23150
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010474889065821509,
+      "loss": 2.159,
+      "step": 23155
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010470850458997346,
+      "loss": 2.1567,
+      "step": 23160
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001046681177520374,
+      "loss": 2.1034,
+      "step": 23165
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00010462773015100891,
+      "loss": 2.1426,
+      "step": 23170
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001045873417934901,
+      "loss": 2.0639,
+      "step": 23175
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010454695268608323,
+      "loss": 2.1228,
+      "step": 23180
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010450656283539066,
+      "loss": 2.1102,
+      "step": 23185
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010446617224801487,
+      "loss": 2.1213,
+      "step": 23190
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010442578093055846,
+      "loss": 2.0765,
+      "step": 23195
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010438538888962417,
+      "loss": 2.1056,
+      "step": 23200
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010434499613181485,
+      "loss": 2.0779,
+      "step": 23205
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010430460266373348,
+      "loss": 2.1022,
+      "step": 23210
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010426420849198308,
+      "loss": 2.092,
+      "step": 23215
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001042238136231669,
+      "loss": 2.1445,
+      "step": 23220
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010418341806388822,
+      "loss": 2.1029,
+      "step": 23225
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010414302182075047,
+      "loss": 2.1201,
+      "step": 23230
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00010410262490035718,
+      "loss": 2.1274,
+      "step": 23235
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000104062227309312,
+      "loss": 2.1097,
+      "step": 23240
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001040218290542187,
+      "loss": 2.1094,
+      "step": 23245
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010398143014168107,
+      "loss": 2.1086,
+      "step": 23250
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010394103057830315,
+      "loss": 2.111,
+      "step": 23255
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000103900630370689,
+      "loss": 2.0793,
+      "step": 23260
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010386022952544279,
+      "loss": 2.1216,
+      "step": 23265
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010381982804916881,
+      "loss": 2.1046,
+      "step": 23270
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010377942594847145,
+      "loss": 2.1014,
+      "step": 23275
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0001037390232299552,
+      "loss": 2.1016,
+      "step": 23280
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010369861990022463,
+      "loss": 2.0641,
+      "step": 23285
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010365821596588445,
+      "loss": 2.0886,
+      "step": 23290
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010361781143353948,
+      "loss": 2.0955,
+      "step": 23295
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010357740630979455,
+      "loss": 2.1161,
+      "step": 23300
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010353700060125466,
+      "loss": 2.1562,
+      "step": 23305
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010349659431452492,
+      "loss": 2.0497,
+      "step": 23310
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010345618745621049,
+      "loss": 2.0979,
+      "step": 23315
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010341578003291661,
+      "loss": 2.1525,
+      "step": 23320
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010337537205124867,
+      "loss": 2.0644,
+      "step": 23325
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010333496351781214,
+      "loss": 2.1184,
+      "step": 23330
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010329455443921252,
+      "loss": 2.0914,
+      "step": 23335
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010325414482205544,
+      "loss": 2.1286,
+      "step": 23340
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010321373467294665,
+      "loss": 2.1533,
+      "step": 23345
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00010317332399849193,
+      "loss": 2.1205,
+      "step": 23350
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001031329128052972,
+      "loss": 2.0962,
+      "step": 23355
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001030925010999684,
+      "loss": 2.0649,
+      "step": 23360
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010305208888911162,
+      "loss": 2.149,
+      "step": 23365
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010301167617933298,
+      "loss": 2.0852,
+      "step": 23370
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010297126297723874,
+      "loss": 2.0825,
+      "step": 23375
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010293084928943514,
+      "loss": 2.111,
+      "step": 23380
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010289043512252865,
+      "loss": 2.1121,
+      "step": 23385
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010285002048312566,
+      "loss": 2.1172,
+      "step": 23390
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010280960537783275,
+      "loss": 2.1087,
+      "step": 23395
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010276918981325652,
+      "loss": 2.09,
+      "step": 23400
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010272877379600365,
+      "loss": 2.1033,
+      "step": 23405
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010268835733268094,
+      "loss": 2.0855,
+      "step": 23410
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010264794042989521,
+      "loss": 2.0862,
+      "step": 23415
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010260752309425334,
+      "loss": 2.1311,
+      "step": 23420
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010256710533236238,
+      "loss": 2.0775,
+      "step": 23425
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010252668715082932,
+      "loss": 2.2022,
+      "step": 23430
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010248626855626126,
+      "loss": 2.0903,
+      "step": 23435
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010244584955526546,
+      "loss": 2.1287,
+      "step": 23440
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010240543015444912,
+      "loss": 2.0616,
+      "step": 23445
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001023650103604196,
+      "loss": 2.0814,
+      "step": 23450
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00010232459017978421,
+      "loss": 2.0721,
+      "step": 23455
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010228416961915048,
+      "loss": 2.1347,
+      "step": 23460
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010224374868512588,
+      "loss": 2.1057,
+      "step": 23465
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010220332738431795,
+      "loss": 2.076,
+      "step": 23470
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010216290572333436,
+      "loss": 2.1536,
+      "step": 23475
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001021224837087828,
+      "loss": 2.1248,
+      "step": 23480
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010208206134727096,
+      "loss": 2.1114,
+      "step": 23485
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001020416386454067,
+      "loss": 2.0509,
+      "step": 23490
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00010200121560979788,
+      "loss": 2.1513,
+      "step": 23495
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010196079224705238,
+      "loss": 2.0993,
+      "step": 23500
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010192036856377816,
+      "loss": 2.1044,
+      "step": 23505
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010187994456658328,
+      "loss": 2.1133,
+      "step": 23510
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010183952026207577,
+      "loss": 2.1266,
+      "step": 23515
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010179909565686377,
+      "loss": 2.1362,
+      "step": 23520
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010175867075755543,
+      "loss": 2.1021,
+      "step": 23525
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010171824557075899,
+      "loss": 2.1751,
+      "step": 23530
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00010167782010308272,
+      "loss": 2.1241,
+      "step": 23535
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010163739436113488,
+      "loss": 2.1222,
+      "step": 23540
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010159696835152388,
+      "loss": 2.1275,
+      "step": 23545
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010155654208085812,
+      "loss": 2.1586,
+      "step": 23550
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010151611555574598,
+      "loss": 2.1568,
+      "step": 23555
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000101475688782796,
+      "loss": 2.1081,
+      "step": 23560
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00010143526176861669,
+      "loss": 2.1071,
+      "step": 23565
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001013948345198166,
+      "loss": 2.0994,
+      "step": 23570
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010135440704300432,
+      "loss": 2.1325,
+      "step": 23575
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010131397934478852,
+      "loss": 2.1658,
+      "step": 23580
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010127355143177788,
+      "loss": 2.1399,
+      "step": 23585
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010123312331058106,
+      "loss": 2.1083,
+      "step": 23590
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010119269498780686,
+      "loss": 2.0926,
+      "step": 23595
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010115226647006401,
+      "loss": 2.1396,
+      "step": 23600
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010111183776396135,
+      "loss": 2.114,
+      "step": 23605
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010107140887610773,
+      "loss": 2.133,
+      "step": 23610
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010103097981311195,
+      "loss": 2.1524,
+      "step": 23615
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.000100990550581583,
+      "loss": 2.102,
+      "step": 23620
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010095012118812974,
+      "loss": 2.1571,
+      "step": 23625
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010090969163936113,
+      "loss": 2.1014,
+      "step": 23630
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010086926194188618,
+      "loss": 2.1536,
+      "step": 23635
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010082883210231384,
+      "loss": 2.0784,
+      "step": 23640
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010078840212725318,
+      "loss": 2.1175,
+      "step": 23645
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00010074797202331323,
+      "loss": 2.1049,
+      "step": 23650
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00010070754179710307,
+      "loss": 2.1416,
+      "step": 23655
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00010066711145523175,
+      "loss": 2.1233,
+      "step": 23660
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010062668100430843,
+      "loss": 2.1163,
+      "step": 23665
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010058625045094219,
+      "loss": 2.0539,
+      "step": 23670
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001005458198017422,
+      "loss": 2.1588,
+      "step": 23675
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00010050538906331762,
+      "loss": 2.0911,
+      "step": 23680
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010046495824227758,
+      "loss": 2.1155,
+      "step": 23685
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00010042452734523128,
+      "loss": 2.1,
+      "step": 23690
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010038409637878796,
+      "loss": 2.1076,
+      "step": 23695
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010034366534955682,
+      "loss": 2.0884,
+      "step": 23700
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.000100303234264147,
+      "loss": 2.1539,
+      "step": 23705
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010026280312916782,
+      "loss": 2.1262,
+      "step": 23710
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00010022237195122846,
+      "loss": 2.1265,
+      "step": 23715
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00010018194073693821,
+      "loss": 2.0879,
+      "step": 23720
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010014150949290625,
+      "loss": 2.1682,
+      "step": 23725
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00010010107822574193,
+      "loss": 2.1392,
+      "step": 23730
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00010006064694205441,
+      "loss": 2.1586,
+      "step": 23735
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010002021564845303,
+      "loss": 2.1356,
+      "step": 23740
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.9979784351547e-05,
+      "loss": 2.1384,
+      "step": 23745
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.99393530579456e-05,
+      "loss": 2.0987,
+      "step": 23750
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.98989217742581e-05,
+      "loss": 2.1327,
+      "step": 23755
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.985849050709376e-05,
+      "loss": 2.0822,
+      "step": 23760
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.981805926306182e-05,
+      "loss": 2.1221,
+      "step": 23765
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.977762804877155e-05,
+      "loss": 2.1308,
+      "step": 23770
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.97371968708322e-05,
+      "loss": 2.1546,
+      "step": 23775
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.969676573585301e-05,
+      "loss": 2.138,
+      "step": 23780
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.965633465044322e-05,
+      "loss": 2.0949,
+      "step": 23785
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.961590362121205e-05,
+      "loss": 2.124,
+      "step": 23790
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.957547265476871e-05,
+      "loss": 2.1122,
+      "step": 23795
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.953504175772243e-05,
+      "loss": 2.0809,
+      "step": 23800
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.94946109366824e-05,
+      "loss": 2.1077,
+      "step": 23805
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.945418019825779e-05,
+      "loss": 2.1268,
+      "step": 23810
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.94137495490578e-05,
+      "loss": 2.1499,
+      "step": 23815
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.937331899569157e-05,
+      "loss": 2.1244,
+      "step": 23820
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.933288854476824e-05,
+      "loss": 2.0755,
+      "step": 23825
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.929245820289693e-05,
+      "loss": 2.1343,
+      "step": 23830
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.925202797668681e-05,
+      "loss": 2.0751,
+      "step": 23835
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.921159787274687e-05,
+      "loss": 2.1297,
+      "step": 23840
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.91711678976862e-05,
+      "loss": 2.0516,
+      "step": 23845
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.913073805811389e-05,
+      "loss": 2.1398,
+      "step": 23850
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.909030836063892e-05,
+      "loss": 2.0327,
+      "step": 23855
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.904987881187031e-05,
+      "loss": 2.0757,
+      "step": 23860
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.900944941841705e-05,
+      "loss": 2.1384,
+      "step": 23865
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.896902018688807e-05,
+      "loss": 2.0708,
+      "step": 23870
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.892859112389232e-05,
+      "loss": 2.1219,
+      "step": 23875
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.888816223603867e-05,
+      "loss": 2.1305,
+      "step": 23880
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.884773352993602e-05,
+      "loss": 2.0961,
+      "step": 23885
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.880730501219317e-05,
+      "loss": 2.132,
+      "step": 23890
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.876687668941895e-05,
+      "loss": 2.1087,
+      "step": 23895
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.872644856822215e-05,
+      "loss": 2.0564,
+      "step": 23900
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.868602065521149e-05,
+      "loss": 2.1122,
+      "step": 23905
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.86455929569957e-05,
+      "loss": 2.135,
+      "step": 23910
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.860516548018344e-05,
+      "loss": 2.0617,
+      "step": 23915
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.856473823138333e-05,
+      "loss": 2.101,
+      "step": 23920
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.852431121720402e-05,
+      "loss": 2.0989,
+      "step": 23925
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.848388444425404e-05,
+      "loss": 2.0935,
+      "step": 23930
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.844345791914192e-05,
+      "loss": 2.1508,
+      "step": 23935
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.840303164847613e-05,
+      "loss": 2.06,
+      "step": 23940
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.836260563886514e-05,
+      "loss": 2.1622,
+      "step": 23945
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.832217989691731e-05,
+      "loss": 2.1746,
+      "step": 23950
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.828175442924102e-05,
+      "loss": 2.0649,
+      "step": 23955
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.82413292424446e-05,
+      "loss": 2.1635,
+      "step": 23960
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.820090434313625e-05,
+      "loss": 2.0819,
+      "step": 23965
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.816047973792425e-05,
+      "loss": 2.1129,
+      "step": 23970
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.812005543341676e-05,
+      "loss": 2.0626,
+      "step": 23975
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.807963143622187e-05,
+      "loss": 2.1208,
+      "step": 23980
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.803920775294766e-05,
+      "loss": 2.0822,
+      "step": 23985
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.799878439020214e-05,
+      "loss": 2.1254,
+      "step": 23990
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.795836135459332e-05,
+      "loss": 2.1639,
+      "step": 23995
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.791793865272906e-05,
+      "loss": 2.077,
+      "step": 24000
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.787751629121724e-05,
+      "loss": 2.1279,
+      "step": 24005
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.783709427666566e-05,
+      "loss": 2.0772,
+      "step": 24010
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.779667261568209e-05,
+      "loss": 2.0648,
+      "step": 24015
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.775625131487417e-05,
+      "loss": 2.1271,
+      "step": 24020
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.771583038084957e-05,
+      "loss": 2.1213,
+      "step": 24025
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.767540982021582e-05,
+      "loss": 2.1116,
+      "step": 24030
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.763498963958046e-05,
+      "loss": 2.1564,
+      "step": 24035
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.759456984555091e-05,
+      "loss": 2.0887,
+      "step": 24040
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.755415044473459e-05,
+      "loss": 2.0738,
+      "step": 24045
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.751373144373876e-05,
+      "loss": 2.1047,
+      "step": 24050
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.747331284917074e-05,
+      "loss": 2.0806,
+      "step": 24055
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.743289466763766e-05,
+      "loss": 2.0885,
+      "step": 24060
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.739247690574668e-05,
+      "loss": 2.1069,
+      "step": 24065
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.735205957010483e-05,
+      "loss": 2.1036,
+      "step": 24070
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.731164266731908e-05,
+      "loss": 2.0626,
+      "step": 24075
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.727122620399635e-05,
+      "loss": 2.1286,
+      "step": 24080
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.72308101867435e-05,
+      "loss": 2.1277,
+      "step": 24085
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.719039462216727e-05,
+      "loss": 2.1329,
+      "step": 24090
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.714997951687435e-05,
+      "loss": 2.1045,
+      "step": 24095
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.710956487747137e-05,
+      "loss": 2.0953,
+      "step": 24100
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.706915071056487e-05,
+      "loss": 2.0973,
+      "step": 24105
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.702873702276128e-05,
+      "loss": 2.0907,
+      "step": 24110
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.698832382066703e-05,
+      "loss": 2.17,
+      "step": 24115
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.69479111108884e-05,
+      "loss": 2.1031,
+      "step": 24120
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.690749890003161e-05,
+      "loss": 2.1482,
+      "step": 24125
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.686708719470283e-05,
+      "loss": 2.0948,
+      "step": 24130
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.682667600150808e-05,
+      "loss": 2.0102,
+      "step": 24135
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.678626532705337e-05,
+      "loss": 2.1324,
+      "step": 24140
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.674585517794457e-05,
+      "loss": 2.1009,
+      "step": 24145
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.67054455607875e-05,
+      "loss": 2.1439,
+      "step": 24150
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.666503648218787e-05,
+      "loss": 2.1426,
+      "step": 24155
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.662462794875132e-05,
+      "loss": 2.12,
+      "step": 24160
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.658421996708339e-05,
+      "loss": 2.1391,
+      "step": 24165
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.654381254378951e-05,
+      "loss": 2.1857,
+      "step": 24170
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.650340568547508e-05,
+      "loss": 2.1754,
+      "step": 24175
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.646299939874533e-05,
+      "loss": 2.0974,
+      "step": 24180
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.642259369020546e-05,
+      "loss": 2.1378,
+      "step": 24185
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.638218856646053e-05,
+      "loss": 2.1708,
+      "step": 24190
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.634178403411554e-05,
+      "loss": 2.1685,
+      "step": 24195
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.630138009977538e-05,
+      "loss": 2.0695,
+      "step": 24200
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.626097677004483e-05,
+      "loss": 2.0798,
+      "step": 24205
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.622057405152857e-05,
+      "loss": 2.1155,
+      "step": 24210
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.618017195083121e-05,
+      "loss": 2.1028,
+      "step": 24215
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.613977047455722e-05,
+      "loss": 2.1248,
+      "step": 24220
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.609936962931102e-05,
+      "loss": 2.1034,
+      "step": 24225
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.605896942169687e-05,
+      "loss": 2.0966,
+      "step": 24230
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.601856985831894e-05,
+      "loss": 2.1437,
+      "step": 24235
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.597817094578134e-05,
+      "loss": 2.1239,
+      "step": 24240
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.593777269068802e-05,
+      "loss": 2.1249,
+      "step": 24245
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.589737509964284e-05,
+      "loss": 2.1244,
+      "step": 24250
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.585697817924954e-05,
+      "loss": 2.126,
+      "step": 24255
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.58165819361118e-05,
+      "loss": 2.1534,
+      "step": 24260
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.577618637683311e-05,
+      "loss": 2.1098,
+      "step": 24265
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.573579150801694e-05,
+      "loss": 2.1004,
+      "step": 24270
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.569539733626656e-05,
+      "loss": 2.0988,
+      "step": 24275
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.565500386818517e-05,
+      "loss": 2.0703,
+      "step": 24280
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.561461111037584e-05,
+      "loss": 2.1159,
+      "step": 24285
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.557421906944156e-05,
+      "loss": 2.0766,
+      "step": 24290
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.553382775198515e-05,
+      "loss": 2.1258,
+      "step": 24295
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.549343716460936e-05,
+      "loss": 2.1313,
+      "step": 24300
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.545304731391679e-05,
+      "loss": 2.0695,
+      "step": 24305
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.541265820650991e-05,
+      "loss": 2.1316,
+      "step": 24310
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.53722698489911e-05,
+      "loss": 2.1218,
+      "step": 24315
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.53318822479626e-05,
+      "loss": 2.0494,
+      "step": 24320
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.529149541002655e-05,
+      "loss": 2.075,
+      "step": 24325
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.525110934178492e-05,
+      "loss": 2.1605,
+      "step": 24330
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.52107240498396e-05,
+      "loss": 2.1311,
+      "step": 24335
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.51703395407923e-05,
+      "loss": 2.15,
+      "step": 24340
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.512995582124467e-05,
+      "loss": 2.123,
+      "step": 24345
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.508957289779817e-05,
+      "loss": 2.0275,
+      "step": 24350
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.504919077705417e-05,
+      "loss": 2.1737,
+      "step": 24355
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.500880946561388e-05,
+      "loss": 2.106,
+      "step": 24360
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.49684289700784e-05,
+      "loss": 2.133,
+      "step": 24365
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.49280492970487e-05,
+      "loss": 2.1452,
+      "step": 24370
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.488767045312558e-05,
+      "loss": 2.1558,
+      "step": 24375
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.48472924449098e-05,
+      "loss": 2.132,
+      "step": 24380
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.48069152790018e-05,
+      "loss": 2.1366,
+      "step": 24385
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.476653896200206e-05,
+      "loss": 2.0736,
+      "step": 24390
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.472616350051082e-05,
+      "loss": 2.1221,
+      "step": 24395
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.468578890112825e-05,
+      "loss": 2.1104,
+      "step": 24400
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.464541517045432e-05,
+      "loss": 2.0726,
+      "step": 24405
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.46050423150889e-05,
+      "loss": 2.1091,
+      "step": 24410
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.45646703416317e-05,
+      "loss": 2.0969,
+      "step": 24415
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.452429925668227e-05,
+      "loss": 2.1443,
+      "step": 24420
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.448392906684005e-05,
+      "loss": 2.0499,
+      "step": 24425
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.44435597787043e-05,
+      "loss": 2.0967,
+      "step": 24430
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.440319139887414e-05,
+      "loss": 2.0887,
+      "step": 24435
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.436282393394855e-05,
+      "loss": 2.1515,
+      "step": 24440
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.432245739052637e-05,
+      "loss": 2.0697,
+      "step": 24445
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.428209177520628e-05,
+      "loss": 2.1314,
+      "step": 24450
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.424172709458679e-05,
+      "loss": 2.179,
+      "step": 24455
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.420136335526629e-05,
+      "loss": 2.092,
+      "step": 24460
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.416100056384298e-05,
+      "loss": 2.1222,
+      "step": 24465
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.412063872691495e-05,
+      "loss": 2.1311,
+      "step": 24470
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.408027785108007e-05,
+      "loss": 2.0879,
+      "step": 24475
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.403991794293613e-05,
+      "loss": 2.0535,
+      "step": 24480
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.399955900908069e-05,
+      "loss": 2.1104,
+      "step": 24485
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.395920105611123e-05,
+      "loss": 2.1466,
+      "step": 24490
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.391884409062496e-05,
+      "loss": 2.1256,
+      "step": 24495
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.387848811921905e-05,
+      "loss": 2.1652,
+      "step": 24500
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.383813314849038e-05,
+      "loss": 2.0772,
+      "step": 24505
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.379777918503581e-05,
+      "loss": 2.1642,
+      "step": 24510
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.375742623545194e-05,
+      "loss": 2.1283,
+      "step": 24515
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.371707430633521e-05,
+      "loss": 2.123,
+      "step": 24520
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.367672340428193e-05,
+      "loss": 2.0817,
+      "step": 24525
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.363637353588818e-05,
+      "loss": 2.0941,
+      "step": 24530
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.359602470774997e-05,
+      "loss": 2.1253,
+      "step": 24535
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.355567692646303e-05,
+      "loss": 2.0942,
+      "step": 24540
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.351533019862299e-05,
+      "loss": 2.1013,
+      "step": 24545
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.347498453082528e-05,
+      "loss": 2.0834,
+      "step": 24550
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.34346399296652e-05,
+      "loss": 2.0932,
+      "step": 24555
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.339429640173782e-05,
+      "loss": 2.1153,
+      "step": 24560
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.335395395363803e-05,
+      "loss": 2.1014,
+      "step": 24565
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.331361259196059e-05,
+      "loss": 2.1317,
+      "step": 24570
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.327327232330006e-05,
+      "loss": 2.0971,
+      "step": 24575
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.32329331542508e-05,
+      "loss": 2.1199,
+      "step": 24580
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.319259509140706e-05,
+      "loss": 2.1306,
+      "step": 24585
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.315225814136281e-05,
+      "loss": 2.0959,
+      "step": 24590
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.311192231071194e-05,
+      "loss": 2.1464,
+      "step": 24595
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.307158760604807e-05,
+      "loss": 2.093,
+      "step": 24600
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.303125403396469e-05,
+      "loss": 2.0906,
+      "step": 24605
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.299092160105509e-05,
+      "loss": 2.1248,
+      "step": 24610
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.295059031391236e-05,
+      "loss": 2.1843,
+      "step": 24615
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.291026017912941e-05,
+      "loss": 2.0809,
+      "step": 24620
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.286993120329898e-05,
+      "loss": 2.0877,
+      "step": 24625
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.282960339301362e-05,
+      "loss": 2.1141,
+      "step": 24630
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.278927675486564e-05,
+      "loss": 2.0845,
+      "step": 24635
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.274895129544721e-05,
+      "loss": 2.1045,
+      "step": 24640
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.270862702135031e-05,
+      "loss": 2.0572,
+      "step": 24645
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.266830393916668e-05,
+      "loss": 2.0885,
+      "step": 24650
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.26279820554879e-05,
+      "loss": 2.0758,
+      "step": 24655
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.258766137690535e-05,
+      "loss": 2.107,
+      "step": 24660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.254734191001023e-05,
+      "loss": 2.1329,
+      "step": 24665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.250702366139348e-05,
+      "loss": 2.1585,
+      "step": 24670
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.246670663764592e-05,
+      "loss": 2.1303,
+      "step": 24675
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.242639084535811e-05,
+      "loss": 2.1316,
+      "step": 24680
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.238607629112045e-05,
+      "loss": 2.0949,
+      "step": 24685
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.23457629815231e-05,
+      "loss": 2.0849,
+      "step": 24690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.230545092315605e-05,
+      "loss": 2.1008,
+      "step": 24695
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.226514012260907e-05,
+      "loss": 2.0766,
+      "step": 24700
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.222483058647173e-05,
+      "loss": 2.1271,
+      "step": 24705
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.218452232133337e-05,
+      "loss": 2.1194,
+      "step": 24710
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.214421533378313e-05,
+      "loss": 2.0921,
+      "step": 24715
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.210390963041e-05,
+      "loss": 2.0644,
+      "step": 24720
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.206360521780268e-05,
+      "loss": 2.1115,
+      "step": 24725
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.202330210254968e-05,
+      "loss": 2.0509,
+      "step": 24730
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.198300029123934e-05,
+      "loss": 2.0931,
+      "step": 24735
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.194269979045973e-05,
+      "loss": 2.0892,
+      "step": 24740
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.190240060679878e-05,
+      "loss": 2.0991,
+      "step": 24745
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.186210274684408e-05,
+      "loss": 2.0739,
+      "step": 24750
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.182180621718313e-05,
+      "loss": 2.153,
+      "step": 24755
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.178151102440314e-05,
+      "loss": 2.0465,
+      "step": 24760
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.174121717509114e-05,
+      "loss": 2.0645,
+      "step": 24765
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.170092467583391e-05,
+      "loss": 2.0833,
+      "step": 24770
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.166063353321806e-05,
+      "loss": 2.0872,
+      "step": 24775
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.162034375382989e-05,
+      "loss": 2.0869,
+      "step": 24780
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.158005534425556e-05,
+      "loss": 2.117,
+      "step": 24785
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.153976831108098e-05,
+      "loss": 2.1356,
+      "step": 24790
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.149948266089182e-05,
+      "loss": 2.0765,
+      "step": 24795
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.145919840027351e-05,
+      "loss": 2.1381,
+      "step": 24800
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.14189155358113e-05,
+      "loss": 2.0961,
+      "step": 24805
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.137863407409022e-05,
+      "loss": 2.1162,
+      "step": 24810
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.133835402169499e-05,
+      "loss": 2.1305,
+      "step": 24815
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.129807538521015e-05,
+      "loss": 2.076,
+      "step": 24820
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.125779817122005e-05,
+      "loss": 2.1297,
+      "step": 24825
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.121752238630873e-05,
+      "loss": 2.1102,
+      "step": 24830
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.117724803706003e-05,
+      "loss": 2.0883,
+      "step": 24835
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.113697513005758e-05,
+      "loss": 2.1221,
+      "step": 24840
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.109670367188474e-05,
+      "loss": 2.1086,
+      "step": 24845
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.105643366912464e-05,
+      "loss": 2.1466,
+      "step": 24850
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.101616512836017e-05,
+      "loss": 2.0825,
+      "step": 24855
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.0975898056174e-05,
+      "loss": 2.1403,
+      "step": 24860
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.093563245914854e-05,
+      "loss": 2.1357,
+      "step": 24865
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.089536834386598e-05,
+      "loss": 2.088,
+      "step": 24870
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.085510571690823e-05,
+      "loss": 2.1856,
+      "step": 24875
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.0814844584857e-05,
+      "loss": 2.1013,
+      "step": 24880
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.077458495429372e-05,
+      "loss": 2.1058,
+      "step": 24885
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.073432683179961e-05,
+      "loss": 2.1049,
+      "step": 24890
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.069407022395561e-05,
+      "loss": 2.1641,
+      "step": 24895
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.065381513734243e-05,
+      "loss": 2.1631,
+      "step": 24900
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.061356157854053e-05,
+      "loss": 2.1179,
+      "step": 24905
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.057330955413012e-05,
+      "loss": 2.1744,
+      "step": 24910
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.053305907069114e-05,
+      "loss": 2.1128,
+      "step": 24915
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.04928101348033e-05,
+      "loss": 2.0845,
+      "step": 24920
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.045256275304613e-05,
+      "loss": 2.0877,
+      "step": 24925
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.041231693199869e-05,
+      "loss": 2.0911,
+      "step": 24930
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.037207267824001e-05,
+      "loss": 2.1198,
+      "step": 24935
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.033182999834875e-05,
+      "loss": 2.1163,
+      "step": 24940
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.029158889890336e-05,
+      "loss": 2.0934,
+      "step": 24945
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.0251349386482e-05,
+      "loss": 2.1233,
+      "step": 24950
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.021111146766257e-05,
+      "loss": 2.0593,
+      "step": 24955
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.017087514902275e-05,
+      "loss": 2.0972,
+      "step": 24960
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.01306404371399e-05,
+      "loss": 2.1049,
+      "step": 24965
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.009040733859117e-05,
+      "loss": 2.1268,
+      "step": 24970
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.005017585995341e-05,
+      "loss": 2.0961,
+      "step": 24975
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.000994600780322e-05,
+      "loss": 2.0754,
+      "step": 24980
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.996971778871695e-05,
+      "loss": 2.1573,
+      "step": 24985
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.992949120927065e-05,
+      "loss": 2.0683,
+      "step": 24990
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.988926627604011e-05,
+      "loss": 2.1297,
+      "step": 24995
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.984904299560087e-05,
+      "loss": 2.0843,
+      "step": 25000
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.980882137452817e-05,
+      "loss": 2.0636,
+      "step": 25005
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.976860141939703e-05,
+      "loss": 2.0951,
+      "step": 25010
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.972838313678215e-05,
+      "loss": 2.0432,
+      "step": 25015
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.968816653325798e-05,
+      "loss": 2.1193,
+      "step": 25020
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.964795161539867e-05,
+      "loss": 2.0768,
+      "step": 25025
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.960773838977811e-05,
+      "loss": 2.1389,
+      "step": 25030
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.956752686296994e-05,
+      "loss": 2.1372,
+      "step": 25035
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.952731704154746e-05,
+      "loss": 2.1616,
+      "step": 25040
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.948710893208376e-05,
+      "loss": 2.1022,
+      "step": 25045
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.94469025411516e-05,
+      "loss": 2.1229,
+      "step": 25050
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.940669787532349e-05,
+      "loss": 2.1266,
+      "step": 25055
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.936649494117165e-05,
+      "loss": 2.1557,
+      "step": 25060
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.9326293745268e-05,
+      "loss": 2.0867,
+      "step": 25065
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.928609429418418e-05,
+      "loss": 2.1095,
+      "step": 25070
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.924589659449158e-05,
+      "loss": 2.086,
+      "step": 25075
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.920570065276124e-05,
+      "loss": 2.1039,
+      "step": 25080
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.9165506475564e-05,
+      "loss": 2.0964,
+      "step": 25085
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.912531406947031e-05,
+      "loss": 2.095,
+      "step": 25090
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.908512344105041e-05,
+      "loss": 2.108,
+      "step": 25095
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.904493459687421e-05,
+      "loss": 2.0912,
+      "step": 25100
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.90047475435114e-05,
+      "loss": 2.1486,
+      "step": 25105
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.896456228753122e-05,
+      "loss": 2.0605,
+      "step": 25110
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.892437883550274e-05,
+      "loss": 2.1001,
+      "step": 25115
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.888419719399472e-05,
+      "loss": 2.1228,
+      "step": 25120
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.884401736957562e-05,
+      "loss": 2.0732,
+      "step": 25125
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.880383936881357e-05,
+      "loss": 2.0935,
+      "step": 25130
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.876366319827646e-05,
+      "loss": 2.1366,
+      "step": 25135
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.872348886453185e-05,
+      "loss": 2.0875,
+      "step": 25140
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.868331637414695e-05,
+      "loss": 2.1388,
+      "step": 25145
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.864314573368876e-05,
+      "loss": 2.1111,
+      "step": 25150
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.860297694972391e-05,
+      "loss": 2.082,
+      "step": 25155
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.856281002881876e-05,
+      "loss": 2.0948,
+      "step": 25160
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.852264497753934e-05,
+      "loss": 2.1316,
+      "step": 25165
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.848248180245143e-05,
+      "loss": 2.0923,
+      "step": 25170
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.844232051012043e-05,
+      "loss": 2.077,
+      "step": 25175
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.840216110711146e-05,
+      "loss": 2.1062,
+      "step": 25180
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.836200359998936e-05,
+      "loss": 2.078,
+      "step": 25185
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.832184799531861e-05,
+      "loss": 2.0878,
+      "step": 25190
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.828169429966345e-05,
+      "loss": 2.1302,
+      "step": 25195
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.824154251958771e-05,
+      "loss": 2.1612,
+      "step": 25200
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.820139266165499e-05,
+      "loss": 2.0522,
+      "step": 25205
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.816124473242853e-05,
+      "loss": 2.1189,
+      "step": 25210
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.81210987384713e-05,
+      "loss": 2.1225,
+      "step": 25215
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.80809546863459e-05,
+      "loss": 2.1166,
+      "step": 25220
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.804081258261466e-05,
+      "loss": 2.0787,
+      "step": 25225
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.800067243383952e-05,
+      "loss": 2.1412,
+      "step": 25230
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.79605342465822e-05,
+      "loss": 2.097,
+      "step": 25235
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.792039802740403e-05,
+      "loss": 2.1135,
+      "step": 25240
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.788026378286604e-05,
+      "loss": 2.1407,
+      "step": 25245
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.784013151952891e-05,
+      "loss": 2.1206,
+      "step": 25250
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.780000124395305e-05,
+      "loss": 2.1218,
+      "step": 25255
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.775987296269852e-05,
+      "loss": 2.1253,
+      "step": 25260
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.7719746682325e-05,
+      "loss": 2.1022,
+      "step": 25265
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.767962240939194e-05,
+      "loss": 2.0672,
+      "step": 25270
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.763950015045842e-05,
+      "loss": 2.1088,
+      "step": 25275
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.759937991208316e-05,
+      "loss": 2.1163,
+      "step": 25280
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.755926170082458e-05,
+      "loss": 2.0554,
+      "step": 25285
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.75191455232408e-05,
+      "loss": 2.1367,
+      "step": 25290
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.747903138588949e-05,
+      "loss": 2.122,
+      "step": 25295
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.74389192953281e-05,
+      "loss": 2.1256,
+      "step": 25300
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.739880925811375e-05,
+      "loss": 2.1281,
+      "step": 25305
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.735870128080314e-05,
+      "loss": 2.1033,
+      "step": 25310
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.73185953699527e-05,
+      "loss": 2.1231,
+      "step": 25315
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.727849153211849e-05,
+      "loss": 2.1218,
+      "step": 25320
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.723838977385628e-05,
+      "loss": 2.0517,
+      "step": 25325
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.719829010172141e-05,
+      "loss": 2.1164,
+      "step": 25330
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.715819252226898e-05,
+      "loss": 2.0777,
+      "step": 25335
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.711809704205368e-05,
+      "loss": 2.1293,
+      "step": 25340
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.707800366762988e-05,
+      "loss": 2.1216,
+      "step": 25345
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.703791240555158e-05,
+      "loss": 2.1182,
+      "step": 25350
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.699782326237249e-05,
+      "loss": 2.1183,
+      "step": 25355
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.695773624464594e-05,
+      "loss": 2.0897,
+      "step": 25360
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.69176513589249e-05,
+      "loss": 2.1265,
+      "step": 25365
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.687756861176199e-05,
+      "loss": 2.1149,
+      "step": 25370
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.683748800970953e-05,
+      "loss": 2.1101,
+      "step": 25375
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.679740955931943e-05,
+      "loss": 2.088,
+      "step": 25380
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.675733326714329e-05,
+      "loss": 2.1126,
+      "step": 25385
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.671725913973233e-05,
+      "loss": 2.1395,
+      "step": 25390
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.667718718363743e-05,
+      "loss": 2.114,
+      "step": 25395
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.663711740540912e-05,
+      "loss": 2.1065,
+      "step": 25400
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.659704981159754e-05,
+      "loss": 2.1068,
+      "step": 25405
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.655698440875253e-05,
+      "loss": 2.0586,
+      "step": 25410
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.651692120342352e-05,
+      "loss": 2.1144,
+      "step": 25415
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.64768602021596e-05,
+      "loss": 2.112,
+      "step": 25420
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.643680141150951e-05,
+      "loss": 2.0977,
+      "step": 25425
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.639674483802163e-05,
+      "loss": 2.1093,
+      "step": 25430
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.635669048824396e-05,
+      "loss": 2.1334,
+      "step": 25435
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.631663836872411e-05,
+      "loss": 2.1215,
+      "step": 25440
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.62765884860094e-05,
+      "loss": 2.1239,
+      "step": 25445
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.623654084664674e-05,
+      "loss": 2.074,
+      "step": 25450
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.619649545718266e-05,
+      "loss": 2.0952,
+      "step": 25455
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.615645232416333e-05,
+      "loss": 2.1315,
+      "step": 25460
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.61164114541346e-05,
+      "loss": 2.1137,
+      "step": 25465
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.60763728536419e-05,
+      "loss": 2.1669,
+      "step": 25470
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.603633652923026e-05,
+      "loss": 2.1003,
+      "step": 25475
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.599630248744442e-05,
+      "loss": 2.1083,
+      "step": 25480
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.595627073482865e-05,
+      "loss": 2.157,
+      "step": 25485
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.591624127792697e-05,
+      "loss": 2.1189,
+      "step": 25490
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.587621412328291e-05,
+      "loss": 2.0517,
+      "step": 25495
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.583618927743967e-05,
+      "loss": 2.0891,
+      "step": 25500
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.579616674694007e-05,
+      "loss": 2.1019,
+      "step": 25505
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.575614653832658e-05,
+      "loss": 2.199,
+      "step": 25510
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.571612865814123e-05,
+      "loss": 2.1147,
+      "step": 25515
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.567611311292572e-05,
+      "loss": 2.0954,
+      "step": 25520
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.563609990922136e-05,
+      "loss": 2.1366,
+      "step": 25525
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.559608905356904e-05,
+      "loss": 2.1095,
+      "step": 25530
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.55560805525093e-05,
+      "loss": 2.1531,
+      "step": 25535
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.551607441258231e-05,
+      "loss": 2.1013,
+      "step": 25540
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.547607064032781e-05,
+      "loss": 2.0872,
+      "step": 25545
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.543606924228518e-05,
+      "loss": 2.1016,
+      "step": 25550
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.53960702249934e-05,
+      "loss": 2.083,
+      "step": 25555
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.53560735949911e-05,
+      "loss": 2.1748,
+      "step": 25560
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.531607935881647e-05,
+      "loss": 2.1366,
+      "step": 25565
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.52760875230073e-05,
+      "loss": 2.1275,
+      "step": 25570
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.523609809410107e-05,
+      "loss": 2.1756,
+      "step": 25575
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.519611107863476e-05,
+      "loss": 2.1276,
+      "step": 25580
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.515612648314504e-05,
+      "loss": 2.1488,
+      "step": 25585
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.511614431416814e-05,
+      "loss": 2.0951,
+      "step": 25590
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.507616457823988e-05,
+      "loss": 2.1266,
+      "step": 25595
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.503618728189575e-05,
+      "loss": 2.0753,
+      "step": 25600
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.499621243167076e-05,
+      "loss": 2.0906,
+      "step": 25605
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.495624003409959e-05,
+      "loss": 2.1266,
+      "step": 25610
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.491627009571646e-05,
+      "loss": 2.0273,
+      "step": 25615
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.487630262305524e-05,
+      "loss": 2.1196,
+      "step": 25620
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.483633762264935e-05,
+      "loss": 2.0889,
+      "step": 25625
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.479637510103183e-05,
+      "loss": 2.104,
+      "step": 25630
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.475641506473532e-05,
+      "loss": 2.1048,
+      "step": 25635
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.471645752029203e-05,
+      "loss": 2.1212,
+      "step": 25640
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.467650247423383e-05,
+      "loss": 2.0876,
+      "step": 25645
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.46365499330921e-05,
+      "loss": 2.1045,
+      "step": 25650
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.45965999033978e-05,
+      "loss": 2.1018,
+      "step": 25655
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.455665239168158e-05,
+      "loss": 2.0721,
+      "step": 25660
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.451670740447357e-05,
+      "loss": 2.1079,
+      "step": 25665
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.447676494830356e-05,
+      "loss": 2.1088,
+      "step": 25670
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.443682502970092e-05,
+      "loss": 2.1465,
+      "step": 25675
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.439688765519457e-05,
+      "loss": 2.047,
+      "step": 25680
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.435695283131301e-05,
+      "loss": 2.119,
+      "step": 25685
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.431702056458436e-05,
+      "loss": 2.0793,
+      "step": 25690
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.427709086153635e-05,
+      "loss": 2.137,
+      "step": 25695
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.423716372869618e-05,
+      "loss": 2.0464,
+      "step": 25700
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.419723917259074e-05,
+      "loss": 2.1012,
+      "step": 25705
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.415731719974646e-05,
+      "loss": 2.1351,
+      "step": 25710
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.411739781668932e-05,
+      "loss": 2.0576,
+      "step": 25715
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.407748102994491e-05,
+      "loss": 2.131,
+      "step": 25720
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.403756684603838e-05,
+      "loss": 2.1741,
+      "step": 25725
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.399765527149448e-05,
+      "loss": 2.1166,
+      "step": 25730
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.39577463128375e-05,
+      "loss": 2.0992,
+      "step": 25735
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.391783997659131e-05,
+      "loss": 2.1115,
+      "step": 25740
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.387793626927936e-05,
+      "loss": 2.1011,
+      "step": 25745
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.383803519742469e-05,
+      "loss": 2.1105,
+      "step": 25750
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3203125,
+      "learning_rate": 8.379813676754987e-05,
+      "loss": 2.1747,
+      "step": 25755
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.375824098617705e-05,
+      "loss": 2.0877,
+      "step": 25760
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.371834785982797e-05,
+      "loss": 2.1215,
+      "step": 25765
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.36784573950239e-05,
+      "loss": 2.1861,
+      "step": 25770
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.36385695982857e-05,
+      "loss": 2.1083,
+      "step": 25775
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.35986844761338e-05,
+      "loss": 2.1687,
+      "step": 25780
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.355880203508815e-05,
+      "loss": 2.1012,
+      "step": 25785
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.351892228166834e-05,
+      "loss": 2.0954,
+      "step": 25790
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.347904522239342e-05,
+      "loss": 2.0591,
+      "step": 25795
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.343917086378208e-05,
+      "loss": 2.0646,
+      "step": 25800
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.339929921235255e-05,
+      "loss": 2.1192,
+      "step": 25805
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.335943027462257e-05,
+      "loss": 2.1244,
+      "step": 25810
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.331956405710951e-05,
+      "loss": 2.1482,
+      "step": 25815
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.327970056633025e-05,
+      "loss": 2.1097,
+      "step": 25820
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.323983980880123e-05,
+      "loss": 2.1113,
+      "step": 25825
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.319998179103846e-05,
+      "loss": 2.1089,
+      "step": 25830
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.316012651955751e-05,
+      "loss": 2.127,
+      "step": 25835
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.312027400087343e-05,
+      "loss": 2.12,
+      "step": 25840
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.308042424150087e-05,
+      "loss": 2.1526,
+      "step": 25845
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.304057724795404e-05,
+      "loss": 2.1307,
+      "step": 25850
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.300073302674671e-05,
+      "loss": 2.0873,
+      "step": 25855
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.296089158439216e-05,
+      "loss": 2.1089,
+      "step": 25860
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 8.292105292740324e-05,
+      "loss": 2.1284,
+      "step": 25865
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.288121706229232e-05,
+      "loss": 2.1303,
+      "step": 25870
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.284138399557133e-05,
+      "loss": 2.106,
+      "step": 25875
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.280155373375173e-05,
+      "loss": 2.1472,
+      "step": 25880
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.276172628334457e-05,
+      "loss": 2.0862,
+      "step": 25885
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.272190165086038e-05,
+      "loss": 2.1671,
+      "step": 25890
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.268207984280924e-05,
+      "loss": 2.1102,
+      "step": 25895
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.264226086570078e-05,
+      "loss": 2.0499,
+      "step": 25900
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.260244472604419e-05,
+      "loss": 2.0717,
+      "step": 25905
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.256263143034817e-05,
+      "loss": 2.0843,
+      "step": 25910
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.252282098512094e-05,
+      "loss": 2.0963,
+      "step": 25915
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.32421875,
+      "learning_rate": 8.248301339687027e-05,
+      "loss": 2.121,
+      "step": 25920
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.24432086721035e-05,
+      "loss": 2.1367,
+      "step": 25925
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.240340681732746e-05,
+      "loss": 2.1013,
+      "step": 25930
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.236360783904849e-05,
+      "loss": 2.1694,
+      "step": 25935
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.232381174377249e-05,
+      "loss": 2.0853,
+      "step": 25940
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.228401853800491e-05,
+      "loss": 2.1106,
+      "step": 25945
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.224422822825069e-05,
+      "loss": 2.1163,
+      "step": 25950
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.220444082101432e-05,
+      "loss": 2.0288,
+      "step": 25955
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.216465632279981e-05,
+      "loss": 2.1426,
+      "step": 25960
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.21248747401107e-05,
+      "loss": 2.097,
+      "step": 25965
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.208509607945e-05,
+      "loss": 2.0659,
+      "step": 25970
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.204532034732033e-05,
+      "loss": 2.108,
+      "step": 25975
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.200554755022377e-05,
+      "loss": 2.1235,
+      "step": 25980
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.196577769466194e-05,
+      "loss": 2.1193,
+      "step": 25985
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.192601078713598e-05,
+      "loss": 2.107,
+      "step": 25990
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.188624683414657e-05,
+      "loss": 2.0861,
+      "step": 25995
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.184648584219383e-05,
+      "loss": 2.126,
+      "step": 26000
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.180672781777749e-05,
+      "loss": 2.0688,
+      "step": 26005
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.176697276739673e-05,
+      "loss": 2.0893,
+      "step": 26010
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.172722069755031e-05,
+      "loss": 2.0798,
+      "step": 26015
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.16874716147364e-05,
+      "loss": 2.0698,
+      "step": 26020
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.164772552545279e-05,
+      "loss": 2.1219,
+      "step": 26025
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.160798243619668e-05,
+      "loss": 2.1122,
+      "step": 26030
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.156824235346488e-05,
+      "loss": 2.1185,
+      "step": 26035
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.152850528375366e-05,
+      "loss": 2.0763,
+      "step": 26040
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.148877123355876e-05,
+      "loss": 2.1234,
+      "step": 26045
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.144904020937551e-05,
+      "loss": 2.0839,
+      "step": 26050
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.140931221769867e-05,
+      "loss": 2.0747,
+      "step": 26055
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.136958726502255e-05,
+      "loss": 2.1306,
+      "step": 26060
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.132986535784093e-05,
+      "loss": 2.07,
+      "step": 26065
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.129014650264712e-05,
+      "loss": 2.1526,
+      "step": 26070
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.125043070593393e-05,
+      "loss": 2.0725,
+      "step": 26075
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.121071797419364e-05,
+      "loss": 2.1092,
+      "step": 26080
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.117100831391806e-05,
+      "loss": 2.1831,
+      "step": 26085
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.113130173159848e-05,
+      "loss": 2.0863,
+      "step": 26090
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.109159823372571e-05,
+      "loss": 2.1107,
+      "step": 26095
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.105189782679004e-05,
+      "loss": 2.1749,
+      "step": 26100
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.101220051728125e-05,
+      "loss": 2.1496,
+      "step": 26105
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.097250631168862e-05,
+      "loss": 2.0973,
+      "step": 26110
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.09328152165009e-05,
+      "loss": 2.0709,
+      "step": 26115
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.089312723820637e-05,
+      "loss": 2.0844,
+      "step": 26120
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.085344238329279e-05,
+      "loss": 2.1903,
+      "step": 26125
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.08137606582474e-05,
+      "loss": 2.1073,
+      "step": 26130
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.077408206955692e-05,
+      "loss": 2.1169,
+      "step": 26135
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.073440662370758e-05,
+      "loss": 2.1232,
+      "step": 26140
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.069473432718507e-05,
+      "loss": 2.139,
+      "step": 26145
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.06550651864746e-05,
+      "loss": 2.0682,
+      "step": 26150
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.061539920806083e-05,
+      "loss": 2.1613,
+      "step": 26155
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.057573639842792e-05,
+      "loss": 2.1393,
+      "step": 26160
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.053607676405951e-05,
+      "loss": 2.1094,
+      "step": 26165
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.049642031143871e-05,
+      "loss": 2.0842,
+      "step": 26170
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.045676704704814e-05,
+      "loss": 2.158,
+      "step": 26175
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.041711697736986e-05,
+      "loss": 2.0987,
+      "step": 26180
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.037747010888543e-05,
+      "loss": 2.0494,
+      "step": 26185
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.033782644807591e-05,
+      "loss": 2.1369,
+      "step": 26190
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.029818600142178e-05,
+      "loss": 2.1714,
+      "step": 26195
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.025854877540303e-05,
+      "loss": 2.1302,
+      "step": 26200
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.02189147764991e-05,
+      "loss": 2.1227,
+      "step": 26205
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.017928401118892e-05,
+      "loss": 2.1019,
+      "step": 26210
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.013965648595091e-05,
+      "loss": 2.0979,
+      "step": 26215
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.010003220726295e-05,
+      "loss": 2.125,
+      "step": 26220
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.006041118160237e-05,
+      "loss": 2.0773,
+      "step": 26225
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.002079341544597e-05,
+      "loss": 2.0918,
+      "step": 26230
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.998117891527003e-05,
+      "loss": 2.1439,
+      "step": 26235
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.994156768755029e-05,
+      "loss": 2.1664,
+      "step": 26240
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.990195973876195e-05,
+      "loss": 2.0999,
+      "step": 26245
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.986235507537972e-05,
+      "loss": 2.0943,
+      "step": 26250
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.982275370387767e-05,
+      "loss": 2.1175,
+      "step": 26255
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3671875,
+      "learning_rate": 7.978315563072945e-05,
+      "loss": 2.1143,
+      "step": 26260
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.974356086240808e-05,
+      "loss": 2.1257,
+      "step": 26265
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.970396940538612e-05,
+      "loss": 2.103,
+      "step": 26270
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.966438126613549e-05,
+      "loss": 2.0889,
+      "step": 26275
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.962479645112766e-05,
+      "loss": 2.0708,
+      "step": 26280
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.958521496683351e-05,
+      "loss": 2.0585,
+      "step": 26285
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.954563681972339e-05,
+      "loss": 2.0842,
+      "step": 26290
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.950606201626709e-05,
+      "loss": 2.1064,
+      "step": 26295
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.946649056293386e-05,
+      "loss": 2.0348,
+      "step": 26300
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.942692246619242e-05,
+      "loss": 2.1207,
+      "step": 26305
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.938735773251091e-05,
+      "loss": 2.1055,
+      "step": 26310
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.934779636835694e-05,
+      "loss": 2.1685,
+      "step": 26315
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.930823838019757e-05,
+      "loss": 2.1282,
+      "step": 26320
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.926868377449932e-05,
+      "loss": 2.0952,
+      "step": 26325
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.92291325577281e-05,
+      "loss": 2.0745,
+      "step": 26330
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.918958473634934e-05,
+      "loss": 2.0818,
+      "step": 26335
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.915004031682789e-05,
+      "loss": 2.1348,
+      "step": 26340
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.911049930562799e-05,
+      "loss": 2.0653,
+      "step": 26345
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.907096170921342e-05,
+      "loss": 2.0872,
+      "step": 26350
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.903142753404732e-05,
+      "loss": 2.1497,
+      "step": 26355
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.899189678659231e-05,
+      "loss": 2.1086,
+      "step": 26360
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.895236947331044e-05,
+      "loss": 2.134,
+      "step": 26365
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.89128456006632e-05,
+      "loss": 2.1159,
+      "step": 26370
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.887332517511156e-05,
+      "loss": 2.1131,
+      "step": 26375
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.883380820311578e-05,
+      "loss": 2.127,
+      "step": 26380
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.879429469113576e-05,
+      "loss": 2.1191,
+      "step": 26385
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.875478464563065e-05,
+      "loss": 2.133,
+      "step": 26390
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.871527807305919e-05,
+      "loss": 2.142,
+      "step": 26395
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.867577497987942e-05,
+      "loss": 2.0964,
+      "step": 26400
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.863627537254891e-05,
+      "loss": 2.1559,
+      "step": 26405
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.859677925752463e-05,
+      "loss": 2.1079,
+      "step": 26410
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.855728664126292e-05,
+      "loss": 2.0528,
+      "step": 26415
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.851779753021963e-05,
+      "loss": 2.1213,
+      "step": 26420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.847831193085e-05,
+      "loss": 2.159,
+      "step": 26425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.84388298496087e-05,
+      "loss": 2.0848,
+      "step": 26430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.839935129294984e-05,
+      "loss": 2.1136,
+      "step": 26435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.835987626732693e-05,
+      "loss": 2.159,
+      "step": 26440
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.83204047791929e-05,
+      "loss": 2.1423,
+      "step": 26445
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.828093683500008e-05,
+      "loss": 2.1304,
+      "step": 26450
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.824147244120034e-05,
+      "loss": 2.0842,
+      "step": 26455
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.820201160424484e-05,
+      "loss": 2.0474,
+      "step": 26460
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.816255433058419e-05,
+      "loss": 2.1217,
+      "step": 26465
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.812310062666844e-05,
+      "loss": 2.1836,
+      "step": 26470
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.808365049894706e-05,
+      "loss": 2.1555,
+      "step": 26475
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.804420395386892e-05,
+      "loss": 2.1467,
+      "step": 26480
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.800476099788227e-05,
+      "loss": 2.1436,
+      "step": 26485
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.796532163743487e-05,
+      "loss": 2.1505,
+      "step": 26490
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.792588587897378e-05,
+      "loss": 2.1426,
+      "step": 26495
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.788645372894553e-05,
+      "loss": 2.1159,
+      "step": 26500
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.78470251937961e-05,
+      "loss": 2.0916,
+      "step": 26505
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.780760027997079e-05,
+      "loss": 2.1462,
+      "step": 26510
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.776817899391433e-05,
+      "loss": 2.1144,
+      "step": 26515
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.772876134207093e-05,
+      "loss": 2.1258,
+      "step": 26520
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.768934733088413e-05,
+      "loss": 2.0852,
+      "step": 26525
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.764993696679688e-05,
+      "loss": 2.1266,
+      "step": 26530
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.761053025625158e-05,
+      "loss": 2.0716,
+      "step": 26535
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.757112720569e-05,
+      "loss": 2.1082,
+      "step": 26540
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.75317278215533e-05,
+      "loss": 2.0825,
+      "step": 26545
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.749233211028206e-05,
+      "loss": 2.1437,
+      "step": 26550
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.745294007831627e-05,
+      "loss": 2.1629,
+      "step": 26555
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.741355173209534e-05,
+      "loss": 2.1321,
+      "step": 26560
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.737416707805797e-05,
+      "loss": 2.0677,
+      "step": 26565
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.733478612264235e-05,
+      "loss": 2.1119,
+      "step": 26570
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.729540887228606e-05,
+      "loss": 2.1622,
+      "step": 26575
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.725603533342605e-05,
+      "loss": 2.1485,
+      "step": 26580
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.721666551249868e-05,
+      "loss": 2.1493,
+      "step": 26585
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.717729941593968e-05,
+      "loss": 2.1297,
+      "step": 26590
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.713793705018421e-05,
+      "loss": 2.1517,
+      "step": 26595
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.709857842166676e-05,
+      "loss": 2.1131,
+      "step": 26600
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.705922353682129e-05,
+      "loss": 2.094,
+      "step": 26605
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.701987240208107e-05,
+      "loss": 2.1689,
+      "step": 26610
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.698052502387881e-05,
+      "loss": 2.1419,
+      "step": 26615
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.694118140864655e-05,
+      "loss": 2.1495,
+      "step": 26620
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.690184156281579e-05,
+      "loss": 2.1274,
+      "step": 26625
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.686250549281735e-05,
+      "loss": 2.1449,
+      "step": 26630
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.682317320508148e-05,
+      "loss": 2.102,
+      "step": 26635
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.678384470603777e-05,
+      "loss": 2.1215,
+      "step": 26640
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.67445200021152e-05,
+      "loss": 2.1395,
+      "step": 26645
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.670519909974218e-05,
+      "loss": 2.0977,
+      "step": 26650
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.666588200534642e-05,
+      "loss": 2.1199,
+      "step": 26655
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.662656872535504e-05,
+      "loss": 2.096,
+      "step": 26660
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.658725926619458e-05,
+      "loss": 2.101,
+      "step": 26665
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.654795363429088e-05,
+      "loss": 2.1012,
+      "step": 26670
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.650865183606922e-05,
+      "loss": 2.1191,
+      "step": 26675
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.646935387795419e-05,
+      "loss": 2.0792,
+      "step": 26680
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.643005976636983e-05,
+      "loss": 2.1181,
+      "step": 26685
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.639076950773949e-05,
+      "loss": 2.0782,
+      "step": 26690
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.63514831084859e-05,
+      "loss": 2.1109,
+      "step": 26695
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.631220057503116e-05,
+      "loss": 2.1095,
+      "step": 26700
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.627292191379677e-05,
+      "loss": 2.1225,
+      "step": 26705
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.623364713120355e-05,
+      "loss": 2.0973,
+      "step": 26710
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.619437623367174e-05,
+      "loss": 2.1111,
+      "step": 26715
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.615510922762088e-05,
+      "loss": 2.1028,
+      "step": 26720
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.611584611946991e-05,
+      "loss": 2.1078,
+      "step": 26725
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.607658691563717e-05,
+      "loss": 2.1611,
+      "step": 26730
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.603733162254025e-05,
+      "loss": 2.0865,
+      "step": 26735
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.599808024659627e-05,
+      "loss": 2.0912,
+      "step": 26740
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.595883279422153e-05,
+      "loss": 2.0867,
+      "step": 26745
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.591958927183179e-05,
+      "loss": 2.1203,
+      "step": 26750
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.588034968584214e-05,
+      "loss": 2.0936,
+      "step": 26755
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.584111404266708e-05,
+      "loss": 2.1391,
+      "step": 26760
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.580188234872039e-05,
+      "loss": 2.081,
+      "step": 26765
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.576265461041521e-05,
+      "loss": 2.1631,
+      "step": 26770
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.572343083416411e-05,
+      "loss": 2.1258,
+      "step": 26775
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.568421102637892e-05,
+      "loss": 2.0925,
+      "step": 26780
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.56449951934709e-05,
+      "loss": 2.0912,
+      "step": 26785
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.560578334185058e-05,
+      "loss": 2.0644,
+      "step": 26790
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.55665754779279e-05,
+      "loss": 2.1052,
+      "step": 26795
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.552737160811214e-05,
+      "loss": 2.1255,
+      "step": 26800
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.54881717388119e-05,
+      "loss": 2.1118,
+      "step": 26805
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.544897587643515e-05,
+      "loss": 2.0934,
+      "step": 26810
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.540978402738919e-05,
+      "loss": 2.0904,
+      "step": 26815
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.537059619808068e-05,
+      "loss": 2.0793,
+      "step": 26820
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.533141239491563e-05,
+      "loss": 2.1017,
+      "step": 26825
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.529223262429934e-05,
+      "loss": 2.1309,
+      "step": 26830
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.525305689263652e-05,
+      "loss": 2.0737,
+      "step": 26835
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.521388520633116e-05,
+      "loss": 2.0743,
+      "step": 26840
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.517471757178664e-05,
+      "loss": 2.1256,
+      "step": 26845
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.513555399540563e-05,
+      "loss": 2.1143,
+      "step": 26850
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.509639448359018e-05,
+      "loss": 2.0976,
+      "step": 26855
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.505723904274164e-05,
+      "loss": 2.0918,
+      "step": 26860
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.501808767926071e-05,
+      "loss": 2.1623,
+      "step": 26865
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.497894039954743e-05,
+      "loss": 2.132,
+      "step": 26870
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.493979721000117e-05,
+      "loss": 2.1213,
+      "step": 26875
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.490065811702063e-05,
+      "loss": 2.0994,
+      "step": 26880
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.486152312700382e-05,
+      "loss": 2.1078,
+      "step": 26885
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.482239224634811e-05,
+      "loss": 2.0999,
+      "step": 26890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.478326548145018e-05,
+      "loss": 2.097,
+      "step": 26895
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.474414283870605e-05,
+      "loss": 2.1641,
+      "step": 26900
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.470502432451103e-05,
+      "loss": 2.0743,
+      "step": 26905
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.466590994525984e-05,
+      "loss": 2.0625,
+      "step": 26910
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.46267997073464e-05,
+      "loss": 2.144,
+      "step": 26915
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.45876936171641e-05,
+      "loss": 2.0602,
+      "step": 26920
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.454859168110548e-05,
+      "loss": 2.1108,
+      "step": 26925
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.450949390556257e-05,
+      "loss": 2.1224,
+      "step": 26930
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.44704002969266e-05,
+      "loss": 2.1246,
+      "step": 26935
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.443131086158817e-05,
+      "loss": 2.0616,
+      "step": 26940
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.43922256059372e-05,
+      "loss": 2.1235,
+      "step": 26945
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.435314453636291e-05,
+      "loss": 2.0541,
+      "step": 26950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.431406765925384e-05,
+      "loss": 2.1211,
+      "step": 26955
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.427499498099786e-05,
+      "loss": 2.0566,
+      "step": 26960
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.423592650798213e-05,
+      "loss": 2.0892,
+      "step": 26965
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.419686224659314e-05,
+      "loss": 2.0703,
+      "step": 26970
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.415780220321667e-05,
+      "loss": 2.1224,
+      "step": 26975
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.411874638423785e-05,
+      "loss": 2.1785,
+      "step": 26980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.407969479604106e-05,
+      "loss": 2.0637,
+      "step": 26985
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.404064744501007e-05,
+      "loss": 2.0753,
+      "step": 26990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.400160433752787e-05,
+      "loss": 2.1239,
+      "step": 26995
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.396256547997682e-05,
+      "loss": 2.0954,
+      "step": 27000
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.392353087873854e-05,
+      "loss": 2.1042,
+      "step": 27005
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.3884500540194e-05,
+      "loss": 2.0674,
+      "step": 27010
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.384547447072343e-05,
+      "loss": 2.1036,
+      "step": 27015
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.38064526767064e-05,
+      "loss": 2.1717,
+      "step": 27020
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.376743516452175e-05,
+      "loss": 2.1141,
+      "step": 27025
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.372842194054763e-05,
+      "loss": 2.1398,
+      "step": 27030
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.368941301116152e-05,
+      "loss": 2.0353,
+      "step": 27035
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.365040838274013e-05,
+      "loss": 2.0954,
+      "step": 27040
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.361140806165952e-05,
+      "loss": 2.1389,
+      "step": 27045
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.357241205429504e-05,
+      "loss": 2.0355,
+      "step": 27050
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.353342036702133e-05,
+      "loss": 2.0824,
+      "step": 27055
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.349443300621233e-05,
+      "loss": 2.0861,
+      "step": 27060
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.345544997824123e-05,
+      "loss": 2.0664,
+      "step": 27065
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.341647128948058e-05,
+      "loss": 2.1066,
+      "step": 27070
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.337749694630216e-05,
+      "loss": 2.1243,
+      "step": 27075
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.333852695507708e-05,
+      "loss": 2.0653,
+      "step": 27080
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.329956132217573e-05,
+      "loss": 2.1475,
+      "step": 27085
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.326060005396775e-05,
+      "loss": 2.1013,
+      "step": 27090
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.322164315682215e-05,
+      "loss": 2.156,
+      "step": 27095
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.318269063710715e-05,
+      "loss": 2.1407,
+      "step": 27100
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.31437425011903e-05,
+      "loss": 2.0924,
+      "step": 27105
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.310479875543835e-05,
+      "loss": 2.0753,
+      "step": 27110
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.306585940621743e-05,
+      "loss": 2.0988,
+      "step": 27115
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.302692445989293e-05,
+      "loss": 2.127,
+      "step": 27120
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.29879939228295e-05,
+      "loss": 2.0657,
+      "step": 27125
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.294906780139104e-05,
+      "loss": 2.1473,
+      "step": 27130
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.291014610194083e-05,
+      "loss": 2.0872,
+      "step": 27135
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.28712288308413e-05,
+      "loss": 2.1125,
+      "step": 27140
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.283231599445422e-05,
+      "loss": 2.0863,
+      "step": 27145
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.279340759914068e-05,
+      "loss": 2.1591,
+      "step": 27150
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.275450365126096e-05,
+      "loss": 2.1451,
+      "step": 27155
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.271560415717464e-05,
+      "loss": 2.1071,
+      "step": 27160
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.267670912324062e-05,
+      "loss": 2.0952,
+      "step": 27165
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.263781855581699e-05,
+      "loss": 2.0883,
+      "step": 27170
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.259893246126117e-05,
+      "loss": 2.1387,
+      "step": 27175
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.256005084592983e-05,
+      "loss": 2.0978,
+      "step": 27180
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.25211737161789e-05,
+      "loss": 2.1177,
+      "step": 27185
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.24823010783636e-05,
+      "loss": 2.0655,
+      "step": 27190
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.244343293883837e-05,
+      "loss": 2.1222,
+      "step": 27195
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.2404569303957e-05,
+      "loss": 2.1509,
+      "step": 27200
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.236571018007243e-05,
+      "loss": 2.1079,
+      "step": 27205
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.232685557353693e-05,
+      "loss": 2.0392,
+      "step": 27210
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.228800549070206e-05,
+      "loss": 2.0582,
+      "step": 27215
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.224915993791859e-05,
+      "loss": 2.0972,
+      "step": 27220
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.221031892153653e-05,
+      "loss": 2.0965,
+      "step": 27225
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.217148244790523e-05,
+      "loss": 2.079,
+      "step": 27230
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.213265052337321e-05,
+      "loss": 2.1277,
+      "step": 27235
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.20938231542883e-05,
+      "loss": 2.111,
+      "step": 27240
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.205500034699758e-05,
+      "loss": 2.1254,
+      "step": 27245
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.201618210784737e-05,
+      "loss": 2.0993,
+      "step": 27250
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.197736844318322e-05,
+      "loss": 2.094,
+      "step": 27255
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.193855935934999e-05,
+      "loss": 2.1207,
+      "step": 27260
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.189975486269176e-05,
+      "loss": 2.1185,
+      "step": 27265
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.186095495955186e-05,
+      "loss": 2.1166,
+      "step": 27270
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.182215965627285e-05,
+      "loss": 2.0816,
+      "step": 27275
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.178336895919657e-05,
+      "loss": 2.1281,
+      "step": 27280
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.174458287466417e-05,
+      "loss": 2.0539,
+      "step": 27285
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.170580140901584e-05,
+      "loss": 2.074,
+      "step": 27290
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.16670245685912e-05,
+      "loss": 2.0691,
+      "step": 27295
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.162825235972907e-05,
+      "loss": 2.1509,
+      "step": 27300
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.15894847887675e-05,
+      "loss": 2.0912,
+      "step": 27305
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.15507218620438e-05,
+      "loss": 2.1017,
+      "step": 27310
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.151196358589448e-05,
+      "loss": 2.1122,
+      "step": 27315
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.147320996665533e-05,
+      "loss": 2.1144,
+      "step": 27320
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.143446101066138e-05,
+      "loss": 2.0522,
+      "step": 27325
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.139571672424684e-05,
+      "loss": 2.0892,
+      "step": 27330
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.135697711374522e-05,
+      "loss": 2.0883,
+      "step": 27335
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.131824218548927e-05,
+      "loss": 2.1496,
+      "step": 27340
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.127951194581093e-05,
+      "loss": 2.1136,
+      "step": 27345
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.124078640104137e-05,
+      "loss": 2.0824,
+      "step": 27350
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.120206555751104e-05,
+      "loss": 2.0773,
+      "step": 27355
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.116334942154959e-05,
+      "loss": 2.0956,
+      "step": 27360
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.112463799948592e-05,
+      "loss": 2.1988,
+      "step": 27365
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.108593129764812e-05,
+      "loss": 2.1153,
+      "step": 27370
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.104722932236356e-05,
+      "loss": 2.1287,
+      "step": 27375
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.100853207995881e-05,
+      "loss": 2.0961,
+      "step": 27380
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.096983957675964e-05,
+      "loss": 2.0853,
+      "step": 27385
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.093115181909111e-05,
+      "loss": 2.1161,
+      "step": 27390
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.089246881327745e-05,
+      "loss": 2.1229,
+      "step": 27395
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.085379056564215e-05,
+      "loss": 2.1088,
+      "step": 27400
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.081511708250785e-05,
+      "loss": 2.12,
+      "step": 27405
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.077644837019654e-05,
+      "loss": 2.1453,
+      "step": 27410
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.07377844350293e-05,
+      "loss": 2.1162,
+      "step": 27415
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.06991252833265e-05,
+      "loss": 2.0702,
+      "step": 27420
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.066047092140772e-05,
+      "loss": 2.0616,
+      "step": 27425
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.062182135559175e-05,
+      "loss": 2.0694,
+      "step": 27430
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.058317659219658e-05,
+      "loss": 2.0969,
+      "step": 27435
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.054453663753944e-05,
+      "loss": 2.0704,
+      "step": 27440
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.050590149793676e-05,
+      "loss": 2.0862,
+      "step": 27445
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.046727117970418e-05,
+      "loss": 2.0946,
+      "step": 27450
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.042864568915657e-05,
+      "loss": 2.1202,
+      "step": 27455
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.039002503260801e-05,
+      "loss": 2.1473,
+      "step": 27460
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.035140921637177e-05,
+      "loss": 2.0759,
+      "step": 27465
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.031279824676032e-05,
+      "loss": 2.1359,
+      "step": 27470
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.027419213008537e-05,
+      "loss": 2.1043,
+      "step": 27475
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.02355908726578e-05,
+      "loss": 2.051,
+      "step": 27480
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.019699448078778e-05,
+      "loss": 2.0655,
+      "step": 27485
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.015840296078455e-05,
+      "loss": 2.0784,
+      "step": 27490
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.011981631895668e-05,
+      "loss": 2.1147,
+      "step": 27495
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.008123456161186e-05,
+      "loss": 2.1257,
+      "step": 27500
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.0042657695057e-05,
+      "loss": 2.1016,
+      "step": 27505
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.000408572559827e-05,
+      "loss": 2.052,
+      "step": 27510
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.996551865954093e-05,
+      "loss": 2.0759,
+      "step": 27515
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.992695650318954e-05,
+      "loss": 2.0761,
+      "step": 27520
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.98883992628478e-05,
+      "loss": 2.0769,
+      "step": 27525
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.984984694481864e-05,
+      "loss": 2.0979,
+      "step": 27530
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.981129955540415e-05,
+      "loss": 2.1145,
+      "step": 27535
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.977275710090563e-05,
+      "loss": 2.1021,
+      "step": 27540
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.973421958762359e-05,
+      "loss": 2.0803,
+      "step": 27545
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.969568702185771e-05,
+      "loss": 2.1061,
+      "step": 27550
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.965715940990687e-05,
+      "loss": 2.164,
+      "step": 27555
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.961863675806912e-05,
+      "loss": 2.094,
+      "step": 27560
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.958011907264177e-05,
+      "loss": 2.1495,
+      "step": 27565
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.954160635992121e-05,
+      "loss": 2.0787,
+      "step": 27570
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.95030986262031e-05,
+      "loss": 2.0536,
+      "step": 27575
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.946459587778227e-05,
+      "loss": 2.103,
+      "step": 27580
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.942609812095271e-05,
+      "loss": 2.1169,
+      "step": 27585
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.93876053620076e-05,
+      "loss": 2.1555,
+      "step": 27590
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.934911760723934e-05,
+      "loss": 2.1698,
+      "step": 27595
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.931063486293945e-05,
+      "loss": 2.1029,
+      "step": 27600
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.92721571353987e-05,
+      "loss": 2.1424,
+      "step": 27605
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.923368443090696e-05,
+      "loss": 2.1379,
+      "step": 27610
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.919521675575338e-05,
+      "loss": 2.1037,
+      "step": 27615
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.915675411622618e-05,
+      "loss": 2.1064,
+      "step": 27620
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.911829651861284e-05,
+      "loss": 2.1362,
+      "step": 27625
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.907984396919998e-05,
+      "loss": 2.0905,
+      "step": 27630
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.90413964742734e-05,
+      "loss": 2.0639,
+      "step": 27635
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.900295404011805e-05,
+      "loss": 2.1081,
+      "step": 27640
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.896451667301809e-05,
+      "loss": 2.1743,
+      "step": 27645
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.892608437925687e-05,
+      "loss": 2.0649,
+      "step": 27650
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.888765716511683e-05,
+      "loss": 2.0675,
+      "step": 27655
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.884923503687964e-05,
+      "loss": 2.0692,
+      "step": 27660
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.881081800082612e-05,
+      "loss": 2.1054,
+      "step": 27665
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.877240606323627e-05,
+      "loss": 2.1176,
+      "step": 27670
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.873399923038926e-05,
+      "loss": 2.0461,
+      "step": 27675
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.86955975085634e-05,
+      "loss": 2.1417,
+      "step": 27680
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.865720090403621e-05,
+      "loss": 2.1351,
+      "step": 27685
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.861880942308432e-05,
+      "loss": 2.1053,
+      "step": 27690
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.858042307198353e-05,
+      "loss": 2.138,
+      "step": 27695
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.854204185700885e-05,
+      "loss": 2.1049,
+      "step": 27700
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.850366578443441e-05,
+      "loss": 2.1675,
+      "step": 27705
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.84652948605335e-05,
+      "loss": 2.1529,
+      "step": 27710
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.842692909157857e-05,
+      "loss": 2.1635,
+      "step": 27715
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.838856848384124e-05,
+      "loss": 2.056,
+      "step": 27720
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.835021304359228e-05,
+      "loss": 2.1881,
+      "step": 27725
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.831186277710162e-05,
+      "loss": 2.1725,
+      "step": 27730
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.827351769063834e-05,
+      "loss": 2.1178,
+      "step": 27735
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.823517779047065e-05,
+      "loss": 2.1064,
+      "step": 27740
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.819684308286595e-05,
+      "loss": 2.0814,
+      "step": 27745
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.815851357409079e-05,
+      "loss": 2.0818,
+      "step": 27750
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.812018927041083e-05,
+      "loss": 2.1132,
+      "step": 27755
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.808187017809091e-05,
+      "loss": 2.0759,
+      "step": 27760
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.804355630339503e-05,
+      "loss": 2.1336,
+      "step": 27765
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.80052476525863e-05,
+      "loss": 2.1164,
+      "step": 27770
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.796694423192701e-05,
+      "loss": 2.1218,
+      "step": 27775
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.792864604767857e-05,
+      "loss": 2.0935,
+      "step": 27780
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.789035310610156e-05,
+      "loss": 2.1104,
+      "step": 27785
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.785206541345566e-05,
+      "loss": 2.0475,
+      "step": 27790
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.781378297599975e-05,
+      "loss": 2.1541,
+      "step": 27795
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.777550579999181e-05,
+      "loss": 2.1693,
+      "step": 27800
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.773723389168897e-05,
+      "loss": 2.0665,
+      "step": 27805
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.769896725734749e-05,
+      "loss": 2.1745,
+      "step": 27810
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.76607059032228e-05,
+      "loss": 2.102,
+      "step": 27815
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.762244983556943e-05,
+      "loss": 2.125,
+      "step": 27820
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.758419906064104e-05,
+      "loss": 2.1254,
+      "step": 27825
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.754595358469053e-05,
+      "loss": 2.1192,
+      "step": 27830
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.750771341396973e-05,
+      "loss": 2.1369,
+      "step": 27835
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.74694785547298e-05,
+      "loss": 2.106,
+      "step": 27840
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.74312490132209e-05,
+      "loss": 2.1375,
+      "step": 27845
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.739302479569243e-05,
+      "loss": 2.0879,
+      "step": 27850
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.735480590839283e-05,
+      "loss": 2.07,
+      "step": 27855
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.73165923575697e-05,
+      "loss": 2.0661,
+      "step": 27860
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.727838414946979e-05,
+      "loss": 2.1524,
+      "step": 27865
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.724018129033894e-05,
+      "loss": 2.1313,
+      "step": 27870
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.720198378642215e-05,
+      "loss": 2.0849,
+      "step": 27875
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.716379164396351e-05,
+      "loss": 2.1267,
+      "step": 27880
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.712560486920625e-05,
+      "loss": 2.1335,
+      "step": 27885
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.708742346839275e-05,
+      "loss": 2.046,
+      "step": 27890
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.704924744776446e-05,
+      "loss": 2.1254,
+      "step": 27895
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.701107681356195e-05,
+      "loss": 2.11,
+      "step": 27900
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.6972911572025e-05,
+      "loss": 2.1376,
+      "step": 27905
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.693475172939239e-05,
+      "loss": 2.1285,
+      "step": 27910
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.68965972919021e-05,
+      "loss": 2.0701,
+      "step": 27915
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.685844826579117e-05,
+      "loss": 2.0888,
+      "step": 27920
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.68203046572958e-05,
+      "loss": 2.0766,
+      "step": 27925
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.678216647265127e-05,
+      "loss": 2.1102,
+      "step": 27930
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.674403371809204e-05,
+      "loss": 2.1064,
+      "step": 27935
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.670590639985157e-05,
+      "loss": 2.1281,
+      "step": 27940
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.666778452416252e-05,
+      "loss": 2.1495,
+      "step": 27945
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.662966809725664e-05,
+      "loss": 2.143,
+      "step": 27950
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.659155712536476e-05,
+      "loss": 2.0227,
+      "step": 27955
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.655345161471688e-05,
+      "loss": 2.1499,
+      "step": 27960
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.651535157154203e-05,
+      "loss": 2.1317,
+      "step": 27965
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.64772570020684e-05,
+      "loss": 2.1171,
+      "step": 27970
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.643916791252327e-05,
+      "loss": 2.0899,
+      "step": 27975
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.640108430913304e-05,
+      "loss": 2.1321,
+      "step": 27980
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.636300619812319e-05,
+      "loss": 2.0772,
+      "step": 27985
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.632493358571829e-05,
+      "loss": 2.1158,
+      "step": 27990
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.628686647814207e-05,
+      "loss": 2.1515,
+      "step": 27995
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.624880488161727e-05,
+      "loss": 2.109,
+      "step": 28000
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.621074880236582e-05,
+      "loss": 2.1294,
+      "step": 28005
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.617269824660873e-05,
+      "loss": 2.0626,
+      "step": 28010
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.613465322056602e-05,
+      "loss": 2.0979,
+      "step": 28015
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.609661373045689e-05,
+      "loss": 2.1516,
+      "step": 28020
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.605857978249964e-05,
+      "loss": 2.0322,
+      "step": 28025
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.602055138291163e-05,
+      "loss": 2.1205,
+      "step": 28030
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.59825285379093e-05,
+      "loss": 2.1055,
+      "step": 28035
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.594451125370825e-05,
+      "loss": 2.1086,
+      "step": 28040
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.59064995365231e-05,
+      "loss": 2.1417,
+      "step": 28045
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.586849339256759e-05,
+      "loss": 2.0949,
+      "step": 28050
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.583049282805454e-05,
+      "loss": 2.1094,
+      "step": 28055
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.579249784919588e-05,
+      "loss": 2.1202,
+      "step": 28060
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.575450846220261e-05,
+      "loss": 2.1206,
+      "step": 28065
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.571652467328479e-05,
+      "loss": 2.1008,
+      "step": 28070
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.567854648865162e-05,
+      "loss": 2.0594,
+      "step": 28075
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.564057391451132e-05,
+      "loss": 2.1592,
+      "step": 28080
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.560260695707127e-05,
+      "loss": 2.0793,
+      "step": 28085
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.556464562253788e-05,
+      "loss": 2.0725,
+      "step": 28090
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.552668991711662e-05,
+      "loss": 2.1154,
+      "step": 28095
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.54887398470121e-05,
+      "loss": 2.1246,
+      "step": 28100
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.545079541842797e-05,
+      "loss": 2.1239,
+      "step": 28105
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.541285663756698e-05,
+      "loss": 2.1411,
+      "step": 28110
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.53749235106309e-05,
+      "loss": 2.0989,
+      "step": 28115
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.533699604382069e-05,
+      "loss": 2.1464,
+      "step": 28120
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.529907424333625e-05,
+      "loss": 2.1792,
+      "step": 28125
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.526115811537664e-05,
+      "loss": 2.0344,
+      "step": 28130
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.522324766613999e-05,
+      "loss": 2.1686,
+      "step": 28135
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.518534290182344e-05,
+      "loss": 2.1076,
+      "step": 28140
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.514744382862329e-05,
+      "loss": 2.1185,
+      "step": 28145
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.510955045273484e-05,
+      "loss": 2.1119,
+      "step": 28150
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 6.507166278035247e-05,
+      "loss": 2.0915,
+      "step": 28155
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.503378081766966e-05,
+      "loss": 2.0616,
+      "step": 28160
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.499590457087895e-05,
+      "loss": 2.1277,
+      "step": 28165
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.495803404617187e-05,
+      "loss": 2.1022,
+      "step": 28170
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.492016924973914e-05,
+      "loss": 2.0892,
+      "step": 28175
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.488231018777044e-05,
+      "loss": 2.0671,
+      "step": 28180
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.484445686645457e-05,
+      "loss": 2.1174,
+      "step": 28185
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.480660929197937e-05,
+      "loss": 2.1051,
+      "step": 28190
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.476876747053177e-05,
+      "loss": 2.0993,
+      "step": 28195
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.47309314082977e-05,
+      "loss": 2.0803,
+      "step": 28200
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.469310111146217e-05,
+      "loss": 2.096,
+      "step": 28205
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.465527658620928e-05,
+      "loss": 2.1179,
+      "step": 28210
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.461745783872217e-05,
+      "loss": 2.0385,
+      "step": 28215
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.457964487518304e-05,
+      "loss": 2.0865,
+      "step": 28220
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.45418377017731e-05,
+      "loss": 2.0819,
+      "step": 28225
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.45040363246727e-05,
+      "loss": 2.0879,
+      "step": 28230
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.446624075006117e-05,
+      "loss": 2.1208,
+      "step": 28235
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.44284509841169e-05,
+      "loss": 2.1085,
+      "step": 28240
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.439066703301737e-05,
+      "loss": 2.1417,
+      "step": 28245
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.435288890293905e-05,
+      "loss": 2.1409,
+      "step": 28250
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.431511660005753e-05,
+      "loss": 2.0468,
+      "step": 28255
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.427735013054738e-05,
+      "loss": 2.1572,
+      "step": 28260
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.423958950058227e-05,
+      "loss": 2.1301,
+      "step": 28265
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.420183471633486e-05,
+      "loss": 2.1254,
+      "step": 28270
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.416408578397692e-05,
+      "loss": 2.0541,
+      "step": 28275
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.412634270967921e-05,
+      "loss": 2.087,
+      "step": 28280
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.408860549961155e-05,
+      "loss": 2.1359,
+      "step": 28285
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.405087415994284e-05,
+      "loss": 2.1351,
+      "step": 28290
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.401314869684092e-05,
+      "loss": 2.1795,
+      "step": 28295
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.397542911647278e-05,
+      "loss": 2.0338,
+      "step": 28300
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.393771542500439e-05,
+      "loss": 2.1523,
+      "step": 28305
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.390000762860075e-05,
+      "loss": 2.0789,
+      "step": 28310
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.386230573342595e-05,
+      "loss": 2.1495,
+      "step": 28315
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.382460974564306e-05,
+      "loss": 2.1212,
+      "step": 28320
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.37869196714142e-05,
+      "loss": 2.123,
+      "step": 28325
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.374923551690053e-05,
+      "loss": 2.1195,
+      "step": 28330
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.371155728826224e-05,
+      "loss": 2.1225,
+      "step": 28335
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.367388499165857e-05,
+      "loss": 2.0606,
+      "step": 28340
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.363621863324774e-05,
+      "loss": 2.0818,
+      "step": 28345
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.359855821918706e-05,
+      "loss": 2.0332,
+      "step": 28350
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.356090375563281e-05,
+      "loss": 2.0638,
+      "step": 28355
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.352325524874037e-05,
+      "loss": 2.1405,
+      "step": 28360
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.348561270466407e-05,
+      "loss": 2.0901,
+      "step": 28365
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.34479761295573e-05,
+      "loss": 2.1106,
+      "step": 28370
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.341034552957253e-05,
+      "loss": 2.1357,
+      "step": 28375
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.337272091086108e-05,
+      "loss": 2.1079,
+      "step": 28380
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.333510227957347e-05,
+      "loss": 2.1259,
+      "step": 28385
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.329748964185918e-05,
+      "loss": 2.163,
+      "step": 28390
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.325988300386672e-05,
+      "loss": 2.0885,
+      "step": 28395
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.322228237174361e-05,
+      "loss": 2.1141,
+      "step": 28400
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.318468775163636e-05,
+      "loss": 2.1542,
+      "step": 28405
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.314709914969054e-05,
+      "loss": 2.0962,
+      "step": 28410
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.310951657205073e-05,
+      "loss": 2.0842,
+      "step": 28415
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.30719400248605e-05,
+      "loss": 2.1234,
+      "step": 28420
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.303436951426244e-05,
+      "loss": 2.0654,
+      "step": 28425
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.299680504639822e-05,
+      "loss": 2.1703,
+      "step": 28430
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.295924662740839e-05,
+      "loss": 2.1168,
+      "step": 28435
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.292169426343262e-05,
+      "loss": 2.0711,
+      "step": 28440
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.288414796060956e-05,
+      "loss": 2.0536,
+      "step": 28445
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.284660772507686e-05,
+      "loss": 2.0615,
+      "step": 28450
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.280907356297119e-05,
+      "loss": 2.1215,
+      "step": 28455
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.277154548042821e-05,
+      "loss": 2.1281,
+      "step": 28460
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.273402348358262e-05,
+      "loss": 2.0722,
+      "step": 28465
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.269650757856807e-05,
+      "loss": 2.1442,
+      "step": 28470
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.265899777151727e-05,
+      "loss": 2.1733,
+      "step": 28475
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.262149406856191e-05,
+      "loss": 2.1346,
+      "step": 28480
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.258399647583266e-05,
+      "loss": 2.0933,
+      "step": 28485
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.254650499945922e-05,
+      "loss": 2.0819,
+      "step": 28490
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.25090196455703e-05,
+      "loss": 2.1331,
+      "step": 28495
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.247154042029359e-05,
+      "loss": 2.1455,
+      "step": 28500
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.243406732975578e-05,
+      "loss": 2.0669,
+      "step": 28505
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.239660038008253e-05,
+      "loss": 2.0897,
+      "step": 28510
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.235913957739857e-05,
+      "loss": 2.1566,
+      "step": 28515
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.232168492782752e-05,
+      "loss": 2.0822,
+      "step": 28520
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.228423643749213e-05,
+      "loss": 2.1925,
+      "step": 28525
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.2246794112514e-05,
+      "loss": 2.1351,
+      "step": 28530
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.220935795901382e-05,
+      "loss": 2.1064,
+      "step": 28535
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.217192798311122e-05,
+      "loss": 2.1065,
+      "step": 28540
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.213450419092488e-05,
+      "loss": 2.0843,
+      "step": 28545
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.209708658857238e-05,
+      "loss": 2.0543,
+      "step": 28550
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.20596751821704e-05,
+      "loss": 2.1168,
+      "step": 28555
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.202226997783446e-05,
+      "loss": 2.1463,
+      "step": 28560
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.198487098167922e-05,
+      "loss": 2.1564,
+      "step": 28565
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.194747819981821e-05,
+      "loss": 2.0539,
+      "step": 28570
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.191009163836399e-05,
+      "loss": 2.0936,
+      "step": 28575
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.187271130342815e-05,
+      "loss": 2.0843,
+      "step": 28580
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.183533720112119e-05,
+      "loss": 2.1125,
+      "step": 28585
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.17979693375526e-05,
+      "loss": 2.1063,
+      "step": 28590
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.176060771883087e-05,
+      "loss": 2.1131,
+      "step": 28595
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.172325235106349e-05,
+      "loss": 2.1168,
+      "step": 28600
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.168590324035687e-05,
+      "loss": 2.086,
+      "step": 28605
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.164856039281648e-05,
+      "loss": 2.1275,
+      "step": 28610
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.161122381454667e-05,
+      "loss": 2.146,
+      "step": 28615
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.157389351165083e-05,
+      "loss": 2.1076,
+      "step": 28620
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.153656949023133e-05,
+      "loss": 2.0679,
+      "step": 28625
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.149925175638943e-05,
+      "loss": 2.1256,
+      "step": 28630
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.146194031622548e-05,
+      "loss": 2.0955,
+      "step": 28635
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.14246351758387e-05,
+      "loss": 2.0855,
+      "step": 28640
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.138733634132736e-05,
+      "loss": 2.1136,
+      "step": 28645
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.135004381878863e-05,
+      "loss": 2.1113,
+      "step": 28650
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.131275761431872e-05,
+      "loss": 2.106,
+      "step": 28655
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.127547773401272e-05,
+      "loss": 2.1335,
+      "step": 28660
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.123820418396478e-05,
+      "loss": 2.1077,
+      "step": 28665
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.120093697026792e-05,
+      "loss": 2.1212,
+      "step": 28670
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.116367609901423e-05,
+      "loss": 2.1301,
+      "step": 28675
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.112642157629465e-05,
+      "loss": 2.1069,
+      "step": 28680
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.108917340819917e-05,
+      "loss": 2.1105,
+      "step": 28685
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.105193160081672e-05,
+      "loss": 2.1053,
+      "step": 28690
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.101469616023516e-05,
+      "loss": 2.1034,
+      "step": 28695
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.097746709254133e-05,
+      "loss": 2.1446,
+      "step": 28700
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.094024440382104e-05,
+      "loss": 2.1291,
+      "step": 28705
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.090302810015902e-05,
+      "loss": 2.118,
+      "step": 28710
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.086581818763901e-05,
+      "loss": 2.1274,
+      "step": 28715
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.0828614672343665e-05,
+      "loss": 2.1095,
+      "step": 28720
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.07914175603546e-05,
+      "loss": 2.0943,
+      "step": 28725
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.0754226857752386e-05,
+      "loss": 2.1182,
+      "step": 28730
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.071704257061657e-05,
+      "loss": 2.0704,
+      "step": 28735
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.0679864705025654e-05,
+      "loss": 2.1316,
+      "step": 28740
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.0642693267056995e-05,
+      "loss": 2.0812,
+      "step": 28745
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.0605528262787e-05,
+      "loss": 2.1326,
+      "step": 28750
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.0568369698291e-05,
+      "loss": 2.0977,
+      "step": 28755
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.053121757964326e-05,
+      "loss": 2.109,
+      "step": 28760
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.049407191291701e-05,
+      "loss": 2.1098,
+      "step": 28765
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.04569327041844e-05,
+      "loss": 2.0997,
+      "step": 28770
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.041979995951654e-05,
+      "loss": 2.0506,
+      "step": 28775
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.03826736849835e-05,
+      "loss": 2.1163,
+      "step": 28780
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.034555388665426e-05,
+      "loss": 2.1181,
+      "step": 28785
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.0308440570596755e-05,
+      "loss": 2.0858,
+      "step": 28790
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.027133374287788e-05,
+      "loss": 2.0955,
+      "step": 28795
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.023423340956341e-05,
+      "loss": 2.0945,
+      "step": 28800
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.019713957671813e-05,
+      "loss": 2.0659,
+      "step": 28805
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.016005225040573e-05,
+      "loss": 2.1442,
+      "step": 28810
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.012297143668881e-05,
+      "loss": 2.1235,
+      "step": 28815
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.0085897141628976e-05,
+      "loss": 2.0555,
+      "step": 28820
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.0048829371286694e-05,
+      "loss": 2.1458,
+      "step": 28825
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 6.00117681317214e-05,
+      "loss": 2.0791,
+      "step": 28830
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.997471342899147e-05,
+      "loss": 2.1018,
+      "step": 28835
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.9937665269154183e-05,
+      "loss": 2.1022,
+      "step": 28840
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.990062365826576e-05,
+      "loss": 2.1325,
+      "step": 28845
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.986358860238137e-05,
+      "loss": 2.1354,
+      "step": 28850
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.9826560107555096e-05,
+      "loss": 2.1042,
+      "step": 28855
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.978953817983993e-05,
+      "loss": 2.1495,
+      "step": 28860
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.9752522825287824e-05,
+      "loss": 2.0815,
+      "step": 28865
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.971551404994964e-05,
+      "loss": 2.111,
+      "step": 28870
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.9678511859875164e-05,
+      "loss": 2.1352,
+      "step": 28875
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.96415162611131e-05,
+      "loss": 2.0724,
+      "step": 28880
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.9604527259711085e-05,
+      "loss": 2.1206,
+      "step": 28885
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.956754486171566e-05,
+      "loss": 2.1356,
+      "step": 28890
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.953056907317233e-05,
+      "loss": 2.1143,
+      "step": 28895
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.949359990012546e-05,
+      "loss": 2.1567,
+      "step": 28900
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.945663734861837e-05,
+      "loss": 2.1025,
+      "step": 28905
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.941968142469332e-05,
+      "loss": 2.0749,
+      "step": 28910
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.9382732134391403e-05,
+      "loss": 2.1367,
+      "step": 28915
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.934578948375275e-05,
+      "loss": 2.0888,
+      "step": 28920
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.9308853478816275e-05,
+      "loss": 2.1133,
+      "step": 28925
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.927192412561988e-05,
+      "loss": 2.1528,
+      "step": 28930
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.9235001430200384e-05,
+      "loss": 2.1635,
+      "step": 28935
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.919808539859351e-05,
+      "loss": 2.1222,
+      "step": 28940
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.9161176036833864e-05,
+      "loss": 2.148,
+      "step": 28945
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.912427335095501e-05,
+      "loss": 2.1474,
+      "step": 28950
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.908737734698935e-05,
+      "loss": 2.089,
+      "step": 28955
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.9050488030968264e-05,
+      "loss": 2.1425,
+      "step": 28960
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.901360540892202e-05,
+      "loss": 2.1515,
+      "step": 28965
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.8976729486879756e-05,
+      "loss": 2.1492,
+      "step": 28970
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.893986027086956e-05,
+      "loss": 2.1415,
+      "step": 28975
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.89029977669184e-05,
+      "loss": 2.1799,
+      "step": 28980
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.886614198105215e-05,
+      "loss": 2.1743,
+      "step": 28985
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.882929291929558e-05,
+      "loss": 2.1148,
+      "step": 28990
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.8792450587672373e-05,
+      "loss": 2.1362,
+      "step": 28995
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.875561499220513e-05,
+      "loss": 2.1503,
+      "step": 29000
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.871878613891531e-05,
+      "loss": 2.1343,
+      "step": 29005
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.8681964033823265e-05,
+      "loss": 2.1594,
+      "step": 29010
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.86451486829483e-05,
+      "loss": 2.1007,
+      "step": 29015
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.8608340092308575e-05,
+      "loss": 2.1334,
+      "step": 29020
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.8571538267921147e-05,
+      "loss": 2.1756,
+      "step": 29025
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.853474321580198e-05,
+      "loss": 2.135,
+      "step": 29030
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.84979549419659e-05,
+      "loss": 2.0905,
+      "step": 29035
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.84611734524267e-05,
+      "loss": 2.2211,
+      "step": 29040
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.8424398753196954e-05,
+      "loss": 2.0848,
+      "step": 29045
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.838763085028822e-05,
+      "loss": 2.0949,
+      "step": 29050
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.8350869749710914e-05,
+      "loss": 2.1331,
+      "step": 29055
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.8314115457474274e-05,
+      "loss": 2.1151,
+      "step": 29060
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.8277367979586585e-05,
+      "loss": 2.1099,
+      "step": 29065
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.824062732205482e-05,
+      "loss": 2.0418,
+      "step": 29070
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.8203893490885044e-05,
+      "loss": 2.1006,
+      "step": 29075
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.816716649208198e-05,
+      "loss": 2.1197,
+      "step": 29080
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.8130446331649456e-05,
+      "loss": 2.1175,
+      "step": 29085
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.809373301559e-05,
+      "loss": 2.024,
+      "step": 29090
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.805702654990517e-05,
+      "loss": 2.1084,
+      "step": 29095
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.802032694059531e-05,
+      "loss": 2.0918,
+      "step": 29100
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.79836341936596e-05,
+      "loss": 2.1131,
+      "step": 29105
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.794694831509628e-05,
+      "loss": 2.0671,
+      "step": 29110
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.7910269310902246e-05,
+      "loss": 2.0794,
+      "step": 29115
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.787359718707348e-05,
+      "loss": 2.115,
+      "step": 29120
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.783693194960463e-05,
+      "loss": 2.0746,
+      "step": 29125
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.780027360448943e-05,
+      "loss": 2.0604,
+      "step": 29130
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.776362215772029e-05,
+      "loss": 2.0735,
+      "step": 29135
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.772697761528867e-05,
+      "loss": 2.0819,
+      "step": 29140
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.769033998318472e-05,
+      "loss": 2.1172,
+      "step": 29145
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.765370926739765e-05,
+      "loss": 2.1162,
+      "step": 29150
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.761708547391537e-05,
+      "loss": 2.1347,
+      "step": 29155
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.75804686087248e-05,
+      "loss": 2.0777,
+      "step": 29160
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.75438586778116e-05,
+      "loss": 2.1134,
+      "step": 29165
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.750725568716042e-05,
+      "loss": 2.1175,
+      "step": 29170
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.747065964275464e-05,
+      "loss": 2.1215,
+      "step": 29175
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.743407055057666e-05,
+      "loss": 2.0971,
+      "step": 29180
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.739748841660758e-05,
+      "loss": 2.1623,
+      "step": 29185
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.736091324682753e-05,
+      "loss": 2.1453,
+      "step": 29190
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.732434504721532e-05,
+      "loss": 2.1531,
+      "step": 29195
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.728778382374882e-05,
+      "loss": 2.1463,
+      "step": 29200
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.7251229582404545e-05,
+      "loss": 2.1351,
+      "step": 29205
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.721468232915808e-05,
+      "loss": 2.0657,
+      "step": 29210
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.717814206998369e-05,
+      "loss": 2.0903,
+      "step": 29215
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.714160881085464e-05,
+      "loss": 2.1568,
+      "step": 29220
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.710508255774292e-05,
+      "loss": 2.1184,
+      "step": 29225
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.7068563316619496e-05,
+      "loss": 2.1081,
+      "step": 29230
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.7032051093454085e-05,
+      "loss": 2.1677,
+      "step": 29235
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.699554589421534e-05,
+      "loss": 2.1215,
+      "step": 29240
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.6959047724870694e-05,
+      "loss": 2.0989,
+      "step": 29245
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.692255659138652e-05,
+      "loss": 2.1107,
+      "step": 29250
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.6886072499727905e-05,
+      "loss": 2.1302,
+      "step": 29255
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.684959545585895e-05,
+      "loss": 2.1114,
+      "step": 29260
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.6813125465742436e-05,
+      "loss": 2.1017,
+      "step": 29265
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.677666253534016e-05,
+      "loss": 2.1271,
+      "step": 29270
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.6740206670612606e-05,
+      "loss": 2.2029,
+      "step": 29275
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.670375787751925e-05,
+      "loss": 2.1178,
+      "step": 29280
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.66673161620183e-05,
+      "loss": 2.1584,
+      "step": 29285
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.66308815300668e-05,
+      "loss": 2.1236,
+      "step": 29290
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.659445398762078e-05,
+      "loss": 2.1128,
+      "step": 29295
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.655803354063491e-05,
+      "loss": 2.1061,
+      "step": 29300
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.6521620195062884e-05,
+      "loss": 2.0962,
+      "step": 29305
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.6485213956857106e-05,
+      "loss": 2.1071,
+      "step": 29310
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.644881483196891e-05,
+      "loss": 2.1497,
+      "step": 29315
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.6412422826348364e-05,
+      "loss": 2.0959,
+      "step": 29320
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.6376037945944506e-05,
+      "loss": 2.1511,
+      "step": 29325
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.6339660196705054e-05,
+      "loss": 2.0962,
+      "step": 29330
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.6303289584576736e-05,
+      "loss": 2.1304,
+      "step": 29335
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.62669261155049e-05,
+      "loss": 2.0743,
+      "step": 29340
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.6230569795433975e-05,
+      "loss": 2.095,
+      "step": 29345
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.619422063030697e-05,
+      "loss": 2.0404,
+      "step": 29350
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.615787862606595e-05,
+      "loss": 2.1024,
+      "step": 29355
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.612154378865162e-05,
+      "loss": 2.0944,
+      "step": 29360
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.608521612400368e-05,
+      "loss": 2.1467,
+      "step": 29365
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.6048895638060484e-05,
+      "loss": 2.1239,
+      "step": 29370
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.601258233675939e-05,
+      "loss": 2.1469,
+      "step": 29375
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.597627622603643e-05,
+      "loss": 2.0839,
+      "step": 29380
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.593997731182661e-05,
+      "loss": 2.1208,
+      "step": 29385
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.590368560006356e-05,
+      "loss": 2.11,
+      "step": 29390
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.5867401096679964e-05,
+      "loss": 2.1732,
+      "step": 29395
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.583112380760711e-05,
+      "loss": 2.0844,
+      "step": 29400
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.579485373877531e-05,
+      "loss": 2.1149,
+      "step": 29405
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.57585908961135e-05,
+      "loss": 2.0818,
+      "step": 29410
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.572233528554962e-05,
+      "loss": 2.1286,
+      "step": 29415
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.568608691301025e-05,
+      "loss": 2.0855,
+      "step": 29420
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.564984578442096e-05,
+      "loss": 2.0879,
+      "step": 29425
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.5613611905705966e-05,
+      "loss": 2.1082,
+      "step": 29430
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.557738528278846e-05,
+      "loss": 2.1295,
+      "step": 29435
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.55411659215903e-05,
+      "loss": 2.1263,
+      "step": 29440
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.55049538280323e-05,
+      "loss": 2.1451,
+      "step": 29445
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.546874900803394e-05,
+      "loss": 2.1194,
+      "step": 29450
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.543255146751366e-05,
+      "loss": 2.1076,
+      "step": 29455
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.539636121238856e-05,
+      "loss": 2.0183,
+      "step": 29460
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.536017824857473e-05,
+      "loss": 2.113,
+      "step": 29465
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.5324002581986824e-05,
+      "loss": 2.0913,
+      "step": 29470
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.528783421853856e-05,
+      "loss": 2.0664,
+      "step": 29475
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.525167316414224e-05,
+      "loss": 2.1254,
+      "step": 29480
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.521551942470916e-05,
+      "loss": 2.1386,
+      "step": 29485
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.5179373006149285e-05,
+      "loss": 2.1526,
+      "step": 29490
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.5143233914371465e-05,
+      "loss": 2.1017,
+      "step": 29495
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.510710215528328e-05,
+      "loss": 2.1809,
+      "step": 29500
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.507097773479122e-05,
+      "loss": 2.1071,
+      "step": 29505
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.503486065880042e-05,
+      "loss": 2.0653,
+      "step": 29510
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.4998750933214995e-05,
+      "loss": 2.1981,
+      "step": 29515
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.496264856393767e-05,
+      "loss": 2.1059,
+      "step": 29520
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.492655355687015e-05,
+      "loss": 2.1367,
+      "step": 29525
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.4890465917912756e-05,
+      "loss": 2.0827,
+      "step": 29530
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.485438565296477e-05,
+      "loss": 2.1598,
+      "step": 29535
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.481831276792419e-05,
+      "loss": 2.1044,
+      "step": 29540
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.478224726868779e-05,
+      "loss": 2.0479,
+      "step": 29545
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.4746189161151194e-05,
+      "loss": 2.1219,
+      "step": 29550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.471013845120871e-05,
+      "loss": 2.1292,
+      "step": 29555
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.467409514475361e-05,
+      "loss": 2.1478,
+      "step": 29560
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.463805924767776e-05,
+      "loss": 2.0584,
+      "step": 29565
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.4602030765872004e-05,
+      "loss": 2.0899,
+      "step": 29570
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.4566009705225785e-05,
+      "loss": 2.0958,
+      "step": 29575
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.452999607162753e-05,
+      "loss": 2.1128,
+      "step": 29580
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.449398987096426e-05,
+      "loss": 2.1154,
+      "step": 29585
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.445799110912193e-05,
+      "loss": 2.1,
+      "step": 29590
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.442199979198519e-05,
+      "loss": 2.1391,
+      "step": 29595
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.438601592543753e-05,
+      "loss": 2.1336,
+      "step": 29600
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.435003951536115e-05,
+      "loss": 2.1449,
+      "step": 29605
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.431407056763715e-05,
+      "loss": 2.0706,
+      "step": 29610
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.4278109088145267e-05,
+      "loss": 2.0703,
+      "step": 29615
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.424215508276414e-05,
+      "loss": 2.0601,
+      "step": 29620
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.420620855737108e-05,
+      "loss": 2.1573,
+      "step": 29625
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.417026951784229e-05,
+      "loss": 2.1172,
+      "step": 29630
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.413433797005262e-05,
+      "loss": 2.0972,
+      "step": 29635
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.4098413919875835e-05,
+      "loss": 2.1096,
+      "step": 29640
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.4062497373184364e-05,
+      "loss": 2.1044,
+      "step": 29645
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.402658833584946e-05,
+      "loss": 2.1015,
+      "step": 29650
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.399068681374109e-05,
+      "loss": 2.0752,
+      "step": 29655
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.3954792812728105e-05,
+      "loss": 2.1116,
+      "step": 29660
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.3918906338678e-05,
+      "loss": 2.103,
+      "step": 29665
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.3883027397457166e-05,
+      "loss": 2.1068,
+      "step": 29670
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.384715599493062e-05,
+      "loss": 2.0818,
+      "step": 29675
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.381129213696232e-05,
+      "loss": 2.0976,
+      "step": 29680
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.37754358294148e-05,
+      "loss": 2.1716,
+      "step": 29685
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.3739587078149524e-05,
+      "loss": 2.1538,
+      "step": 29690
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.370374588902659e-05,
+      "loss": 2.1323,
+      "step": 29695
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.3667912267904996e-05,
+      "loss": 2.1151,
+      "step": 29700
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.363208622064233e-05,
+      "loss": 2.1275,
+      "step": 29705
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.359626775309514e-05,
+      "loss": 2.1254,
+      "step": 29710
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.356045687111855e-05,
+      "loss": 2.1353,
+      "step": 29715
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.35246535805666e-05,
+      "loss": 2.1145,
+      "step": 29720
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.348885788729194e-05,
+      "loss": 2.1367,
+      "step": 29725
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.345306979714615e-05,
+      "loss": 2.1411,
+      "step": 29730
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.341728931597938e-05,
+      "loss": 2.0783,
+      "step": 29735
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.338151644964069e-05,
+      "loss": 2.1465,
+      "step": 29740
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.334575120397779e-05,
+      "loss": 2.1099,
+      "step": 29745
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.330999358483725e-05,
+      "loss": 2.1291,
+      "step": 29750
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.327424359806426e-05,
+      "loss": 2.1001,
+      "step": 29755
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.3238501249502884e-05,
+      "loss": 2.14,
+      "step": 29760
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.3202766544995844e-05,
+      "loss": 2.133,
+      "step": 29765
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.3167039490384705e-05,
+      "loss": 2.1263,
+      "step": 29770
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.31313200915097e-05,
+      "loss": 2.0797,
+      "step": 29775
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.3095608354209855e-05,
+      "loss": 2.1044,
+      "step": 29780
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.3059904284322907e-05,
+      "loss": 2.1116,
+      "step": 29785
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.302420788768543e-05,
+      "loss": 2.1306,
+      "step": 29790
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.2988519170132566e-05,
+      "loss": 2.1518,
+      "step": 29795
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.295283813749844e-05,
+      "loss": 2.1176,
+      "step": 29800
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.291716479561568e-05,
+      "loss": 2.0946,
+      "step": 29805
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.288149915031587e-05,
+      "loss": 2.1445,
+      "step": 29810
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.2845841207429147e-05,
+      "loss": 2.112,
+      "step": 29815
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.281019097278457e-05,
+      "loss": 2.0824,
+      "step": 29820
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.277454845220978e-05,
+      "loss": 2.1611,
+      "step": 29825
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.273891365153121e-05,
+      "loss": 2.1323,
+      "step": 29830
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.270328657657412e-05,
+      "loss": 2.1131,
+      "step": 29835
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.2667667233162346e-05,
+      "loss": 2.1348,
+      "step": 29840
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.2632055627118614e-05,
+      "loss": 2.1041,
+      "step": 29845
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.259645176426425e-05,
+      "loss": 2.0658,
+      "step": 29850
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.256085565041946e-05,
+      "loss": 2.1455,
+      "step": 29855
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.252526729140304e-05,
+      "loss": 2.1094,
+      "step": 29860
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.2489686693032635e-05,
+      "loss": 2.0811,
+      "step": 29865
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.245411386112451e-05,
+      "loss": 2.1395,
+      "step": 29870
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.241854880149377e-05,
+      "loss": 2.1252,
+      "step": 29875
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.238299151995416e-05,
+      "loss": 2.1359,
+      "step": 29880
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.2347442022318235e-05,
+      "loss": 2.1034,
+      "step": 29885
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.2311900314397186e-05,
+      "loss": 2.0416,
+      "step": 29890
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.2276366402001044e-05,
+      "loss": 2.1245,
+      "step": 29895
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.224084029093841e-05,
+      "loss": 2.1191,
+      "step": 29900
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.22053219870168e-05,
+      "loss": 2.0584,
+      "step": 29905
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.2169811496042274e-05,
+      "loss": 2.107,
+      "step": 29910
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.213430882381977e-05,
+      "loss": 2.1401,
+      "step": 29915
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.2098813976152785e-05,
+      "loss": 2.1201,
+      "step": 29920
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.206332695884372e-05,
+      "loss": 2.135,
+      "step": 29925
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.2027847777693514e-05,
+      "loss": 2.1075,
+      "step": 29930
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.1992376438501986e-05,
+      "loss": 2.1081,
+      "step": 29935
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.195691294706755e-05,
+      "loss": 2.1484,
+      "step": 29940
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.1921457309187426e-05,
+      "loss": 2.1398,
+      "step": 29945
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.1886009530657457e-05,
+      "loss": 2.0783,
+      "step": 29950
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.1850569617272316e-05,
+      "loss": 2.1172,
+      "step": 29955
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.181513757482527e-05,
+      "loss": 2.082,
+      "step": 29960
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.177971340910842e-05,
+      "loss": 2.0978,
+      "step": 29965
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.1744297125912454e-05,
+      "loss": 2.1184,
+      "step": 29970
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.17088887310269e-05,
+      "loss": 2.1185,
+      "step": 29975
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.167348823023987e-05,
+      "loss": 2.1164,
+      "step": 29980
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.1638095629338327e-05,
+      "loss": 2.0999,
+      "step": 29985
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.160271093410776e-05,
+      "loss": 2.1421,
+      "step": 29990
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.156733415033257e-05,
+      "loss": 2.1472,
+      "step": 29995
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.1531965283795683e-05,
+      "loss": 2.0954,
+      "step": 30000
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.149660434027887e-05,
+      "loss": 2.0967,
+      "step": 30005
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.146125132556254e-05,
+      "loss": 2.1281,
+      "step": 30010
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.142590624542576e-05,
+      "loss": 2.1025,
+      "step": 30015
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.139056910564641e-05,
+      "loss": 2.109,
+      "step": 30020
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.1355239912000977e-05,
+      "loss": 2.087,
+      "step": 30025
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.1319918670264746e-05,
+      "loss": 2.1067,
+      "step": 30030
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.128460538621156e-05,
+      "loss": 2.0938,
+      "step": 30035
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.1249300065614144e-05,
+      "loss": 2.1001,
+      "step": 30040
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.121400271424373e-05,
+      "loss": 2.0917,
+      "step": 30045
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.117871333787042e-05,
+      "loss": 2.0897,
+      "step": 30050
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.1143431942262856e-05,
+      "loss": 2.0727,
+      "step": 30055
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.1108158533188536e-05,
+      "loss": 2.0912,
+      "step": 30060
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.107289311641347e-05,
+      "loss": 2.0992,
+      "step": 30065
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.1037635697702546e-05,
+      "loss": 2.1064,
+      "step": 30070
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.100238628281919e-05,
+      "loss": 2.1264,
+      "step": 30075
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.0967144877525675e-05,
+      "loss": 2.1199,
+      "step": 30080
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.093191148758276e-05,
+      "loss": 2.1206,
+      "step": 30085
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.0896686118750135e-05,
+      "loss": 2.0537,
+      "step": 30090
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.086146877678595e-05,
+      "loss": 2.1008,
+      "step": 30095
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.0826259467447234e-05,
+      "loss": 2.128,
+      "step": 30100
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.079105819648953e-05,
+      "loss": 2.1064,
+      "step": 30105
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.0755864969667246e-05,
+      "loss": 2.1216,
+      "step": 30110
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.07206797927333e-05,
+      "loss": 2.1139,
+      "step": 30115
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.068550267143943e-05,
+      "loss": 2.1449,
+      "step": 30120
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.065033361153597e-05,
+      "loss": 2.1318,
+      "step": 30125
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.061517261877201e-05,
+      "loss": 2.1265,
+      "step": 30130
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.058001969889523e-05,
+      "loss": 2.14,
+      "step": 30135
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.054487485765211e-05,
+      "loss": 2.1252,
+      "step": 30140
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.0509738100787664e-05,
+      "loss": 2.1325,
+      "step": 30145
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.047460943404574e-05,
+      "loss": 2.1268,
+      "step": 30150
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.043948886316869e-05,
+      "loss": 2.1496,
+      "step": 30155
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.0404376393897746e-05,
+      "loss": 2.088,
+      "step": 30160
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.0369272031972616e-05,
+      "loss": 2.1634,
+      "step": 30165
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.033417578313184e-05,
+      "loss": 2.1051,
+      "step": 30170
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.029908765311251e-05,
+      "loss": 2.1541,
+      "step": 30175
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.026400764765051e-05,
+      "loss": 2.0718,
+      "step": 30180
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.022893577248027e-05,
+      "loss": 2.107,
+      "step": 30185
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.019387203333504e-05,
+      "loss": 2.1092,
+      "step": 30190
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.0158816435946535e-05,
+      "loss": 2.1014,
+      "step": 30195
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.012376898604534e-05,
+      "loss": 2.1044,
+      "step": 30200
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.008872968936058e-05,
+      "loss": 2.1174,
+      "step": 30205
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.0053698551620144e-05,
+      "loss": 2.0927,
+      "step": 30210
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.001867557855047e-05,
+      "loss": 2.103,
+      "step": 30215
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.99836607758768e-05,
+      "loss": 2.0788,
+      "step": 30220
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.99486541493229e-05,
+      "loss": 2.1049,
+      "step": 30225
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.9913655704611336e-05,
+      "loss": 2.0852,
+      "step": 30230
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.9878665447463205e-05,
+      "loss": 2.0835,
+      "step": 30235
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.9843683383598384e-05,
+      "loss": 2.113,
+      "step": 30240
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.98087095187353e-05,
+      "loss": 2.1423,
+      "step": 30245
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.977374385859115e-05,
+      "loss": 2.0648,
+      "step": 30250
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.9738786408881676e-05,
+      "loss": 2.1642,
+      "step": 30255
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.970383717532141e-05,
+      "loss": 2.0943,
+      "step": 30260
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.966889616362338e-05,
+      "loss": 2.0605,
+      "step": 30265
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.9633963379499436e-05,
+      "loss": 2.139,
+      "step": 30270
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.959903882865995e-05,
+      "loss": 2.1471,
+      "step": 30275
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.956412251681405e-05,
+      "loss": 2.0771,
+      "step": 30280
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.95292144496694e-05,
+      "loss": 2.0655,
+      "step": 30285
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.9494314632932485e-05,
+      "loss": 2.1736,
+      "step": 30290
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.945942307230824e-05,
+      "loss": 2.1422,
+      "step": 30295
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.94245397735004e-05,
+      "loss": 2.132,
+      "step": 30300
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.938966474221134e-05,
+      "loss": 2.0634,
+      "step": 30305
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.935479798414196e-05,
+      "loss": 2.0718,
+      "step": 30310
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.9319939504992e-05,
+      "loss": 2.0817,
+      "step": 30315
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.928508931045963e-05,
+      "loss": 2.1009,
+      "step": 30320
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.925024740624187e-05,
+      "loss": 2.0941,
+      "step": 30325
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.9215413798034204e-05,
+      "loss": 2.0738,
+      "step": 30330
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.9180588491530924e-05,
+      "loss": 2.1123,
+      "step": 30335
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.914577149242482e-05,
+      "loss": 2.1185,
+      "step": 30340
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.911096280640746e-05,
+      "loss": 2.1277,
+      "step": 30345
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.9076162439168916e-05,
+      "loss": 2.1188,
+      "step": 30350
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.9041370396398035e-05,
+      "loss": 2.0845,
+      "step": 30355
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.9006586683782154e-05,
+      "loss": 2.1294,
+      "step": 30360
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.897181130700743e-05,
+      "loss": 2.0808,
+      "step": 30365
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.89370442717585e-05,
+      "loss": 2.0931,
+      "step": 30370
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.890228558371871e-05,
+      "loss": 2.1285,
+      "step": 30375
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.886753524856997e-05,
+      "loss": 2.122,
+      "step": 30380
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.8832793271992984e-05,
+      "loss": 2.0914,
+      "step": 30385
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.8798059659666885e-05,
+      "loss": 2.0675,
+      "step": 30390
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.8763334417269634e-05,
+      "loss": 2.1356,
+      "step": 30395
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.872861755047765e-05,
+      "loss": 2.1023,
+      "step": 30400
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.869390906496615e-05,
+      "loss": 2.1298,
+      "step": 30405
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.8659208966408796e-05,
+      "loss": 2.0938,
+      "step": 30410
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.862451726047806e-05,
+      "loss": 2.1122,
+      "step": 30415
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.858983395284491e-05,
+      "loss": 2.1236,
+      "step": 30420
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.855515904917903e-05,
+      "loss": 2.0845,
+      "step": 30425
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.852049255514864e-05,
+      "loss": 2.0608,
+      "step": 30430
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.84858344764207e-05,
+      "loss": 2.0884,
+      "step": 30435
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.845118481866067e-05,
+      "loss": 2.0848,
+      "step": 30440
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.8416543587532746e-05,
+      "loss": 2.1304,
+      "step": 30445
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.838191078869964e-05,
+      "loss": 2.1271,
+      "step": 30450
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.834728642782279e-05,
+      "loss": 2.0832,
+      "step": 30455
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.831267051056216e-05,
+      "loss": 2.1331,
+      "step": 30460
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.8278063042576436e-05,
+      "loss": 2.0862,
+      "step": 30465
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.82434640295228e-05,
+      "loss": 2.1017,
+      "step": 30470
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.820887347705718e-05,
+      "loss": 2.1543,
+      "step": 30475
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.8174291390834005e-05,
+      "loss": 2.1192,
+      "step": 30480
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.813971777650642e-05,
+      "loss": 2.0825,
+      "step": 30485
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.810515263972608e-05,
+      "loss": 2.0923,
+      "step": 30490
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.8070595986143384e-05,
+      "loss": 2.085,
+      "step": 30495
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.803604782140719e-05,
+      "loss": 2.1141,
+      "step": 30500
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.800150815116515e-05,
+      "loss": 2.1487,
+      "step": 30505
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.796697698106334e-05,
+      "loss": 2.0806,
+      "step": 30510
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.793245431674659e-05,
+      "loss": 2.1041,
+      "step": 30515
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.7897940163858236e-05,
+      "loss": 2.1348,
+      "step": 30520
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.786343452804034e-05,
+      "loss": 2.1137,
+      "step": 30525
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.782893741493343e-05,
+      "loss": 2.1429,
+      "step": 30530
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.779444883017678e-05,
+      "loss": 2.1285,
+      "step": 30535
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.775996877940812e-05,
+      "loss": 2.0685,
+      "step": 30540
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.772549726826399e-05,
+      "loss": 2.141,
+      "step": 30545
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.769103430237928e-05,
+      "loss": 2.0774,
+      "step": 30550
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.765657988738772e-05,
+      "loss": 2.1187,
+      "step": 30555
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.762213402892149e-05,
+      "loss": 2.1412,
+      "step": 30560
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.75876967326114e-05,
+      "loss": 2.1209,
+      "step": 30565
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.7553268004086924e-05,
+      "loss": 2.1627,
+      "step": 30570
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.751884784897605e-05,
+      "loss": 2.0907,
+      "step": 30575
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.7484436272905465e-05,
+      "loss": 2.1525,
+      "step": 30580
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.745003328150032e-05,
+      "loss": 2.1201,
+      "step": 30585
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.7415638880384495e-05,
+      "loss": 2.1158,
+      "step": 30590
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.7381253075180365e-05,
+      "loss": 2.1548,
+      "step": 30595
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.734687587150899e-05,
+      "loss": 2.1209,
+      "step": 30600
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.7312507274989914e-05,
+      "loss": 2.1145,
+      "step": 30605
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.7278147291241416e-05,
+      "loss": 2.144,
+      "step": 30610
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.7243795925880205e-05,
+      "loss": 2.0959,
+      "step": 30615
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.720945318452174e-05,
+      "loss": 2.1448,
+      "step": 30620
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.7175119072779904e-05,
+      "loss": 2.1022,
+      "step": 30625
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.714079359626736e-05,
+      "loss": 2.1847,
+      "step": 30630
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.7106476760595174e-05,
+      "loss": 2.0982,
+      "step": 30635
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.7072168571373155e-05,
+      "loss": 2.1131,
+      "step": 30640
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.703786903420955e-05,
+      "loss": 2.0721,
+      "step": 30645
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.7003578154711345e-05,
+      "loss": 2.1156,
+      "step": 30650
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.696929593848397e-05,
+      "loss": 2.0633,
+      "step": 30655
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.6935022391131576e-05,
+      "loss": 2.1487,
+      "step": 30660
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.690075751825674e-05,
+      "loss": 2.137,
+      "step": 30665
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.686650132546079e-05,
+      "loss": 2.1254,
+      "step": 30670
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.6832253818343474e-05,
+      "loss": 2.132,
+      "step": 30675
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.6798015002503256e-05,
+      "loss": 2.1083,
+      "step": 30680
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.6763784883537076e-05,
+      "loss": 2.1808,
+      "step": 30685
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.672956346704055e-05,
+      "loss": 2.1021,
+      "step": 30690
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.6695350758607746e-05,
+      "loss": 2.1486,
+      "step": 30695
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.666114676383147e-05,
+      "loss": 2.1323,
+      "step": 30700
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.662695148830292e-05,
+      "loss": 2.0857,
+      "step": 30705
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.6592764937612035e-05,
+      "loss": 2.1163,
+      "step": 30710
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.6558587117347195e-05,
+      "loss": 2.0881,
+      "step": 30715
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.652441803309548e-05,
+      "loss": 2.1344,
+      "step": 30720
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.649025769044242e-05,
+      "loss": 2.1289,
+      "step": 30725
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.645610609497222e-05,
+      "loss": 2.0824,
+      "step": 30730
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.6421963252267577e-05,
+      "loss": 2.1261,
+      "step": 30735
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.638782916790975e-05,
+      "loss": 2.0683,
+      "step": 30740
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.635370384747867e-05,
+      "loss": 2.1163,
+      "step": 30745
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.6319587296552714e-05,
+      "loss": 2.1169,
+      "step": 30750
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.628547952070894e-05,
+      "loss": 2.1067,
+      "step": 30755
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.6251380525522836e-05,
+      "loss": 2.1138,
+      "step": 30760
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.6217290316568604e-05,
+      "loss": 2.1323,
+      "step": 30765
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.6183208899418876e-05,
+      "loss": 2.0636,
+      "step": 30770
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.614913627964495e-05,
+      "loss": 2.1306,
+      "step": 30775
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.611507246281661e-05,
+      "loss": 2.111,
+      "step": 30780
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.6081017454502275e-05,
+      "loss": 2.1348,
+      "step": 30785
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.604697126026881e-05,
+      "loss": 2.0379,
+      "step": 30790
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.6012933885681805e-05,
+      "loss": 2.1062,
+      "step": 30795
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.5978905336305224e-05,
+      "loss": 2.1172,
+      "step": 30800
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.594488561770176e-05,
+      "loss": 2.1638,
+      "step": 30805
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.5910874735432504e-05,
+      "loss": 2.1306,
+      "step": 30810
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.587687269505726e-05,
+      "loss": 2.1299,
+      "step": 30815
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.584287950213423e-05,
+      "loss": 2.0675,
+      "step": 30820
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.580889516222031e-05,
+      "loss": 2.0629,
+      "step": 30825
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.577491968087083e-05,
+      "loss": 2.0989,
+      "step": 30830
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.5740953063639794e-05,
+      "loss": 2.0905,
+      "step": 30835
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.57069953160796e-05,
+      "loss": 2.0927,
+      "step": 30840
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.5673046443741385e-05,
+      "loss": 2.1248,
+      "step": 30845
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.563910645217465e-05,
+      "loss": 2.1345,
+      "step": 30850
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.56051753469276e-05,
+      "loss": 2.0968,
+      "step": 30855
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.557125313354685e-05,
+      "loss": 2.12,
+      "step": 30860
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.553733981757772e-05,
+      "loss": 2.0829,
+      "step": 30865
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.550343540456388e-05,
+      "loss": 2.1138,
+      "step": 30870
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.546953990004774e-05,
+      "loss": 2.1086,
+      "step": 30875
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.543565330957009e-05,
+      "loss": 2.1547,
+      "step": 30880
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.540177563867042e-05,
+      "loss": 2.1491,
+      "step": 30885
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.536790689288658e-05,
+      "loss": 2.1464,
+      "step": 30890
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.533404707775515e-05,
+      "loss": 2.1941,
+      "step": 30895
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.530019619881107e-05,
+      "loss": 2.1516,
+      "step": 30900
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.5266354261588004e-05,
+      "loss": 2.1891,
+      "step": 30905
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.5232521271617976e-05,
+      "loss": 2.1877,
+      "step": 30910
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.519869723443174e-05,
+      "loss": 2.104,
+      "step": 30915
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.516488215555833e-05,
+      "loss": 2.1324,
+      "step": 30920
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.5131076040525566e-05,
+      "loss": 2.1365,
+      "step": 30925
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.509727889485964e-05,
+      "loss": 2.1175,
+      "step": 30930
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.506349072408539e-05,
+      "loss": 2.1441,
+      "step": 30935
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.502971153372608e-05,
+      "loss": 2.1265,
+      "step": 30940
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.499594132930363e-05,
+      "loss": 2.0903,
+      "step": 30945
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.496218011633833e-05,
+      "loss": 2.1669,
+      "step": 30950
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.492842790034918e-05,
+      "loss": 2.0986,
+      "step": 30955
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.489468468685354e-05,
+      "loss": 2.0986,
+      "step": 30960
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.486095048136746e-05,
+      "loss": 2.071,
+      "step": 30965
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.482722528940536e-05,
+      "loss": 2.1114,
+      "step": 30970
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.4793509116480334e-05,
+      "loss": 2.0977,
+      "step": 30975
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.475980196810385e-05,
+      "loss": 2.1509,
+      "step": 30980
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.472610384978607e-05,
+      "loss": 2.1099,
+      "step": 30985
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.469241476703551e-05,
+      "loss": 2.1117,
+      "step": 30990
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.4658734725359354e-05,
+      "loss": 2.1057,
+      "step": 30995
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.462506373026319e-05,
+      "loss": 2.1081,
+      "step": 31000
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.459140178725123e-05,
+      "loss": 2.1169,
+      "step": 31005
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.45577489018261e-05,
+      "loss": 2.1422,
+      "step": 31010
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.452410507948908e-05,
+      "loss": 2.0939,
+      "step": 31015
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.449047032573981e-05,
+      "loss": 2.1481,
+      "step": 31020
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.44568446460766e-05,
+      "loss": 2.1494,
+      "step": 31025
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.4423228045996134e-05,
+      "loss": 2.102,
+      "step": 31030
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.438962053099377e-05,
+      "loss": 2.1131,
+      "step": 31035
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.43560221065632e-05,
+      "loss": 2.1245,
+      "step": 31040
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.432243277819681e-05,
+      "loss": 2.1112,
+      "step": 31045
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.428885255138533e-05,
+      "loss": 2.1102,
+      "step": 31050
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.425528143161813e-05,
+      "loss": 2.0972,
+      "step": 31055
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.422171942438309e-05,
+      "loss": 2.1128,
+      "step": 31060
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.418816653516647e-05,
+      "loss": 2.1309,
+      "step": 31065
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.415462276945319e-05,
+      "loss": 2.1211,
+      "step": 31070
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.4121088132726574e-05,
+      "loss": 2.0948,
+      "step": 31075
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.408756263046855e-05,
+      "loss": 2.1625,
+      "step": 31080
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.4054046268159424e-05,
+      "loss": 2.0923,
+      "step": 31085
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.402053905127815e-05,
+      "loss": 2.149,
+      "step": 31090
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.398704098530205e-05,
+      "loss": 2.0448,
+      "step": 31095
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.395355207570714e-05,
+      "loss": 2.1127,
+      "step": 31100
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.392007232796766e-05,
+      "loss": 2.114,
+      "step": 31105
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.388660174755661e-05,
+      "loss": 2.1033,
+      "step": 31110
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.385314033994533e-05,
+      "loss": 2.1035,
+      "step": 31115
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.38196881106038e-05,
+      "loss": 2.0729,
+      "step": 31120
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.378624506500033e-05,
+      "loss": 2.1336,
+      "step": 31125
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.375281120860192e-05,
+      "loss": 2.1132,
+      "step": 31130
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.371938654687389e-05,
+      "loss": 2.1154,
+      "step": 31135
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.3685971085280174e-05,
+      "loss": 2.1542,
+      "step": 31140
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.365256482928314e-05,
+      "loss": 2.0785,
+      "step": 31145
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.361916778434371e-05,
+      "loss": 2.0998,
+      "step": 31150
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.3585779955921216e-05,
+      "loss": 2.0539,
+      "step": 31155
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.355240134947358e-05,
+      "loss": 2.1359,
+      "step": 31160
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.351903197045711e-05,
+      "loss": 2.1304,
+      "step": 31165
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.3485671824326755e-05,
+      "loss": 2.0388,
+      "step": 31170
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.345232091653576e-05,
+      "loss": 2.0736,
+      "step": 31175
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.341897925253605e-05,
+      "loss": 2.0896,
+      "step": 31180
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.338564683777789e-05,
+      "loss": 2.1471,
+      "step": 31185
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.335232367771015e-05,
+      "loss": 2.1241,
+      "step": 31190
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.331900977778007e-05,
+      "loss": 2.0958,
+      "step": 31195
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.328570514343352e-05,
+      "loss": 2.1103,
+      "step": 31200
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.325240978011468e-05,
+      "loss": 2.1441,
+      "step": 31205
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.321912369326639e-05,
+      "loss": 2.1366,
+      "step": 31210
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.318584688832984e-05,
+      "loss": 2.1377,
+      "step": 31215
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.31525793707448e-05,
+      "loss": 2.1053,
+      "step": 31220
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.311932114594942e-05,
+      "loss": 2.0972,
+      "step": 31225
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.308607221938046e-05,
+      "loss": 2.1168,
+      "step": 31230
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.3052832596473015e-05,
+      "loss": 2.1234,
+      "step": 31235
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.3019602282660786e-05,
+      "loss": 2.0699,
+      "step": 31240
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.2986381283375855e-05,
+      "loss": 2.1216,
+      "step": 31245
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.2953169604048884e-05,
+      "loss": 2.101,
+      "step": 31250
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.291996725010887e-05,
+      "loss": 2.1174,
+      "step": 31255
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.2886774226983464e-05,
+      "loss": 2.153,
+      "step": 31260
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.28535905400986e-05,
+      "loss": 2.0935,
+      "step": 31265
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.282041619487887e-05,
+      "loss": 2.1462,
+      "step": 31270
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.2787251196747155e-05,
+      "loss": 2.1265,
+      "step": 31275
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.275409555112501e-05,
+      "loss": 2.1106,
+      "step": 31280
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.272094926343229e-05,
+      "loss": 2.1392,
+      "step": 31285
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.268781233908736e-05,
+      "loss": 2.1207,
+      "step": 31290
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.265468478350715e-05,
+      "loss": 2.1357,
+      "step": 31295
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.262156660210691e-05,
+      "loss": 2.1401,
+      "step": 31300
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.2588457800300516e-05,
+      "loss": 2.0892,
+      "step": 31305
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.2555358383500165e-05,
+      "loss": 2.1173,
+      "step": 31310
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.2522268357116636e-05,
+      "loss": 2.1161,
+      "step": 31315
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.248918772655906e-05,
+      "loss": 2.1224,
+      "step": 31320
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.245611649723518e-05,
+      "loss": 2.0791,
+      "step": 31325
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.242305467455101e-05,
+      "loss": 2.1378,
+      "step": 31330
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.239000226391123e-05,
+      "loss": 2.1203,
+      "step": 31335
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.235695927071881e-05,
+      "loss": 2.0862,
+      "step": 31340
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.232392570037532e-05,
+      "loss": 2.078,
+      "step": 31345
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.229090155828066e-05,
+      "loss": 2.1916,
+      "step": 31350
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.225788684983331e-05,
+      "loss": 2.1292,
+      "step": 31355
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.22248815804301e-05,
+      "loss": 2.1245,
+      "step": 31360
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.2191885755466406e-05,
+      "loss": 2.1187,
+      "step": 31365
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.215889938033599e-05,
+      "loss": 2.1473,
+      "step": 31370
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.212592246043114e-05,
+      "loss": 2.1307,
+      "step": 31375
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.209295500114252e-05,
+      "loss": 2.1133,
+      "step": 31380
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.205999700785933e-05,
+      "loss": 2.0812,
+      "step": 31385
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.2027048485969115e-05,
+      "loss": 2.1703,
+      "step": 31390
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.199410944085801e-05,
+      "loss": 2.1226,
+      "step": 31395
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.196117987791046e-05,
+      "loss": 2.1416,
+      "step": 31400
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.19282598025095e-05,
+      "loss": 2.1498,
+      "step": 31405
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.189534922003645e-05,
+      "loss": 2.1123,
+      "step": 31410
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.186244813587126e-05,
+      "loss": 2.0919,
+      "step": 31415
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.1829556555392154e-05,
+      "loss": 2.0671,
+      "step": 31420
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.1796674483975986e-05,
+      "loss": 2.1054,
+      "step": 31425
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.1763801926997836e-05,
+      "loss": 2.1256,
+      "step": 31430
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.173093888983145e-05,
+      "loss": 2.1384,
+      "step": 31435
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.169808537784884e-05,
+      "loss": 2.1116,
+      "step": 31440
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.1665241396420595e-05,
+      "loss": 2.1468,
+      "step": 31445
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.1632406950915624e-05,
+      "loss": 2.1032,
+      "step": 31450
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.1599582046701424e-05,
+      "loss": 2.1021,
+      "step": 31455
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.156676668914379e-05,
+      "loss": 2.1232,
+      "step": 31460
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.1533960883607027e-05,
+      "loss": 2.0959,
+      "step": 31465
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.1501164635453824e-05,
+      "loss": 2.0865,
+      "step": 31470
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.14683779500454e-05,
+      "loss": 2.1047,
+      "step": 31475
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.143560083274138e-05,
+      "loss": 2.159,
+      "step": 31480
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.140283328889974e-05,
+      "loss": 2.1038,
+      "step": 31485
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.1370075323877044e-05,
+      "loss": 2.126,
+      "step": 31490
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.133732694302811e-05,
+      "loss": 2.1264,
+      "step": 31495
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.130458815170636e-05,
+      "loss": 2.1534,
+      "step": 31500
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.1271858955263496e-05,
+      "loss": 2.1433,
+      "step": 31505
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.123913935904981e-05,
+      "loss": 2.1137,
+      "step": 31510
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.1206429368413855e-05,
+      "loss": 2.1159,
+      "step": 31515
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.1173728988702784e-05,
+      "loss": 2.0984,
+      "step": 31520
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.114103822526202e-05,
+      "loss": 2.0245,
+      "step": 31525
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.1108357083435554e-05,
+      "loss": 2.1068,
+      "step": 31530
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.107568556856567e-05,
+      "loss": 2.1033,
+      "step": 31535
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.104302368599322e-05,
+      "loss": 2.1303,
+      "step": 31540
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.101037144105734e-05,
+      "loss": 2.1619,
+      "step": 31545
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.097772883909572e-05,
+      "loss": 2.125,
+      "step": 31550
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.094509588544435e-05,
+      "loss": 2.1073,
+      "step": 31555
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.091247258543778e-05,
+      "loss": 2.1028,
+      "step": 31560
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.0879858944408824e-05,
+      "loss": 2.1362,
+      "step": 31565
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.0847254967688886e-05,
+      "loss": 2.1219,
+      "step": 31570
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.081466066060762e-05,
+      "loss": 2.1247,
+      "step": 31575
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.0782076028493265e-05,
+      "loss": 2.1239,
+      "step": 31580
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.074950107667233e-05,
+      "loss": 2.0836,
+      "step": 31585
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.071693581046986e-05,
+      "loss": 2.1371,
+      "step": 31590
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.0684380235209216e-05,
+      "loss": 2.1105,
+      "step": 31595
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.065183435621228e-05,
+      "loss": 2.0552,
+      "step": 31600
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.061929817879925e-05,
+      "loss": 2.1057,
+      "step": 31605
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.058677170828882e-05,
+      "loss": 2.1142,
+      "step": 31610
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.0554254949997994e-05,
+      "loss": 2.1042,
+      "step": 31615
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.052174790924234e-05,
+      "loss": 2.1047,
+      "step": 31620
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.048925059133566e-05,
+      "loss": 2.1267,
+      "step": 31625
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.045676300159035e-05,
+      "loss": 2.1334,
+      "step": 31630
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.042428514531703e-05,
+      "loss": 2.0678,
+      "step": 31635
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.039181702782495e-05,
+      "loss": 2.1721,
+      "step": 31640
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.035935865442149e-05,
+      "loss": 2.1623,
+      "step": 31645
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.032691003041268e-05,
+      "loss": 2.0972,
+      "step": 31650
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.0294471161102806e-05,
+      "loss": 2.0981,
+      "step": 31655
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.026204205179469e-05,
+      "loss": 2.1282,
+      "step": 31660
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.02296227077894e-05,
+      "loss": 2.1048,
+      "step": 31665
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.019721313438659e-05,
+      "loss": 2.1363,
+      "step": 31670
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.016481333688412e-05,
+      "loss": 2.1094,
+      "step": 31675
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.0132423320578435e-05,
+      "loss": 2.0676,
+      "step": 31680
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.0100043090764225e-05,
+      "loss": 2.1198,
+      "step": 31685
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.0067672652734735e-05,
+      "loss": 2.126,
+      "step": 31690
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.003531201178145e-05,
+      "loss": 2.1048,
+      "step": 31695
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.000296117319439e-05,
+      "loss": 2.0923,
+      "step": 31700
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.997062014226185e-05,
+      "loss": 2.0832,
+      "step": 31705
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.993828892427066e-05,
+      "loss": 2.0617,
+      "step": 31710
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.990596752450589e-05,
+      "loss": 2.1589,
+      "step": 31715
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.987365594825119e-05,
+      "loss": 2.135,
+      "step": 31720
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.984135420078838e-05,
+      "loss": 2.1242,
+      "step": 31725
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.9809062287397894e-05,
+      "loss": 2.0535,
+      "step": 31730
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.977678021335839e-05,
+      "loss": 2.0474,
+      "step": 31735
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.9744507983947046e-05,
+      "loss": 2.1476,
+      "step": 31740
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.971224560443929e-05,
+      "loss": 2.1102,
+      "step": 31745
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.9679993080109124e-05,
+      "loss": 2.0448,
+      "step": 31750
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.9647750416228724e-05,
+      "loss": 2.1697,
+      "step": 31755
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.961551761806886e-05,
+      "loss": 2.1095,
+      "step": 31760
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.958329469089853e-05,
+      "loss": 2.1072,
+      "step": 31765
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.955108163998523e-05,
+      "loss": 2.0779,
+      "step": 31770
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.9518878470594746e-05,
+      "loss": 2.107,
+      "step": 31775
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.9486685187991354e-05,
+      "loss": 2.1245,
+      "step": 31780
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.945450179743759e-05,
+      "loss": 2.1031,
+      "step": 31785
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.942232830419452e-05,
+      "loss": 2.0719,
+      "step": 31790
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.939016471352143e-05,
+      "loss": 2.1252,
+      "step": 31795
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.935801103067613e-05,
+      "loss": 2.1105,
+      "step": 31800
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.932586726091471e-05,
+      "loss": 2.1019,
+      "step": 31805
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.929373340949172e-05,
+      "loss": 2.0714,
+      "step": 31810
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.926160948165999e-05,
+      "loss": 2.093,
+      "step": 31815
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.922949548267082e-05,
+      "loss": 2.1427,
+      "step": 31820
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.919739141777392e-05,
+      "loss": 2.1125,
+      "step": 31825
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.916529729221717e-05,
+      "loss": 2.1451,
+      "step": 31830
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.913321311124708e-05,
+      "loss": 2.1344,
+      "step": 31835
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.910113888010833e-05,
+      "loss": 2.1117,
+      "step": 31840
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.906907460404414e-05,
+      "loss": 2.027,
+      "step": 31845
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.903702028829596e-05,
+      "loss": 2.1003,
+      "step": 31850
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.9004975938103736e-05,
+      "loss": 2.0735,
+      "step": 31855
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.8972941558705656e-05,
+      "loss": 2.1654,
+      "step": 31860
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.894091715533843e-05,
+      "loss": 2.0646,
+      "step": 31865
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.8908902733236966e-05,
+      "loss": 2.061,
+      "step": 31870
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.8876898297634714e-05,
+      "loss": 2.1218,
+      "step": 31875
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.884490385376333e-05,
+      "loss": 2.1329,
+      "step": 31880
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.881291940685298e-05,
+      "loss": 2.0606,
+      "step": 31885
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.878094496213206e-05,
+      "loss": 2.1036,
+      "step": 31890
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.874898052482747e-05,
+      "loss": 2.0849,
+      "step": 31895
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.871702610016434e-05,
+      "loss": 2.0915,
+      "step": 31900
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.86850816933663e-05,
+      "loss": 2.102,
+      "step": 31905
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.865314730965518e-05,
+      "loss": 2.0932,
+      "step": 31910
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.862122295425135e-05,
+      "loss": 2.1166,
+      "step": 31915
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.858930863237338e-05,
+      "loss": 2.1376,
+      "step": 31920
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.855740434923833e-05,
+      "loss": 2.0749,
+      "step": 31925
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.852551011006149e-05,
+      "loss": 2.181,
+      "step": 31930
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.849362592005666e-05,
+      "loss": 2.0925,
+      "step": 31935
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.846175178443585e-05,
+      "loss": 2.0952,
+      "step": 31940
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.842988770840955e-05,
+      "loss": 2.1251,
+      "step": 31945
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.839803369718649e-05,
+      "loss": 2.0939,
+      "step": 31950
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.836618975597386e-05,
+      "loss": 2.0804,
+      "step": 31955
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.833435588997711e-05,
+      "loss": 2.1408,
+      "step": 31960
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.830253210440015e-05,
+      "loss": 2.0776,
+      "step": 31965
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.82707184044451e-05,
+      "loss": 2.116,
+      "step": 31970
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.823891479531261e-05,
+      "loss": 2.0796,
+      "step": 31975
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.8207121282201494e-05,
+      "loss": 2.1734,
+      "step": 31980
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.817533787030908e-05,
+      "loss": 2.1469,
+      "step": 31985
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.8143564564830904e-05,
+      "loss": 2.1033,
+      "step": 31990
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.811180137096098e-05,
+      "loss": 2.1537,
+      "step": 31995
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.808004829389154e-05,
+      "loss": 2.1038,
+      "step": 32000
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.804830533881328e-05,
+      "loss": 2.1261,
+      "step": 32005
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.801657251091517e-05,
+      "loss": 2.1202,
+      "step": 32010
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.798484981538451e-05,
+      "loss": 2.0841,
+      "step": 32015
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.795313725740703e-05,
+      "loss": 2.1029,
+      "step": 32020
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.7921434842166715e-05,
+      "loss": 2.1145,
+      "step": 32025
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.7889742574845956e-05,
+      "loss": 2.098,
+      "step": 32030
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.78580604606254e-05,
+      "loss": 2.0999,
+      "step": 32035
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.782638850468416e-05,
+      "loss": 2.1633,
+      "step": 32040
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.7794726712199556e-05,
+      "loss": 2.1547,
+      "step": 32045
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.776307508834737e-05,
+      "loss": 2.0791,
+      "step": 32050
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.7731433638301586e-05,
+      "loss": 2.0836,
+      "step": 32055
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.7699802367234685e-05,
+      "loss": 2.1108,
+      "step": 32060
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.7668181280317315e-05,
+      "loss": 2.12,
+      "step": 32065
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.763657038271862e-05,
+      "loss": 2.1241,
+      "step": 32070
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.7604969679605926e-05,
+      "loss": 2.1117,
+      "step": 32075
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.757337917614504e-05,
+      "loss": 2.1136,
+      "step": 32080
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.7541798877499965e-05,
+      "loss": 2.1444,
+      "step": 32085
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.751022878883316e-05,
+      "loss": 2.1604,
+      "step": 32090
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.747866891530529e-05,
+      "loss": 2.1287,
+      "step": 32095
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.744711926207549e-05,
+      "loss": 2.1048,
+      "step": 32100
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.741557983430107e-05,
+      "loss": 2.1066,
+      "step": 32105
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.738405063713783e-05,
+      "loss": 2.0798,
+      "step": 32110
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.735253167573972e-05,
+      "loss": 2.1181,
+      "step": 32115
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.732102295525922e-05,
+      "loss": 2.115,
+      "step": 32120
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.7289524480846925e-05,
+      "loss": 2.157,
+      "step": 32125
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.725803625765195e-05,
+      "loss": 2.141,
+      "step": 32130
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.7226558290821565e-05,
+      "loss": 2.0719,
+      "step": 32135
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.71950905855015e-05,
+      "loss": 2.1223,
+      "step": 32140
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.716363314683571e-05,
+      "loss": 2.1117,
+      "step": 32145
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.713218597996654e-05,
+      "loss": 2.1012,
+      "step": 32150
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.7100749090034584e-05,
+      "loss": 2.1418,
+      "step": 32155
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.706932248217886e-05,
+      "loss": 2.1078,
+      "step": 32160
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.703790616153657e-05,
+      "loss": 2.1286,
+      "step": 32165
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.700650013324339e-05,
+      "loss": 2.0754,
+      "step": 32170
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.697510440243315e-05,
+      "loss": 2.1211,
+      "step": 32175
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.694371897423815e-05,
+      "loss": 2.1197,
+      "step": 32180
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.691234385378889e-05,
+      "loss": 2.1493,
+      "step": 32185
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.688097904621425e-05,
+      "loss": 2.1566,
+      "step": 32190
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.684962455664135e-05,
+      "loss": 2.0903,
+      "step": 32195
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.681828039019576e-05,
+      "loss": 2.0335,
+      "step": 32200
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.67869465520012e-05,
+      "loss": 2.0946,
+      "step": 32205
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.6755623047179844e-05,
+      "loss": 2.1491,
+      "step": 32210
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.672430988085206e-05,
+      "loss": 2.1131,
+      "step": 32215
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.669300705813663e-05,
+      "loss": 2.1162,
+      "step": 32220
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.666171458415054e-05,
+      "loss": 2.1972,
+      "step": 32225
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.663043246400921e-05,
+      "loss": 2.1041,
+      "step": 32230
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.6599160702826216e-05,
+      "loss": 2.1268,
+      "step": 32235
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.656789930571356e-05,
+      "loss": 2.1062,
+      "step": 32240
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.653664827778155e-05,
+      "loss": 2.1317,
+      "step": 32245
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.650540762413868e-05,
+      "loss": 2.0502,
+      "step": 32250
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.647417734989189e-05,
+      "loss": 2.0947,
+      "step": 32255
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.644295746014632e-05,
+      "loss": 2.0588,
+      "step": 32260
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.64117479600055e-05,
+      "loss": 2.0878,
+      "step": 32265
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.638054885457116e-05,
+      "loss": 2.0714,
+      "step": 32270
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.6349360148943444e-05,
+      "loss": 2.1659,
+      "step": 32275
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.6318181848220666e-05,
+      "loss": 2.1245,
+      "step": 32280
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.628701395749959e-05,
+      "loss": 2.1136,
+      "step": 32285
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.6255856481875125e-05,
+      "loss": 2.1399,
+      "step": 32290
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.622470942644063e-05,
+      "loss": 2.1214,
+      "step": 32295
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.61935727962876e-05,
+      "loss": 2.1115,
+      "step": 32300
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.616244659650598e-05,
+      "loss": 2.0757,
+      "step": 32305
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.6131330832183876e-05,
+      "loss": 2.1035,
+      "step": 32310
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.61002255084078e-05,
+      "loss": 2.1192,
+      "step": 32315
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.606913063026247e-05,
+      "loss": 2.1192,
+      "step": 32320
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.6038046202830986e-05,
+      "loss": 2.1236,
+      "step": 32325
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.600697223119461e-05,
+      "loss": 2.1315,
+      "step": 32330
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.597590872043305e-05,
+      "loss": 2.1363,
+      "step": 32335
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.594485567562417e-05,
+      "loss": 2.1081,
+      "step": 32340
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.5913813101844216e-05,
+      "loss": 2.1254,
+      "step": 32345
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.588278100416764e-05,
+      "loss": 2.1263,
+      "step": 32350
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.5851759387667304e-05,
+      "loss": 2.1654,
+      "step": 32355
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.582074825741419e-05,
+      "loss": 2.0796,
+      "step": 32360
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.5789747618477745e-05,
+      "loss": 2.0745,
+      "step": 32365
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.575875747592556e-05,
+      "loss": 2.1702,
+      "step": 32370
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.572777783482356e-05,
+      "loss": 2.1212,
+      "step": 32375
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 2.147568941116333,
+      "eval_runtime": 188.6791,
+      "eval_samples_per_second": 25.732,
+      "eval_steps_per_second": 3.217,
+      "step": 32376
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.569680870023594e-05,
+      "loss": 2.1229,
+      "step": 32380
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.566585007722525e-05,
+      "loss": 2.1351,
+      "step": 32385
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.56349019708522e-05,
+      "loss": 2.1124,
+      "step": 32390
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.56039643861759e-05,
+      "loss": 2.1189,
+      "step": 32395
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.5573037328253633e-05,
+      "loss": 2.1065,
+      "step": 32400
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.5542120802141076e-05,
+      "loss": 2.0533,
+      "step": 32405
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.5511214812892056e-05,
+      "loss": 2.0774,
+      "step": 32410
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.54803193655588e-05,
+      "loss": 2.1059,
+      "step": 32415
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.5449434465191714e-05,
+      "loss": 2.1062,
+      "step": 32420
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.541856011683955e-05,
+      "loss": 2.1049,
+      "step": 32425
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.538769632554928e-05,
+      "loss": 2.0857,
+      "step": 32430
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.535684309636621e-05,
+      "loss": 2.0965,
+      "step": 32435
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.532600043433384e-05,
+      "loss": 2.1051,
+      "step": 32440
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.5295168344494046e-05,
+      "loss": 2.0904,
+      "step": 32445
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.526434683188684e-05,
+      "loss": 2.1332,
+      "step": 32450
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.523353590155069e-05,
+      "loss": 2.0772,
+      "step": 32455
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.5202735558522104e-05,
+      "loss": 2.091,
+      "step": 32460
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.5171945807836095e-05,
+      "loss": 2.1152,
+      "step": 32465
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.5141166654525745e-05,
+      "loss": 2.1346,
+      "step": 32470
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.5110398103622565e-05,
+      "loss": 2.1273,
+      "step": 32475
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.507964016015619e-05,
+      "loss": 2.0815,
+      "step": 32480
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.5048892829154644e-05,
+      "loss": 2.1316,
+      "step": 32485
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.501815611564412e-05,
+      "loss": 2.1457,
+      "step": 32490
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.498743002464916e-05,
+      "loss": 2.0754,
+      "step": 32495
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.495671456119247e-05,
+      "loss": 2.1398,
+      "step": 32500
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.492600973029514e-05,
+      "loss": 2.1049,
+      "step": 32505
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.48953155369764e-05,
+      "loss": 2.1563,
+      "step": 32510
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.486463198625385e-05,
+      "loss": 2.064,
+      "step": 32515
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.483395908314324e-05,
+      "loss": 2.0636,
+      "step": 32520
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.4803296832658705e-05,
+      "loss": 2.1399,
+      "step": 32525
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.4772645239812506e-05,
+      "loss": 2.1293,
+      "step": 32530
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.4742004309615294e-05,
+      "loss": 2.1169,
+      "step": 32535
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.471137404707584e-05,
+      "loss": 2.1418,
+      "step": 32540
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.46807544572013e-05,
+      "loss": 2.1349,
+      "step": 32545
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.465014554499702e-05,
+      "loss": 2.135,
+      "step": 32550
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.461954731546655e-05,
+      "loss": 2.1444,
+      "step": 32555
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.458895977361183e-05,
+      "loss": 2.067,
+      "step": 32560
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.4558382924432895e-05,
+      "loss": 2.1076,
+      "step": 32565
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.45278167729282e-05,
+      "loss": 2.0666,
+      "step": 32570
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.449726132409428e-05,
+      "loss": 2.0944,
+      "step": 32575
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.446671658292606e-05,
+      "loss": 2.1574,
+      "step": 32580
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.4436182554416605e-05,
+      "loss": 2.0581,
+      "step": 32585
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.440565924355737e-05,
+      "loss": 2.1858,
+      "step": 32590
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.437514665533785e-05,
+      "loss": 2.0587,
+      "step": 32595
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.434464479474602e-05,
+      "loss": 2.069,
+      "step": 32600
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.431415366676789e-05,
+      "loss": 2.1062,
+      "step": 32605
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.428367327638789e-05,
+      "loss": 2.0855,
+      "step": 32610
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.425320362858857e-05,
+      "loss": 2.1335,
+      "step": 32615
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.422274472835082e-05,
+      "loss": 2.1064,
+      "step": 32620
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.419229658065366e-05,
+      "loss": 2.0912,
+      "step": 32625
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.416185919047449e-05,
+      "loss": 2.0801,
+      "step": 32630
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.413143256278881e-05,
+      "loss": 2.1582,
+      "step": 32635
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.410101670257049e-05,
+      "loss": 2.1659,
+      "step": 32640
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.4070611614791525e-05,
+      "loss": 2.0932,
+      "step": 32645
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.404021730442226e-05,
+      "loss": 2.1012,
+      "step": 32650
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.4009833776431156e-05,
+      "loss": 2.1233,
+      "step": 32655
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.3979461035785044e-05,
+      "loss": 2.077,
+      "step": 32660
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.394909908744887e-05,
+      "loss": 2.0467,
+      "step": 32665
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.391874793638592e-05,
+      "loss": 2.1706,
+      "step": 32670
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.3888407587557616e-05,
+      "loss": 2.105,
+      "step": 32675
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.3858078045923714e-05,
+      "loss": 2.1168,
+      "step": 32680
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.38277593164421e-05,
+      "loss": 2.1097,
+      "step": 32685
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.3797451404069006e-05,
+      "loss": 2.1473,
+      "step": 32690
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.376715431375876e-05,
+      "loss": 2.1295,
+      "step": 32695
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.373686805046408e-05,
+      "loss": 2.1092,
+      "step": 32700
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.3706592619135756e-05,
+      "loss": 2.0842,
+      "step": 32705
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.3676328024722945e-05,
+      "loss": 2.1177,
+      "step": 32710
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.3646074272172914e-05,
+      "loss": 2.1235,
+      "step": 32715
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.361583136643126e-05,
+      "loss": 2.0986,
+      "step": 32720
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.3585599312441706e-05,
+      "loss": 2.1135,
+      "step": 32725
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.355537811514633e-05,
+      "loss": 2.1044,
+      "step": 32730
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.3525167779485314e-05,
+      "loss": 2.0889,
+      "step": 32735
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.3494968310397076e-05,
+      "loss": 2.0996,
+      "step": 32740
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.346477971281836e-05,
+      "loss": 2.1198,
+      "step": 32745
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.343460199168401e-05,
+      "loss": 2.1332,
+      "step": 32750
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.340443515192721e-05,
+      "loss": 2.1071,
+      "step": 32755
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.337427919847923e-05,
+      "loss": 2.0825,
+      "step": 32760
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.334413413626969e-05,
+      "loss": 2.0153,
+      "step": 32765
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.331399997022633e-05,
+      "loss": 2.0883,
+      "step": 32770
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.3283876705275196e-05,
+      "loss": 2.0709,
+      "step": 32775
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.3253764346340454e-05,
+      "loss": 2.1007,
+      "step": 32780
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.322366289834461e-05,
+      "loss": 2.0748,
+      "step": 32785
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.3193572366208234e-05,
+      "loss": 2.1302,
+      "step": 32790
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.316349275485027e-05,
+      "loss": 2.1421,
+      "step": 32795
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.3133424069187734e-05,
+      "loss": 2.0654,
+      "step": 32800
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.3103366314136e-05,
+      "loss": 2.1261,
+      "step": 32805
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.3073319494608504e-05,
+      "loss": 2.037,
+      "step": 32810
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.304328361551702e-05,
+      "loss": 2.1411,
+      "step": 32815
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.301325868177144e-05,
+      "loss": 2.1358,
+      "step": 32820
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.2983244698279955e-05,
+      "loss": 2.1383,
+      "step": 32825
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.295324166994888e-05,
+      "loss": 2.146,
+      "step": 32830
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.292324960168282e-05,
+      "loss": 2.0802,
+      "step": 32835
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.28932684983845e-05,
+      "loss": 2.1069,
+      "step": 32840
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.286329836495494e-05,
+      "loss": 2.0808,
+      "step": 32845
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.28333392062933e-05,
+      "loss": 2.0554,
+      "step": 32850
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.2803391027297006e-05,
+      "loss": 2.0801,
+      "step": 32855
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.277345383286161e-05,
+      "loss": 2.1778,
+      "step": 32860
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.274352762788096e-05,
+      "loss": 2.1175,
+      "step": 32865
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.271361241724702e-05,
+      "loss": 2.0687,
+      "step": 32870
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.268370820585006e-05,
+      "loss": 2.1108,
+      "step": 32875
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.2653814998578404e-05,
+      "loss": 2.157,
+      "step": 32880
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.262393280031876e-05,
+      "loss": 2.1145,
+      "step": 32885
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.259406161595584e-05,
+      "loss": 2.1022,
+      "step": 32890
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.2564201450372756e-05,
+      "loss": 2.1194,
+      "step": 32895
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.253435230845063e-05,
+      "loss": 2.13,
+      "step": 32900
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.250451419506895e-05,
+      "loss": 2.0895,
+      "step": 32905
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.247468711510525e-05,
+      "loss": 2.0851,
+      "step": 32910
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.2444871073435424e-05,
+      "loss": 2.1152,
+      "step": 32915
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.2415066074933353e-05,
+      "loss": 2.069,
+      "step": 32920
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.238527212447132e-05,
+      "loss": 2.0658,
+      "step": 32925
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.235548922691964e-05,
+      "loss": 2.0859,
+      "step": 32930
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.232571738714697e-05,
+      "loss": 2.1202,
+      "step": 32935
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.229595661002002e-05,
+      "loss": 2.0504,
+      "step": 32940
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.22662069004038e-05,
+      "loss": 2.0821,
+      "step": 32945
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.2236468263161415e-05,
+      "loss": 2.0971,
+      "step": 32950
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.220674070315427e-05,
+      "loss": 2.0666,
+      "step": 32955
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.217702422524183e-05,
+      "loss": 2.1336,
+      "step": 32960
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.214731883428188e-05,
+      "loss": 2.1363,
+      "step": 32965
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.211762453513029e-05,
+      "loss": 2.1024,
+      "step": 32970
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.208794133264119e-05,
+      "loss": 2.1149,
+      "step": 32975
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.205826923166682e-05,
+      "loss": 2.0797,
+      "step": 32980
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.202860823705769e-05,
+      "loss": 2.1169,
+      "step": 32985
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.19989583536624e-05,
+      "loss": 2.1001,
+      "step": 32990
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.1969319586327825e-05,
+      "loss": 2.1246,
+      "step": 32995
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.193969193989902e-05,
+      "loss": 2.1034,
+      "step": 33000
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.19100754192191e-05,
+      "loss": 2.1729,
+      "step": 33005
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.188047002912953e-05,
+      "loss": 2.141,
+      "step": 33010
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.1850875774469805e-05,
+      "loss": 2.1327,
+      "step": 33015
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.182129266007772e-05,
+      "loss": 2.1,
+      "step": 33020
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.179172069078915e-05,
+      "loss": 2.0861,
+      "step": 33025
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.176215987143824e-05,
+      "loss": 2.1232,
+      "step": 33030
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.1732610206857225e-05,
+      "loss": 2.1149,
+      "step": 33035
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1703071701876605e-05,
+      "loss": 2.1336,
+      "step": 33040
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.167354436132496e-05,
+      "loss": 2.0868,
+      "step": 33045
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.164402819002914e-05,
+      "loss": 2.0783,
+      "step": 33050
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.1614523192814064e-05,
+      "loss": 2.1502,
+      "step": 33055
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.158502937450296e-05,
+      "loss": 2.1266,
+      "step": 33060
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.155554673991707e-05,
+      "loss": 2.0993,
+      "step": 33065
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.152607529387598e-05,
+      "loss": 2.1111,
+      "step": 33070
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1496615041197264e-05,
+      "loss": 2.0308,
+      "step": 33075
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.146716598669684e-05,
+      "loss": 2.0864,
+      "step": 33080
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.143772813518865e-05,
+      "loss": 2.0603,
+      "step": 33085
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.1408301491484937e-05,
+      "loss": 2.141,
+      "step": 33090
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1378886060396e-05,
+      "loss": 2.1034,
+      "step": 33095
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.1349481846730357e-05,
+      "loss": 2.1145,
+      "step": 33100
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.132008885529465e-05,
+      "loss": 2.0443,
+      "step": 33105
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.129070709089378e-05,
+      "loss": 2.0859,
+      "step": 33110
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.12613365583307e-05,
+      "loss": 2.0867,
+      "step": 33115
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.123197726240665e-05,
+      "loss": 2.0912,
+      "step": 33120
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.1202629207920884e-05,
+      "loss": 2.0803,
+      "step": 33125
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.1173292399670985e-05,
+      "loss": 2.0439,
+      "step": 33130
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.1143966842452524e-05,
+      "loss": 2.0853,
+      "step": 33135
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.111465254105939e-05,
+      "loss": 2.1158,
+      "step": 33140
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.108534950028349e-05,
+      "loss": 2.0846,
+      "step": 33145
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1056057724915047e-05,
+      "loss": 2.0711,
+      "step": 33150
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.102677721974228e-05,
+      "loss": 2.0809,
+      "step": 33155
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.099750798955171e-05,
+      "loss": 2.157,
+      "step": 33160
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.096825003912789e-05,
+      "loss": 2.1097,
+      "step": 33165
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.093900337325364e-05,
+      "loss": 2.0523,
+      "step": 33170
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.090976799670982e-05,
+      "loss": 2.1305,
+      "step": 33175
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.0880543914275574e-05,
+      "loss": 2.1224,
+      "step": 33180
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.0851331130728075e-05,
+      "loss": 2.1058,
+      "step": 33185
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.0822129650842766e-05,
+      "loss": 2.1147,
+      "step": 33190
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.0792939479393124e-05,
+      "loss": 2.1534,
+      "step": 33195
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.0763760621150886e-05,
+      "loss": 2.0463,
+      "step": 33200
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.073459308088583e-05,
+      "loss": 2.0548,
+      "step": 33205
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.0705436863366024e-05,
+      "loss": 2.1069,
+      "step": 33210
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.067629197335753e-05,
+      "loss": 2.1261,
+      "step": 33215
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.064715841562469e-05,
+      "loss": 2.1145,
+      "step": 33220
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.061803619492989e-05,
+      "loss": 2.134,
+      "step": 33225
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.0588925316033755e-05,
+      "loss": 2.1158,
+      "step": 33230
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.055982578369496e-05,
+      "loss": 2.1269,
+      "step": 33235
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.053073760267044e-05,
+      "loss": 2.0685,
+      "step": 33240
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.0501660777715136e-05,
+      "loss": 2.1237,
+      "step": 33245
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.0472595313582276e-05,
+      "loss": 2.0819,
+      "step": 33250
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.0443541215023098e-05,
+      "loss": 2.1578,
+      "step": 33255
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.0414498486787114e-05,
+      "loss": 2.0808,
+      "step": 33260
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.0385467133621838e-05,
+      "loss": 2.1093,
+      "step": 33265
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.0356447160273062e-05,
+      "loss": 2.1209,
+      "step": 33270
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.0327438571484613e-05,
+      "loss": 2.0825,
+      "step": 33275
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.0298441371998476e-05,
+      "loss": 2.0699,
+      "step": 33280
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.0269455566554862e-05,
+      "loss": 2.0593,
+      "step": 33285
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.0240481159891964e-05,
+      "loss": 2.0887,
+      "step": 33290
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.0211518156746276e-05,
+      "loss": 2.0938,
+      "step": 33295
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.0182566561852287e-05,
+      "loss": 2.0934,
+      "step": 33300
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.0153626379942758e-05,
+      "loss": 2.1645,
+      "step": 33305
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.012469761574842e-05,
+      "loss": 2.0653,
+      "step": 33310
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.009578027399832e-05,
+      "loss": 2.0988,
+      "step": 33315
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.0066874359419472e-05,
+      "loss": 2.072,
+      "step": 33320
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.0037979876737155e-05,
+      "loss": 2.0606,
+      "step": 33325
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.0009096830674653e-05,
+      "loss": 2.0771,
+      "step": 33330
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.998022522595353e-05,
+      "loss": 2.1285,
+      "step": 33335
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.9951365067293325e-05,
+      "loss": 2.0487,
+      "step": 33340
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9922516359411835e-05,
+      "loss": 2.0599,
+      "step": 33345
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9893679107024873e-05,
+      "loss": 2.127,
+      "step": 33350
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9864853314846498e-05,
+      "loss": 2.1415,
+      "step": 33355
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.9836038987588765e-05,
+      "loss": 2.1125,
+      "step": 33360
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.9807236129962003e-05,
+      "loss": 2.1609,
+      "step": 33365
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.9778444746674495e-05,
+      "loss": 2.0967,
+      "step": 33370
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.9749664842432812e-05,
+      "loss": 2.1073,
+      "step": 33375
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.9720896421941525e-05,
+      "loss": 2.104,
+      "step": 33380
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9692139489903426e-05,
+      "loss": 2.0985,
+      "step": 33385
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.9663394051019334e-05,
+      "loss": 2.1121,
+      "step": 33390
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.963466010998829e-05,
+      "loss": 2.1728,
+      "step": 33395
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9605937671507333e-05,
+      "loss": 2.098,
+      "step": 33400
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.957722674027176e-05,
+      "loss": 2.1859,
+      "step": 33405
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.9548527320974852e-05,
+      "loss": 2.1225,
+      "step": 33410
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.9519839418308137e-05,
+      "loss": 2.1343,
+      "step": 33415
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.9491163036961145e-05,
+      "loss": 2.1592,
+      "step": 33420
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.9462498181621624e-05,
+      "loss": 2.162,
+      "step": 33425
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9433844856975335e-05,
+      "loss": 2.0826,
+      "step": 33430
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9405203067706256e-05,
+      "loss": 2.2001,
+      "step": 33435
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.9376572818496383e-05,
+      "loss": 2.0925,
+      "step": 33440
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.9347954114025934e-05,
+      "loss": 2.0889,
+      "step": 33445
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.9319346958973116e-05,
+      "loss": 2.1169,
+      "step": 33450
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.929075135801437e-05,
+      "loss": 2.1003,
+      "step": 33455
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.926216731582415e-05,
+      "loss": 2.0865,
+      "step": 33460
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.9233594837075053e-05,
+      "loss": 2.0726,
+      "step": 33465
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.920503392643784e-05,
+      "loss": 2.0895,
+      "step": 33470
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.9176484588581264e-05,
+      "loss": 2.0826,
+      "step": 33475
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.9147946828172334e-05,
+      "loss": 2.1132,
+      "step": 33480
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.911942064987603e-05,
+      "loss": 2.0874,
+      "step": 33485
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.9090906058355528e-05,
+      "loss": 2.1231,
+      "step": 33490
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9062403058272048e-05,
+      "loss": 2.1344,
+      "step": 33495
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.9033911654285006e-05,
+      "loss": 2.0437,
+      "step": 33500
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.900543185105178e-05,
+      "loss": 2.178,
+      "step": 33505
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.8976963653228017e-05,
+      "loss": 2.1349,
+      "step": 33510
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.894850706546731e-05,
+      "loss": 2.0411,
+      "step": 33515
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.8920062092421507e-05,
+      "loss": 2.1007,
+      "step": 33520
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.8891628738740394e-05,
+      "loss": 2.0779,
+      "step": 33525
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.8863207009072012e-05,
+      "loss": 2.1085,
+      "step": 33530
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.883479690806238e-05,
+      "loss": 2.163,
+      "step": 33535
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.880639844035572e-05,
+      "loss": 2.1564,
+      "step": 33540
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.8778011610594235e-05,
+      "loss": 2.0918,
+      "step": 33545
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.874963642341837e-05,
+      "loss": 2.0645,
+      "step": 33550
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.872127288346651e-05,
+      "loss": 2.0674,
+      "step": 33555
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.8692920995375283e-05,
+      "loss": 2.0558,
+      "step": 33560
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.866458076377927e-05,
+      "loss": 2.0842,
+      "step": 33565
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.8636252193311297e-05,
+      "loss": 2.0695,
+      "step": 33570
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.860793528860214e-05,
+      "loss": 2.1346,
+      "step": 33575
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.8579630054280804e-05,
+      "loss": 2.1021,
+      "step": 33580
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.8551336494974234e-05,
+      "loss": 2.0791,
+      "step": 33585
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.8523054615307622e-05,
+      "loss": 2.1164,
+      "step": 33590
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.849478441990412e-05,
+      "loss": 2.1033,
+      "step": 33595
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.8466525913385078e-05,
+      "loss": 2.0837,
+      "step": 33600
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.8438279100369848e-05,
+      "loss": 2.1653,
+      "step": 33605
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.8410043985475944e-05,
+      "loss": 2.087,
+      "step": 33610
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.8381820573318873e-05,
+      "loss": 2.1333,
+      "step": 33615
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.835360886851236e-05,
+      "loss": 2.1194,
+      "step": 33620
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.832540887566808e-05,
+      "loss": 2.0311,
+      "step": 33625
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.829722059939591e-05,
+      "loss": 2.0869,
+      "step": 33630
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8269044044303706e-05,
+      "loss": 2.1308,
+      "step": 33635
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.824087921499755e-05,
+      "loss": 2.1545,
+      "step": 33640
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.82127261160814e-05,
+      "loss": 2.1317,
+      "step": 33645
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.8184584752157482e-05,
+      "loss": 2.1332,
+      "step": 33650
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.815645512782601e-05,
+      "loss": 2.1189,
+      "step": 33655
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.8128337247685343e-05,
+      "loss": 2.185,
+      "step": 33660
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.810023111633183e-05,
+      "loss": 2.1136,
+      "step": 33665
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.807213673836001e-05,
+      "loss": 2.1561,
+      "step": 33670
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.8044054118362374e-05,
+      "loss": 2.0794,
+      "step": 33675
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.8015983260929635e-05,
+      "loss": 2.0373,
+      "step": 33680
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.7987924170650436e-05,
+      "loss": 2.0572,
+      "step": 33685
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.7959876852111632e-05,
+      "loss": 2.1006,
+      "step": 33690
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.793184130989802e-05,
+      "loss": 2.1103,
+      "step": 33695
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.7903817548592627e-05,
+      "loss": 2.0721,
+      "step": 33700
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.787580557277638e-05,
+      "loss": 2.095,
+      "step": 33705
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.7847805387028437e-05,
+      "loss": 2.0967,
+      "step": 33710
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.7819816995925908e-05,
+      "loss": 2.113,
+      "step": 33715
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.7791840404044077e-05,
+      "loss": 2.093,
+      "step": 33720
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.776387561595619e-05,
+      "loss": 2.0584,
+      "step": 33725
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.773592263623369e-05,
+      "loss": 2.1211,
+      "step": 33730
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.7707981469445955e-05,
+      "loss": 2.1221,
+      "step": 33735
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.768005212016056e-05,
+      "loss": 2.1245,
+      "step": 33740
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.7652134592943035e-05,
+      "loss": 2.1292,
+      "step": 33745
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.7624228892357073e-05,
+      "loss": 2.035,
+      "step": 33750
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.759633502296435e-05,
+      "loss": 2.0618,
+      "step": 33755
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.7568452989324668e-05,
+      "loss": 2.0993,
+      "step": 33760
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7540582795995894e-05,
+      "loss": 2.098,
+      "step": 33765
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.75127244475339e-05,
+      "loss": 2.0977,
+      "step": 33770
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7484877948492704e-05,
+      "loss": 2.0504,
+      "step": 33775
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.74570433034243e-05,
+      "loss": 2.168,
+      "step": 33780
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.742922051687884e-05,
+      "loss": 2.1293,
+      "step": 33785
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.7401409593404425e-05,
+      "loss": 2.1098,
+      "step": 33790
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.7373610537547345e-05,
+      "loss": 2.1049,
+      "step": 33795
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.734582335385182e-05,
+      "loss": 2.1242,
+      "step": 33800
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7318048046860258e-05,
+      "loss": 2.1309,
+      "step": 33805
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.7290284621112982e-05,
+      "loss": 2.0963,
+      "step": 33810
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.7262533081148533e-05,
+      "loss": 2.0832,
+      "step": 33815
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.723479343150339e-05,
+      "loss": 2.0851,
+      "step": 33820
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.7207065676712118e-05,
+      "loss": 2.0619,
+      "step": 33825
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.717934982130732e-05,
+      "loss": 2.1049,
+      "step": 33830
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.7151645869819732e-05,
+      "loss": 2.025,
+      "step": 33835
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.7123953826778036e-05,
+      "loss": 2.1361,
+      "step": 33840
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.7096273696709095e-05,
+      "loss": 2.1521,
+      "step": 33845
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.7068605484137667e-05,
+      "loss": 2.1194,
+      "step": 33850
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.704094919358672e-05,
+      "loss": 2.1378,
+      "step": 33855
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.7013304829577135e-05,
+      "loss": 2.0659,
+      "step": 33860
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.698567239662798e-05,
+      "loss": 2.0848,
+      "step": 33865
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.695805189925622e-05,
+      "loss": 2.1395,
+      "step": 33870
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.693044334197703e-05,
+      "loss": 2.1224,
+      "step": 33875
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.6902846729303477e-05,
+      "loss": 2.0938,
+      "step": 33880
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.687526206574682e-05,
+      "loss": 2.1362,
+      "step": 33885
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.684768935581623e-05,
+      "loss": 2.0572,
+      "step": 33890
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6820128604019057e-05,
+      "loss": 2.1437,
+      "step": 33895
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6792579814860563e-05,
+      "loss": 2.0965,
+      "step": 33900
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.676504299284418e-05,
+      "loss": 2.0978,
+      "step": 33905
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.6737518142471262e-05,
+      "loss": 2.1379,
+      "step": 33910
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6710005268241335e-05,
+      "loss": 2.1834,
+      "step": 33915
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.668250437465184e-05,
+      "loss": 2.0805,
+      "step": 33920
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.6655015466198375e-05,
+      "loss": 2.1138,
+      "step": 33925
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.6627538547374476e-05,
+      "loss": 2.0712,
+      "step": 33930
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.6600073622671805e-05,
+      "loss": 2.0325,
+      "step": 33935
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.6572620696579975e-05,
+      "loss": 2.0726,
+      "step": 33940
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.6545179773586758e-05,
+      "loss": 2.0655,
+      "step": 33945
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6517750858177826e-05,
+      "loss": 2.1381,
+      "step": 33950
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.649033395483701e-05,
+      "loss": 2.1029,
+      "step": 33955
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.6462929068046073e-05,
+      "loss": 2.1459,
+      "step": 33960
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6435536202284915e-05,
+      "loss": 2.0674,
+      "step": 33965
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.640815536203136e-05,
+      "loss": 2.1311,
+      "step": 33970
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6380786551761404e-05,
+      "loss": 2.1259,
+      "step": 33975
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.6353429775948912e-05,
+      "loss": 2.0491,
+      "step": 33980
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.632608503906595e-05,
+      "loss": 2.1061,
+      "step": 33985
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6298752345582454e-05,
+      "loss": 2.0706,
+      "step": 33990
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6271431699966552e-05,
+      "loss": 2.0877,
+      "step": 33995
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6244123106684272e-05,
+      "loss": 2.0793,
+      "step": 34000
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6216826570199704e-05,
+      "loss": 2.0653,
+      "step": 34005
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.6189542094975052e-05,
+      "loss": 2.1577,
+      "step": 34010
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.616226968547041e-05,
+      "loss": 2.1103,
+      "step": 34015
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.6135009346144023e-05,
+      "loss": 2.1096,
+      "step": 34020
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.6107761081452065e-05,
+      "loss": 2.1155,
+      "step": 34025
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.6080524895848836e-05,
+      "loss": 2.078,
+      "step": 34030
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.605330079378655e-05,
+      "loss": 2.1121,
+      "step": 34035
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.602608877971556e-05,
+      "loss": 2.0931,
+      "step": 34040
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.5998888858084135e-05,
+      "loss": 2.1019,
+      "step": 34045
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.5971701033338668e-05,
+      "loss": 2.1344,
+      "step": 34050
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.5944525309923474e-05,
+      "loss": 2.1265,
+      "step": 34055
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.5917361692280994e-05,
+      "loss": 2.096,
+      "step": 34060
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.589021018485157e-05,
+      "loss": 2.0602,
+      "step": 34065
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.5863070792073707e-05,
+      "loss": 2.0879,
+      "step": 34070
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5835943518383776e-05,
+      "loss": 2.1326,
+      "step": 34075
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.580882836821632e-05,
+      "loss": 2.1221,
+      "step": 34080
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.578172534600376e-05,
+      "loss": 2.1313,
+      "step": 34085
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5754634456176663e-05,
+      "loss": 2.0721,
+      "step": 34090
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.5727555703163474e-05,
+      "loss": 2.0872,
+      "step": 34095
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.5700489091390812e-05,
+      "loss": 2.0368,
+      "step": 34100
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5673434625283144e-05,
+      "loss": 2.0742,
+      "step": 34105
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.5646392309263124e-05,
+      "loss": 2.1267,
+      "step": 34110
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.5619362147751237e-05,
+      "loss": 2.0428,
+      "step": 34115
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.559234414516617e-05,
+      "loss": 2.0596,
+      "step": 34120
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.5565338305924446e-05,
+      "loss": 2.0743,
+      "step": 34125
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.553834463444075e-05,
+      "loss": 2.1176,
+      "step": 34130
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.551136313512764e-05,
+      "loss": 2.1379,
+      "step": 34135
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.5484393812395836e-05,
+      "loss": 2.1306,
+      "step": 34140
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.54574366706539e-05,
+      "loss": 2.0883,
+      "step": 34145
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.5430491714308568e-05,
+      "loss": 2.0848,
+      "step": 34150
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.540355894776444e-05,
+      "loss": 2.1198,
+      "step": 34155
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5376638375424244e-05,
+      "loss": 2.0912,
+      "step": 34160
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.53497300016886e-05,
+      "loss": 2.1491,
+      "step": 34165
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.532283383095625e-05,
+      "loss": 2.1523,
+      "step": 34170
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.5295949867623814e-05,
+      "loss": 2.1161,
+      "step": 34175
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.5269078116086077e-05,
+      "loss": 2.0658,
+      "step": 34180
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.5242218580735678e-05,
+      "loss": 2.1073,
+      "step": 34185
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5215371265963294e-05,
+      "loss": 2.0967,
+      "step": 34190
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.5188536176157686e-05,
+      "loss": 2.1135,
+      "step": 34195
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.5161713315705516e-05,
+      "loss": 2.0407,
+      "step": 34200
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.5134902688991525e-05,
+      "loss": 2.0777,
+      "step": 34205
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.5108104300398383e-05,
+      "loss": 2.0524,
+      "step": 34210
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.5081318154306834e-05,
+      "loss": 2.042,
+      "step": 34215
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.5054544255095535e-05,
+      "loss": 2.116,
+      "step": 34220
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.5027782607141238e-05,
+      "loss": 2.1211,
+      "step": 34225
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.500103321481859e-05,
+      "loss": 2.1443,
+      "step": 34230
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4974296082500346e-05,
+      "loss": 2.1363,
+      "step": 34235
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.494757121455714e-05,
+      "loss": 2.0905,
+      "step": 34240
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4920858615357712e-05,
+      "loss": 2.0992,
+      "step": 34245
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.489415828926869e-05,
+      "loss": 2.0903,
+      "step": 34250
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.48674702406548e-05,
+      "loss": 2.0394,
+      "step": 34255
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.484079447387866e-05,
+      "loss": 2.1185,
+      "step": 34260
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.4814130993300977e-05,
+      "loss": 2.1154,
+      "step": 34265
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.4787479803280357e-05,
+      "loss": 2.0972,
+      "step": 34270
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4760840908173498e-05,
+      "loss": 2.1296,
+      "step": 34275
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.473421431233498e-05,
+      "loss": 2.1453,
+      "step": 34280
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4707600020117473e-05,
+      "loss": 2.1497,
+      "step": 34285
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.4680998035871528e-05,
+      "loss": 2.0906,
+      "step": 34290
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.4654408363945803e-05,
+      "loss": 2.0555,
+      "step": 34295
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.462783100868684e-05,
+      "loss": 2.1402,
+      "step": 34300
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4601265974439268e-05,
+      "loss": 2.0727,
+      "step": 34305
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.4574713265545567e-05,
+      "loss": 2.098,
+      "step": 34310
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4548172886346365e-05,
+      "loss": 2.1137,
+      "step": 34315
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.4521644841180115e-05,
+      "loss": 2.1055,
+      "step": 34320
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4495129134383398e-05,
+      "loss": 2.176,
+      "step": 34325
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.4468625770290644e-05,
+      "loss": 2.1235,
+      "step": 34330
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.4442134753234393e-05,
+      "loss": 2.0824,
+      "step": 34335
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.4415656087545047e-05,
+      "loss": 2.0647,
+      "step": 34340
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.4389189777551102e-05,
+      "loss": 2.1104,
+      "step": 34345
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.436273582757892e-05,
+      "loss": 2.138,
+      "step": 34350
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.4336294241952962e-05,
+      "loss": 2.0378,
+      "step": 34355
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.430986502499555e-05,
+      "loss": 2.1553,
+      "step": 34360
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.4283448181027112e-05,
+      "loss": 2.0734,
+      "step": 34365
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4257043714365903e-05,
+      "loss": 2.0913,
+      "step": 34370
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4230651629328283e-05,
+      "loss": 2.1171,
+      "step": 34375
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.4204271930228495e-05,
+      "loss": 2.1305,
+      "step": 34380
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.4177904621378866e-05,
+      "loss": 2.1497,
+      "step": 34385
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.415154970708957e-05,
+      "loss": 2.1717,
+      "step": 34390
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4125207191668875e-05,
+      "loss": 2.0707,
+      "step": 34395
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.40988770794229e-05,
+      "loss": 2.1512,
+      "step": 34400
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.4072559374655878e-05,
+      "loss": 2.1072,
+      "step": 34405
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.404625408166986e-05,
+      "loss": 2.093,
+      "step": 34410
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4019961204765018e-05,
+      "loss": 2.1087,
+      "step": 34415
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.3993680748239357e-05,
+      "loss": 2.1267,
+      "step": 34420
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.396741271638897e-05,
+      "loss": 2.0623,
+      "step": 34425
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.394115711350782e-05,
+      "loss": 2.088,
+      "step": 34430
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.391491394388793e-05,
+      "loss": 2.0742,
+      "step": 34435
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.3888683211819187e-05,
+      "loss": 2.1181,
+      "step": 34440
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3862464921589565e-05,
+      "loss": 2.1762,
+      "step": 34445
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.383625907748488e-05,
+      "loss": 2.1222,
+      "step": 34450
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.381006568378904e-05,
+      "loss": 2.1203,
+      "step": 34455
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.378388474478378e-05,
+      "loss": 2.1054,
+      "step": 34460
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.375771626474894e-05,
+      "loss": 2.1171,
+      "step": 34465
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.3731560247962192e-05,
+      "loss": 2.1079,
+      "step": 34470
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.370541669869929e-05,
+      "loss": 2.1648,
+      "step": 34475
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.367928562123385e-05,
+      "loss": 2.0862,
+      "step": 34480
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.3653167019837517e-05,
+      "loss": 2.1002,
+      "step": 34485
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.362706089877985e-05,
+      "loss": 2.0946,
+      "step": 34490
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.3600967262328434e-05,
+      "loss": 2.0746,
+      "step": 34495
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3574886114748707e-05,
+      "loss": 2.1187,
+      "step": 34500
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3548817460304196e-05,
+      "loss": 2.0365,
+      "step": 34505
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3522761303256248e-05,
+      "loss": 2.1185,
+      "step": 34510
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.3496717647864298e-05,
+      "loss": 2.0933,
+      "step": 34515
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.3470686498385618e-05,
+      "loss": 2.0718,
+      "step": 34520
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.344466785907552e-05,
+      "loss": 2.0969,
+      "step": 34525
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.3418661734187288e-05,
+      "loss": 2.0864,
+      "step": 34530
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.3392668127972028e-05,
+      "loss": 2.1336,
+      "step": 34535
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.336668704467898e-05,
+      "loss": 2.0866,
+      "step": 34540
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.334071848855519e-05,
+      "loss": 2.1418,
+      "step": 34545
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.331476246384573e-05,
+      "loss": 2.111,
+      "step": 34550
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.328881897479357e-05,
+      "loss": 2.111,
+      "step": 34555
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.326288802563972e-05,
+      "loss": 2.1314,
+      "step": 34560
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.323696962062303e-05,
+      "loss": 2.0797,
+      "step": 34565
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.3211063763980412e-05,
+      "loss": 2.1368,
+      "step": 34570
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.3185170459946603e-05,
+      "loss": 2.1239,
+      "step": 34575
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.315928971275444e-05,
+      "loss": 2.1488,
+      "step": 34580
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3133421526634534e-05,
+      "loss": 2.1098,
+      "step": 34585
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.310756590581561e-05,
+      "loss": 2.1031,
+      "step": 34590
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.308172285452418e-05,
+      "loss": 2.0888,
+      "step": 34595
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.3055892376984866e-05,
+      "loss": 2.1343,
+      "step": 34600
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.3030074477420072e-05,
+      "loss": 2.1046,
+      "step": 34605
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.3004269160050295e-05,
+      "loss": 2.0591,
+      "step": 34610
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.297847642909383e-05,
+      "loss": 2.0741,
+      "step": 34615
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.295269628876706e-05,
+      "loss": 2.119,
+      "step": 34620
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.2926928743284183e-05,
+      "loss": 2.0733,
+      "step": 34625
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2901173796857432e-05,
+      "loss": 2.0518,
+      "step": 34630
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.2875431453696894e-05,
+      "loss": 2.1225,
+      "step": 34635
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.284970171801072e-05,
+      "loss": 2.0859,
+      "step": 34640
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.2823984594004832e-05,
+      "loss": 2.1673,
+      "step": 34645
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.2798280085883263e-05,
+      "loss": 2.0751,
+      "step": 34650
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.277258819784783e-05,
+      "loss": 2.1491,
+      "step": 34655
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2746908934098433e-05,
+      "loss": 2.0995,
+      "step": 34660
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.2721242298832767e-05,
+      "loss": 2.0492,
+      "step": 34665
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2695588296246596e-05,
+      "loss": 2.064,
+      "step": 34670
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2669946930533493e-05,
+      "loss": 2.095,
+      "step": 34675
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.2644318205885075e-05,
+      "loss": 2.126,
+      "step": 34680
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.2618702126490797e-05,
+      "loss": 2.0458,
+      "step": 34685
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.259309869653814e-05,
+      "loss": 2.0812,
+      "step": 34690
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2567507920212426e-05,
+      "loss": 2.0993,
+      "step": 34695
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.2541929801697004e-05,
+      "loss": 2.0729,
+      "step": 34700
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.2516364345173036e-05,
+      "loss": 2.0866,
+      "step": 34705
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.249081155481976e-05,
+      "loss": 2.1185,
+      "step": 34710
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.2465271434814183e-05,
+      "loss": 2.146,
+      "step": 34715
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.2439743989331398e-05,
+      "loss": 2.0908,
+      "step": 34720
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.2414229222544292e-05,
+      "loss": 2.0821,
+      "step": 34725
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.2388727138623778e-05,
+      "loss": 2.1114,
+      "step": 34730
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.236323774173864e-05,
+      "loss": 2.1165,
+      "step": 34735
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.2337761036055582e-05,
+      "loss": 2.094,
+      "step": 34740
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2312297025739293e-05,
+      "loss": 2.1548,
+      "step": 34745
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.228684571495231e-05,
+      "loss": 2.0599,
+      "step": 34750
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.226140710785518e-05,
+      "loss": 2.1429,
+      "step": 34755
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.223598120860627e-05,
+      "loss": 2.1402,
+      "step": 34760
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.2210568021361988e-05,
+      "loss": 2.0807,
+      "step": 34765
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.2185167550276532e-05,
+      "loss": 2.0705,
+      "step": 34770
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.2159779799502168e-05,
+      "loss": 2.1331,
+      "step": 34775
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2134404773188934e-05,
+      "loss": 2.0504,
+      "step": 34780
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.210904247548491e-05,
+      "loss": 2.1241,
+      "step": 34785
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2083692910536e-05,
+      "loss": 2.0734,
+      "step": 34790
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.2058356082486144e-05,
+      "loss": 2.086,
+      "step": 34795
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.203303199547704e-05,
+      "loss": 2.1111,
+      "step": 34800
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.2007720653648455e-05,
+      "loss": 2.104,
+      "step": 34805
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.1982422061137952e-05,
+      "loss": 2.0866,
+      "step": 34810
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.195713622208114e-05,
+      "loss": 2.0772,
+      "step": 34815
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.1931863140611397e-05,
+      "loss": 2.1661,
+      "step": 34820
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1906602820860144e-05,
+      "loss": 2.1635,
+      "step": 34825
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1881355266956617e-05,
+      "loss": 2.0694,
+      "step": 34830
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1856120483028043e-05,
+      "loss": 2.0605,
+      "step": 34835
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1830898473199478e-05,
+      "loss": 2.1608,
+      "step": 34840
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1805689241593997e-05,
+      "loss": 2.1076,
+      "step": 34845
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.1780492792332475e-05,
+      "loss": 2.1373,
+      "step": 34850
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1755309129533807e-05,
+      "loss": 2.0881,
+      "step": 34855
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.173013825731467e-05,
+      "loss": 2.0893,
+      "step": 34860
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.1704980179789792e-05,
+      "loss": 2.1429,
+      "step": 34865
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.1679834901071672e-05,
+      "loss": 2.1071,
+      "step": 34870
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.1654702425270845e-05,
+      "loss": 2.1353,
+      "step": 34875
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1629582756495636e-05,
+      "loss": 2.1051,
+      "step": 34880
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.1604475898852385e-05,
+      "loss": 2.1365,
+      "step": 34885
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1579381856445225e-05,
+      "loss": 2.1278,
+      "step": 34890
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.155430063337631e-05,
+      "loss": 2.1814,
+      "step": 34895
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.15292322337456e-05,
+      "loss": 2.0622,
+      "step": 34900
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.1504176661651033e-05,
+      "loss": 2.0435,
+      "step": 34905
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.147913392118841e-05,
+      "loss": 2.1405,
+      "step": 34910
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1454104016451426e-05,
+      "loss": 2.1246,
+      "step": 34915
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1429086951531686e-05,
+      "loss": 2.1155,
+      "step": 34920
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.140408273051874e-05,
+      "loss": 2.1141,
+      "step": 34925
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1379091357499948e-05,
+      "loss": 2.0694,
+      "step": 34930
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1354112836560702e-05,
+      "loss": 2.1338,
+      "step": 34935
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1329147171784137e-05,
+      "loss": 1.9983,
+      "step": 34940
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1304194367251394e-05,
+      "loss": 2.1115,
+      "step": 34945
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.127925442704152e-05,
+      "loss": 2.137,
+      "step": 34950
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1254327355231362e-05,
+      "loss": 2.1501,
+      "step": 34955
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1229413155895783e-05,
+      "loss": 2.0285,
+      "step": 34960
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.120451183310742e-05,
+      "loss": 2.1073,
+      "step": 34965
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.117962339093692e-05,
+      "loss": 2.0809,
+      "step": 34970
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.1154747833452717e-05,
+      "loss": 2.1488,
+      "step": 34975
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1129885164721252e-05,
+      "loss": 2.1423,
+      "step": 34980
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.110503538880675e-05,
+      "loss": 2.09,
+      "step": 34985
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.108019850977142e-05,
+      "loss": 2.146,
+      "step": 34990
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.1055374531675277e-05,
+      "loss": 2.0952,
+      "step": 34995
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1030563458576323e-05,
+      "loss": 2.0741,
+      "step": 35000
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1005765294530343e-05,
+      "loss": 2.0892,
+      "step": 35005
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0980980043591124e-05,
+      "loss": 2.1707,
+      "step": 35010
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.095620770981023e-05,
+      "loss": 2.111,
+      "step": 35015
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.093144829723721e-05,
+      "loss": 2.1343,
+      "step": 35020
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.0906701809919427e-05,
+      "loss": 2.1513,
+      "step": 35025
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.0881968251902218e-05,
+      "loss": 2.0883,
+      "step": 35030
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0857247627228683e-05,
+      "loss": 2.1498,
+      "step": 35035
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.0832539939939943e-05,
+      "loss": 2.1467,
+      "step": 35040
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.080784519407487e-05,
+      "loss": 2.1182,
+      "step": 35045
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0783163393670356e-05,
+      "loss": 2.0728,
+      "step": 35050
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.0758494542761065e-05,
+      "loss": 2.1421,
+      "step": 35055
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0733838645379612e-05,
+      "loss": 2.1268,
+      "step": 35060
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.070919570555645e-05,
+      "loss": 2.1007,
+      "step": 35065
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.068456572731996e-05,
+      "loss": 2.1195,
+      "step": 35070
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.0659948714696355e-05,
+      "loss": 2.1278,
+      "step": 35075
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0635344671709778e-05,
+      "loss": 2.0919,
+      "step": 35080
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.0610753602382192e-05,
+      "loss": 2.0869,
+      "step": 35085
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.0586175510733553e-05,
+      "loss": 2.1473,
+      "step": 35090
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0561610400781507e-05,
+      "loss": 2.1393,
+      "step": 35095
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.0537058276541765e-05,
+      "loss": 2.0704,
+      "step": 35100
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.0512519142027777e-05,
+      "loss": 2.1136,
+      "step": 35105
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0487993001250995e-05,
+      "loss": 2.0721,
+      "step": 35110
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.0463479858220637e-05,
+      "loss": 2.1216,
+      "step": 35115
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.0438979716943873e-05,
+      "loss": 2.167,
+      "step": 35120
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.041449258142567e-05,
+      "loss": 2.0835,
+      "step": 35125
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.039001845566898e-05,
+      "loss": 2.059,
+      "step": 35130
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.036555734367449e-05,
+      "loss": 2.1008,
+      "step": 35135
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.0341109249440905e-05,
+      "loss": 2.0845,
+      "step": 35140
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0316674176964668e-05,
+      "loss": 2.076,
+      "step": 35145
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.029225213024022e-05,
+      "loss": 2.1151,
+      "step": 35150
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0267843113259734e-05,
+      "loss": 2.1103,
+      "step": 35155
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.0243447130013393e-05,
+      "loss": 2.1538,
+      "step": 35160
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.021906418448912e-05,
+      "loss": 2.0735,
+      "step": 35165
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0194694280672843e-05,
+      "loss": 2.0608,
+      "step": 35170
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0170337422548214e-05,
+      "loss": 2.0842,
+      "step": 35175
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.0145993614096892e-05,
+      "loss": 2.0755,
+      "step": 35180
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.0121662859298262e-05,
+      "loss": 2.0895,
+      "step": 35185
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.009734516212972e-05,
+      "loss": 2.1639,
+      "step": 35190
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.007304052656639e-05,
+      "loss": 2.088,
+      "step": 35195
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.0048748956581387e-05,
+      "loss": 2.0765,
+      "step": 35200
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.002447045614556e-05,
+      "loss": 2.1163,
+      "step": 35205
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.000020502922777e-05,
+      "loss": 2.0823,
+      "step": 35210
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9975952679794584e-05,
+      "loss": 2.1065,
+      "step": 35215
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.995171341181057e-05,
+      "loss": 2.1104,
+      "step": 35220
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9927487229238052e-05,
+      "loss": 2.095,
+      "step": 35225
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9903274136037298e-05,
+      "loss": 2.088,
+      "step": 35230
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9879074136166353e-05,
+      "loss": 2.0503,
+      "step": 35235
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.985488723358122e-05,
+      "loss": 2.0854,
+      "step": 35240
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9830713432235646e-05,
+      "loss": 2.0861,
+      "step": 35245
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.980655273608136e-05,
+      "loss": 2.0979,
+      "step": 35250
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.978240514906784e-05,
+      "loss": 2.1242,
+      "step": 35255
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.97582706751425e-05,
+      "loss": 2.1158,
+      "step": 35260
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9734149318250528e-05,
+      "loss": 2.0748,
+      "step": 35265
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9710041082335097e-05,
+      "loss": 2.0981,
+      "step": 35270
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9685945971337093e-05,
+      "loss": 2.1315,
+      "step": 35275
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9661863989195316e-05,
+      "loss": 2.1229,
+      "step": 35280
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.963779513984647e-05,
+      "loss": 2.1057,
+      "step": 35285
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.961373942722501e-05,
+      "loss": 2.0919,
+      "step": 35290
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9589696855263352e-05,
+      "loss": 2.0914,
+      "step": 35295
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9565667427891654e-05,
+      "loss": 2.1639,
+      "step": 35300
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.954165114903804e-05,
+      "loss": 2.048,
+      "step": 35305
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.951764802262839e-05,
+      "loss": 2.0692,
+      "step": 35310
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9493658052586493e-05,
+      "loss": 2.1035,
+      "step": 35315
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9469681242833926e-05,
+      "loss": 2.0969,
+      "step": 35320
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9445717597290215e-05,
+      "loss": 2.0931,
+      "step": 35325
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.94217671198726e-05,
+      "loss": 2.0854,
+      "step": 35330
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9397829814496303e-05,
+      "loss": 2.047,
+      "step": 35335
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.937390568507429e-05,
+      "loss": 2.1872,
+      "step": 35340
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.934999473551745e-05,
+      "loss": 2.1041,
+      "step": 35345
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9326096969734442e-05,
+      "loss": 2.0934,
+      "step": 35350
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9302212391631845e-05,
+      "loss": 2.1194,
+      "step": 35355
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.927834100511401e-05,
+      "loss": 2.1081,
+      "step": 35360
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9254482814083217e-05,
+      "loss": 2.1296,
+      "step": 35365
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.923063782243949e-05,
+      "loss": 2.1423,
+      "step": 35370
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.920680603408078e-05,
+      "loss": 2.0611,
+      "step": 35375
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.918298745290281e-05,
+      "loss": 2.048,
+      "step": 35380
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.915918208279923e-05,
+      "loss": 2.1125,
+      "step": 35385
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.913538992766143e-05,
+      "loss": 2.1334,
+      "step": 35390
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9111610991378724e-05,
+      "loss": 2.0898,
+      "step": 35395
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.90878452778382e-05,
+      "loss": 2.1515,
+      "step": 35400
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.906409279092486e-05,
+      "loss": 2.0863,
+      "step": 35405
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.904035353452144e-05,
+      "loss": 2.0817,
+      "step": 35410
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9016627512508622e-05,
+      "loss": 2.1881,
+      "step": 35415
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.899291472876483e-05,
+      "loss": 2.0425,
+      "step": 35420
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8969215187166423e-05,
+      "loss": 2.1322,
+      "step": 35425
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8945528891587483e-05,
+      "loss": 2.0847,
+      "step": 35430
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8921855845900037e-05,
+      "loss": 2.1227,
+      "step": 35435
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8898196053973848e-05,
+      "loss": 2.1128,
+      "step": 35440
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8874549519676608e-05,
+      "loss": 2.1627,
+      "step": 35445
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.885091624687373e-05,
+      "loss": 2.0476,
+      "step": 35450
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.882729623942858e-05,
+      "loss": 2.0594,
+      "step": 35455
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.880368950120227e-05,
+      "loss": 2.1284,
+      "step": 35460
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.878009603605374e-05,
+      "loss": 2.1047,
+      "step": 35465
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.875651584783985e-05,
+      "loss": 2.0999,
+      "step": 35470
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8732948940415175e-05,
+      "loss": 2.0282,
+      "step": 35475
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8709395317632216e-05,
+      "loss": 2.0759,
+      "step": 35480
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.868585498334121e-05,
+      "loss": 2.0885,
+      "step": 35485
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8662327941390334e-05,
+      "loss": 2.0685,
+      "step": 35490
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8638814195625475e-05,
+      "loss": 2.092,
+      "step": 35495
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.861531374989044e-05,
+      "loss": 2.0353,
+      "step": 35500
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8591826608026787e-05,
+      "loss": 2.1096,
+      "step": 35505
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8568352773873975e-05,
+      "loss": 2.0513,
+      "step": 35510
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8544892251269207e-05,
+      "loss": 2.1314,
+      "step": 35515
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8521445044047592e-05,
+      "loss": 2.0989,
+      "step": 35520
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8498011156041973e-05,
+      "loss": 2.0845,
+      "step": 35525
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8474590591083118e-05,
+      "loss": 2.1062,
+      "step": 35530
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.84511833529995e-05,
+      "loss": 2.1583,
+      "step": 35535
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.842778944561755e-05,
+      "loss": 2.0846,
+      "step": 35540
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.840440887276137e-05,
+      "loss": 2.0356,
+      "step": 35545
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.838104163825303e-05,
+      "loss": 2.1274,
+      "step": 35550
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8357687745912277e-05,
+      "loss": 2.1108,
+      "step": 35555
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8334347199556812e-05,
+      "loss": 2.0714,
+      "step": 35560
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8311020003002043e-05,
+      "loss": 2.1016,
+      "step": 35565
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.828770616006128e-05,
+      "loss": 2.1204,
+      "step": 35570
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.826440567454558e-05,
+      "loss": 2.0619,
+      "step": 35575
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.824111855026388e-05,
+      "loss": 2.0674,
+      "step": 35580
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8217844791022864e-05,
+      "loss": 2.1071,
+      "step": 35585
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8194584400627122e-05,
+      "loss": 2.0959,
+      "step": 35590
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.817133738287895e-05,
+      "loss": 2.0613,
+      "step": 35595
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.814810374157857e-05,
+      "loss": 2.0705,
+      "step": 35600
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8124883480523902e-05,
+      "loss": 2.0929,
+      "step": 35605
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8101676603510796e-05,
+      "loss": 2.0868,
+      "step": 35610
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8078483114332802e-05,
+      "loss": 2.1244,
+      "step": 35615
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.80553030167814e-05,
+      "loss": 2.0442,
+      "step": 35620
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.803213631464574e-05,
+      "loss": 2.0788,
+      "step": 35625
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.800898301171293e-05,
+      "loss": 2.1555,
+      "step": 35630
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7985843111767787e-05,
+      "loss": 2.0779,
+      "step": 35635
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7962716618592966e-05,
+      "loss": 2.0903,
+      "step": 35640
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7939603535968908e-05,
+      "loss": 2.1432,
+      "step": 35645
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.791650386767394e-05,
+      "loss": 2.0834,
+      "step": 35650
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7893417617484064e-05,
+      "loss": 2.1641,
+      "step": 35655
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7870344789173256e-05,
+      "loss": 2.1296,
+      "step": 35660
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7847285386513124e-05,
+      "loss": 2.1411,
+      "step": 35665
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7824239413273238e-05,
+      "loss": 2.1511,
+      "step": 35670
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7801206873220832e-05,
+      "loss": 2.0776,
+      "step": 35675
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7778187770121078e-05,
+      "loss": 2.1632,
+      "step": 35680
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7755182107736824e-05,
+      "loss": 2.1407,
+      "step": 35685
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7732189889828832e-05,
+      "loss": 2.1777,
+      "step": 35690
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7709211120155577e-05,
+      "loss": 2.1315,
+      "step": 35695
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.768624580247339e-05,
+      "loss": 2.072,
+      "step": 35700
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7663293940536408e-05,
+      "loss": 2.1234,
+      "step": 35705
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.764035553809651e-05,
+      "loss": 2.1101,
+      "step": 35710
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7617430598903462e-05,
+      "loss": 2.098,
+      "step": 35715
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7594519126704722e-05,
+      "loss": 2.0721,
+      "step": 35720
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7571621125245673e-05,
+      "loss": 2.1003,
+      "step": 35725
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7548736598269354e-05,
+      "loss": 2.163,
+      "step": 35730
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.752586554951674e-05,
+      "loss": 2.0763,
+      "step": 35735
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.750300798272648e-05,
+      "loss": 2.0903,
+      "step": 35740
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.748016390163514e-05,
+      "loss": 2.1174,
+      "step": 35745
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7457333309976952e-05,
+      "loss": 2.1115,
+      "step": 35750
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7434516211484077e-05,
+      "loss": 2.1205,
+      "step": 35755
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7411712609886332e-05,
+      "loss": 2.0595,
+      "step": 35760
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7388922508911465e-05,
+      "loss": 2.0546,
+      "step": 35765
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7366145912284902e-05,
+      "loss": 2.0706,
+      "step": 35770
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7343382823729947e-05,
+      "loss": 2.0856,
+      "step": 35775
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.732063324696761e-05,
+      "loss": 2.1243,
+      "step": 35780
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7297897185716806e-05,
+      "loss": 2.0889,
+      "step": 35785
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7275174643694102e-05,
+      "loss": 2.1227,
+      "step": 35790
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7252465624614e-05,
+      "loss": 2.0939,
+      "step": 35795
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.722977013218867e-05,
+      "loss": 2.0788,
+      "step": 35800
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.720708817012816e-05,
+      "loss": 2.0252,
+      "step": 35805
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7184419742140212e-05,
+      "loss": 2.0708,
+      "step": 35810
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7161764851930472e-05,
+      "loss": 2.0291,
+      "step": 35815
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7139123503202294e-05,
+      "loss": 2.1409,
+      "step": 35820
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7116495699656808e-05,
+      "loss": 2.1464,
+      "step": 35825
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7093881444992964e-05,
+      "loss": 2.1234,
+      "step": 35830
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7071280742907525e-05,
+      "loss": 2.1367,
+      "step": 35835
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7048693597094967e-05,
+      "loss": 2.1061,
+      "step": 35840
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7026120011247616e-05,
+      "loss": 2.0452,
+      "step": 35845
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7003559989055516e-05,
+      "loss": 2.1442,
+      "step": 35850
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.698101353420659e-05,
+      "loss": 2.0382,
+      "step": 35855
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.695848065038642e-05,
+      "loss": 2.1142,
+      "step": 35860
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6935961341278484e-05,
+      "loss": 2.0901,
+      "step": 35865
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6913455610563944e-05,
+      "loss": 2.0913,
+      "step": 35870
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.689096346192184e-05,
+      "loss": 2.0831,
+      "step": 35875
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.686848489902889e-05,
+      "loss": 2.1498,
+      "step": 35880
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6846019925559688e-05,
+      "loss": 2.1396,
+      "step": 35885
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6823568545186507e-05,
+      "loss": 2.0996,
+      "step": 35890
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.680113076157952e-05,
+      "loss": 2.1008,
+      "step": 35895
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6778706578406544e-05,
+      "loss": 2.0679,
+      "step": 35900
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6756295999333283e-05,
+      "loss": 2.0842,
+      "step": 35905
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6733899028023136e-05,
+      "loss": 2.1419,
+      "step": 35910
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6711515668137344e-05,
+      "loss": 2.114,
+      "step": 35915
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.668914592333486e-05,
+      "loss": 2.1824,
+      "step": 35920
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6666789797272487e-05,
+      "loss": 2.0917,
+      "step": 35925
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6644447293604694e-05,
+      "loss": 2.1131,
+      "step": 35930
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6622118415983868e-05,
+      "loss": 2.1057,
+      "step": 35935
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.659980316806e-05,
+      "loss": 2.0631,
+      "step": 35940
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6577501553481023e-05,
+      "loss": 2.0432,
+      "step": 35945
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6555213575892503e-05,
+      "loss": 2.0716,
+      "step": 35950
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6532939238937863e-05,
+      "loss": 2.1085,
+      "step": 35955
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.651067854625824e-05,
+      "loss": 2.1118,
+      "step": 35960
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6488431501492596e-05,
+      "loss": 2.1274,
+      "step": 35965
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6466198108277597e-05,
+      "loss": 2.1184,
+      "step": 35970
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.644397837024777e-05,
+      "loss": 2.0922,
+      "step": 35975
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6421772291035275e-05,
+      "loss": 2.1191,
+      "step": 35980
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.63995798742702e-05,
+      "loss": 2.0674,
+      "step": 35985
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.637740112358024e-05,
+      "loss": 2.0529,
+      "step": 35990
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6355236042590992e-05,
+      "loss": 2.09,
+      "step": 35995
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6333084634925744e-05,
+      "loss": 2.1602,
+      "step": 36000
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6310946904205515e-05,
+      "loss": 2.102,
+      "step": 36005
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.628882285404921e-05,
+      "loss": 2.1022,
+      "step": 36010
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6266712488073375e-05,
+      "loss": 2.1348,
+      "step": 36015
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6244615809892404e-05,
+      "loss": 2.1012,
+      "step": 36020
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6222532823118385e-05,
+      "loss": 2.0497,
+      "step": 36025
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.620046353136123e-05,
+      "loss": 2.1349,
+      "step": 36030
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6178407938228556e-05,
+      "loss": 2.0657,
+      "step": 36035
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6156366047325798e-05,
+      "loss": 2.137,
+      "step": 36040
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6134337862256088e-05,
+      "loss": 2.105,
+      "step": 36045
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6112323386620387e-05,
+      "loss": 2.1521,
+      "step": 36050
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.609032262401734e-05,
+      "loss": 2.0647,
+      "step": 36055
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6068335578043438e-05,
+      "loss": 2.0778,
+      "step": 36060
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6046362252292824e-05,
+      "loss": 2.1193,
+      "step": 36065
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6024402650357505e-05,
+      "loss": 2.1671,
+      "step": 36070
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6002456775827157e-05,
+      "loss": 2.0939,
+      "step": 36075
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5980524632289285e-05,
+      "loss": 2.1538,
+      "step": 36080
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5958606223329076e-05,
+      "loss": 2.1432,
+      "step": 36085
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.593670155252954e-05,
+      "loss": 2.1339,
+      "step": 36090
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5914810623471387e-05,
+      "loss": 2.0511,
+      "step": 36095
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5892933439733148e-05,
+      "loss": 2.0857,
+      "step": 36100
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5871070004890998e-05,
+      "loss": 2.1555,
+      "step": 36105
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5849220322518998e-05,
+      "loss": 2.0842,
+      "step": 36110
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.582738439618884e-05,
+      "loss": 2.1061,
+      "step": 36115
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.580556222947005e-05,
+      "loss": 2.0942,
+      "step": 36120
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5783753825929858e-05,
+      "loss": 2.0982,
+      "step": 36125
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.576195918913329e-05,
+      "loss": 2.074,
+      "step": 36130
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.574017832264305e-05,
+      "loss": 2.1093,
+      "step": 36135
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.571841123001968e-05,
+      "loss": 2.148,
+      "step": 36140
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5696657914821378e-05,
+      "loss": 2.0812,
+      "step": 36145
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5674918380604186e-05,
+      "loss": 2.0876,
+      "step": 36150
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5653192630921797e-05,
+      "loss": 2.0475,
+      "step": 36155
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5631480669325736e-05,
+      "loss": 2.0799,
+      "step": 36160
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.56097824993652e-05,
+      "loss": 2.1471,
+      "step": 36165
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.55880981245872e-05,
+      "loss": 2.1094,
+      "step": 36170
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5566427548536424e-05,
+      "loss": 2.0988,
+      "step": 36175
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.554477077475538e-05,
+      "loss": 2.0686,
+      "step": 36180
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5523127806784243e-05,
+      "loss": 2.1555,
+      "step": 36185
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5501498648160973e-05,
+      "loss": 2.0349,
+      "step": 36190
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5479883302421283e-05,
+      "loss": 2.1267,
+      "step": 36195
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.545828177309858e-05,
+      "loss": 2.0615,
+      "step": 36200
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.54366940637241e-05,
+      "loss": 2.0585,
+      "step": 36205
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5415120177826692e-05,
+      "loss": 2.1525,
+      "step": 36210
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.539356011893308e-05,
+      "loss": 2.1054,
+      "step": 36215
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5372013890567625e-05,
+      "loss": 2.1451,
+      "step": 36220
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.53504814962525e-05,
+      "loss": 2.1668,
+      "step": 36225
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.532896293950754e-05,
+      "loss": 2.131,
+      "step": 36230
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5307458223850402e-05,
+      "loss": 2.1033,
+      "step": 36235
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5285967352796404e-05,
+      "loss": 2.0929,
+      "step": 36240
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5264490329858684e-05,
+      "loss": 2.1303,
+      "step": 36245
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5243027158548019e-05,
+      "loss": 2.1125,
+      "step": 36250
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5221577842373003e-05,
+      "loss": 2.1315,
+      "step": 36255
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5200142384839921e-05,
+      "loss": 2.0871,
+      "step": 36260
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5178720789452815e-05,
+      "loss": 2.1071,
+      "step": 36265
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.515731305971343e-05,
+      "loss": 2.1348,
+      "step": 36270
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5135919199121307e-05,
+      "loss": 2.0936,
+      "step": 36275
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5114539211173629e-05,
+      "loss": 2.0713,
+      "step": 36280
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5093173099365399e-05,
+      "loss": 2.0716,
+      "step": 36285
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5071820867189279e-05,
+      "loss": 2.1158,
+      "step": 36290
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5050482518135744e-05,
+      "loss": 2.1198,
+      "step": 36295
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5029158055692894e-05,
+      "loss": 2.1437,
+      "step": 36300
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.500784748334667e-05,
+      "loss": 2.0739,
+      "step": 36305
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4986550804580645e-05,
+      "loss": 2.0916,
+      "step": 36310
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4965268022876222e-05,
+      "loss": 2.0566,
+      "step": 36315
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4943999141712405e-05,
+      "loss": 2.1441,
+      "step": 36320
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4922744164566061e-05,
+      "loss": 2.0952,
+      "step": 36325
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4901503094911662e-05,
+      "loss": 2.1048,
+      "step": 36330
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.488027593622151e-05,
+      "loss": 2.16,
+      "step": 36335
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4859062691965552e-05,
+      "loss": 2.1028,
+      "step": 36340
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4837863365611537e-05,
+      "loss": 2.1072,
+      "step": 36345
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4816677960624858e-05,
+      "loss": 2.0474,
+      "step": 36350
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4795506480468691e-05,
+      "loss": 2.1135,
+      "step": 36355
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.477434892860392e-05,
+      "loss": 2.1232,
+      "step": 36360
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4753205308489149e-05,
+      "loss": 2.0979,
+      "step": 36365
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4732075623580654e-05,
+      "loss": 2.1205,
+      "step": 36370
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4710959877332552e-05,
+      "loss": 2.0985,
+      "step": 36375
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4689858073196572e-05,
+      "loss": 2.0855,
+      "step": 36380
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.466877021462223e-05,
+      "loss": 2.0652,
+      "step": 36385
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4647696305056701e-05,
+      "loss": 2.0412,
+      "step": 36390
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4626636347944955e-05,
+      "loss": 2.1331,
+      "step": 36395
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4605590346729604e-05,
+      "loss": 2.1011,
+      "step": 36400
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.458455830485107e-05,
+      "loss": 2.0794,
+      "step": 36405
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4563540225747374e-05,
+      "loss": 2.1198,
+      "step": 36410
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4542536112854366e-05,
+      "loss": 2.0985,
+      "step": 36415
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4521545969605543e-05,
+      "loss": 2.0851,
+      "step": 36420
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4500569799432174e-05,
+      "loss": 2.0966,
+      "step": 36425
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4479607605763156e-05,
+      "loss": 2.1167,
+      "step": 36430
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.445865939202522e-05,
+      "loss": 2.0686,
+      "step": 36435
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4437725161642701e-05,
+      "loss": 2.1021,
+      "step": 36440
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.441680491803773e-05,
+      "loss": 2.098,
+      "step": 36445
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4395898664630092e-05,
+      "loss": 2.1396,
+      "step": 36450
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4375006404837332e-05,
+      "loss": 2.1273,
+      "step": 36455
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4354128142074663e-05,
+      "loss": 2.1106,
+      "step": 36460
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4333263879755044e-05,
+      "loss": 2.1036,
+      "step": 36465
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4312413621289156e-05,
+      "loss": 2.1486,
+      "step": 36470
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.429157737008533e-05,
+      "loss": 2.1069,
+      "step": 36475
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4270755129549696e-05,
+      "loss": 2.0903,
+      "step": 36480
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4249946903085986e-05,
+      "loss": 2.0915,
+      "step": 36485
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.422915269409575e-05,
+      "loss": 2.0912,
+      "step": 36490
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4208372505978163e-05,
+      "loss": 2.0747,
+      "step": 36495
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4187606342130166e-05,
+      "loss": 2.0685,
+      "step": 36500
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4166854205946357e-05,
+      "loss": 2.1066,
+      "step": 36505
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4146116100819085e-05,
+      "loss": 2.1462,
+      "step": 36510
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4125392030138374e-05,
+      "loss": 2.1425,
+      "step": 36515
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4104681997291991e-05,
+      "loss": 2.1139,
+      "step": 36520
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.408398600566535e-05,
+      "loss": 2.0548,
+      "step": 36525
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4063304058641635e-05,
+      "loss": 2.0739,
+      "step": 36530
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4042636159601663e-05,
+      "loss": 2.1254,
+      "step": 36535
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4021982311924053e-05,
+      "loss": 2.0576,
+      "step": 36540
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4001342518985028e-05,
+      "loss": 2.0826,
+      "step": 36545
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3980716784158564e-05,
+      "loss": 2.1718,
+      "step": 36550
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3960105110816301e-05,
+      "loss": 2.0911,
+      "step": 36555
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3939507502327653e-05,
+      "loss": 2.1274,
+      "step": 36560
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3918923962059648e-05,
+      "loss": 2.1247,
+      "step": 36565
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3898354493377096e-05,
+      "loss": 2.1169,
+      "step": 36570
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.387779909964243e-05,
+      "loss": 2.127,
+      "step": 36575
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.385725778421586e-05,
+      "loss": 2.1682,
+      "step": 36580
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3836730550455213e-05,
+      "loss": 2.1047,
+      "step": 36585
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3816217401716092e-05,
+      "loss": 2.091,
+      "step": 36590
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3795718341351726e-05,
+      "loss": 2.1465,
+      "step": 36595
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3775233372713103e-05,
+      "loss": 2.1105,
+      "step": 36600
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3754762499148854e-05,
+      "loss": 2.051,
+      "step": 36605
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.373430572400537e-05,
+      "loss": 2.1221,
+      "step": 36610
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3713863050626652e-05,
+      "loss": 2.1663,
+      "step": 36615
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3693434482354484e-05,
+      "loss": 2.1464,
+      "step": 36620
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.367302002252826e-05,
+      "loss": 2.1314,
+      "step": 36625
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3652619674485167e-05,
+      "loss": 2.097,
+      "step": 36630
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3632233441559971e-05,
+      "loss": 2.1788,
+      "step": 36635
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.361186132708524e-05,
+      "loss": 2.0932,
+      "step": 36640
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3591503334391132e-05,
+      "loss": 2.1137,
+      "step": 36645
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3571159466805594e-05,
+      "loss": 2.1505,
+      "step": 36650
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.355082972765418e-05,
+      "loss": 2.0506,
+      "step": 36655
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3530514120260219e-05,
+      "loss": 2.1499,
+      "step": 36660
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.351021264794462e-05,
+      "loss": 2.1165,
+      "step": 36665
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3489925314026097e-05,
+      "loss": 2.1139,
+      "step": 36670
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3469652121820964e-05,
+      "loss": 2.0854,
+      "step": 36675
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.344939307464329e-05,
+      "loss": 2.1463,
+      "step": 36680
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3429148175804762e-05,
+      "loss": 2.0926,
+      "step": 36685
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3408917428614842e-05,
+      "loss": 2.1181,
+      "step": 36690
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3388700836380574e-05,
+      "loss": 2.1158,
+      "step": 36695
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3368498402406804e-05,
+      "loss": 2.0665,
+      "step": 36700
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3348310129995945e-05,
+      "loss": 2.0795,
+      "step": 36705
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3328136022448201e-05,
+      "loss": 2.1396,
+      "step": 36710
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3307976083061368e-05,
+      "loss": 2.1139,
+      "step": 36715
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.328783031513101e-05,
+      "loss": 2.0714,
+      "step": 36720
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.326769872195034e-05,
+      "loss": 2.0744,
+      "step": 36725
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3247581306810187e-05,
+      "loss": 2.1555,
+      "step": 36730
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.32274780729992e-05,
+      "loss": 2.1117,
+      "step": 36735
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3207389023803574e-05,
+      "loss": 2.0993,
+      "step": 36740
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.318731416250729e-05,
+      "loss": 2.1214,
+      "step": 36745
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.316725349239193e-05,
+      "loss": 2.0933,
+      "step": 36750
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3147207016736828e-05,
+      "loss": 2.1312,
+      "step": 36755
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3127174738818916e-05,
+      "loss": 2.1191,
+      "step": 36760
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3107156661912889e-05,
+      "loss": 2.0955,
+      "step": 36765
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3087152789291058e-05,
+      "loss": 2.037,
+      "step": 36770
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3067163124223458e-05,
+      "loss": 2.108,
+      "step": 36775
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3047187669977746e-05,
+      "loss": 2.1263,
+      "step": 36780
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3027226429819328e-05,
+      "loss": 2.0591,
+      "step": 36785
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.300727940701122e-05,
+      "loss": 2.0864,
+      "step": 36790
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2987346604814155e-05,
+      "loss": 2.1087,
+      "step": 36795
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2967428026486506e-05,
+      "loss": 2.0454,
+      "step": 36800
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2947523675284378e-05,
+      "loss": 2.0571,
+      "step": 36805
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2927633554461482e-05,
+      "loss": 2.0982,
+      "step": 36810
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.290775766726926e-05,
+      "loss": 2.1212,
+      "step": 36815
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.288789601695678e-05,
+      "loss": 2.1002,
+      "step": 36820
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2868048606770833e-05,
+      "loss": 2.0789,
+      "step": 36825
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2848215439955824e-05,
+      "loss": 2.1246,
+      "step": 36830
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2828396519753894e-05,
+      "loss": 2.0982,
+      "step": 36835
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2808591849404784e-05,
+      "loss": 2.1015,
+      "step": 36840
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.278880143214598e-05,
+      "loss": 2.1115,
+      "step": 36845
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2769025271212576e-05,
+      "loss": 2.1335,
+      "step": 36850
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2749263369837382e-05,
+      "loss": 2.1294,
+      "step": 36855
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2729515731250818e-05,
+      "loss": 2.1675,
+      "step": 36860
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2709782358681066e-05,
+      "loss": 2.1126,
+      "step": 36865
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2690063255353857e-05,
+      "loss": 2.1036,
+      "step": 36870
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2670358424492711e-05,
+      "loss": 2.1517,
+      "step": 36875
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2650667869318711e-05,
+      "loss": 2.0934,
+      "step": 36880
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2630991593050701e-05,
+      "loss": 2.1076,
+      "step": 36885
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2611329598905076e-05,
+      "loss": 2.1158,
+      "step": 36890
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.259168189009603e-05,
+      "loss": 2.096,
+      "step": 36895
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2572048469835295e-05,
+      "loss": 2.1083,
+      "step": 36900
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2552429341332383e-05,
+      "loss": 2.0918,
+      "step": 36905
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2532824507794372e-05,
+      "loss": 2.1675,
+      "step": 36910
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2513233972426042e-05,
+      "loss": 2.1128,
+      "step": 36915
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2493657738429875e-05,
+      "loss": 2.1314,
+      "step": 36920
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2474095809005926e-05,
+      "loss": 2.1425,
+      "step": 36925
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2454548187352011e-05,
+      "loss": 2.1056,
+      "step": 36930
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.243501487666353e-05,
+      "loss": 2.0739,
+      "step": 36935
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2415495880133599e-05,
+      "loss": 2.0783,
+      "step": 36940
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2395991200952927e-05,
+      "loss": 2.1341,
+      "step": 36945
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2376500842309979e-05,
+      "loss": 2.0883,
+      "step": 36950
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2357024807390771e-05,
+      "loss": 2.1354,
+      "step": 36955
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.233756309937908e-05,
+      "loss": 2.1563,
+      "step": 36960
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2318115721456236e-05,
+      "loss": 2.161,
+      "step": 36965
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2298682676801343e-05,
+      "loss": 2.0991,
+      "step": 36970
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2279263968591049e-05,
+      "loss": 2.134,
+      "step": 36975
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2259859599999746e-05,
+      "loss": 2.1036,
+      "step": 36980
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.224046957419941e-05,
+      "loss": 2.1286,
+      "step": 36985
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2221093894359747e-05,
+      "loss": 2.1103,
+      "step": 36990
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2201732563648049e-05,
+      "loss": 2.1273,
+      "step": 36995
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2182385585229328e-05,
+      "loss": 2.1435,
+      "step": 37000
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2163052962266175e-05,
+      "loss": 2.1252,
+      "step": 37005
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2143734697918907e-05,
+      "loss": 2.1125,
+      "step": 37010
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2124430795345432e-05,
+      "loss": 2.1548,
+      "step": 37015
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2105141257701369e-05,
+      "loss": 2.1097,
+      "step": 37020
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.208586608813993e-05,
+      "loss": 2.1223,
+      "step": 37025
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.206660528981204e-05,
+      "loss": 2.033,
+      "step": 37030
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2047358865866199e-05,
+      "loss": 2.1514,
+      "step": 37035
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2028126819448649e-05,
+      "loss": 2.0764,
+      "step": 37040
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2008909153703186e-05,
+      "loss": 2.0949,
+      "step": 37045
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1989705871771339e-05,
+      "loss": 2.1208,
+      "step": 37050
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1970516976792212e-05,
+      "loss": 2.075,
+      "step": 37055
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.195134247190265e-05,
+      "loss": 2.0563,
+      "step": 37060
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1932182360237032e-05,
+      "loss": 2.1215,
+      "step": 37065
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1913036644927477e-05,
+      "loss": 2.1111,
+      "step": 37070
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1893905329103695e-05,
+      "loss": 2.0871,
+      "step": 37075
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.187478841589308e-05,
+      "loss": 2.1233,
+      "step": 37080
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1855685908420633e-05,
+      "loss": 2.1393,
+      "step": 37085
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.183659780980908e-05,
+      "loss": 2.1055,
+      "step": 37090
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.1817524123178635e-05,
+      "loss": 2.049,
+      "step": 37095
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.179846485164734e-05,
+      "loss": 2.1154,
+      "step": 37100
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.1779419998330733e-05,
+      "loss": 2.0919,
+      "step": 37105
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1760389566342112e-05,
+      "loss": 2.1236,
+      "step": 37110
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1741373558792301e-05,
+      "loss": 2.0746,
+      "step": 37115
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1722371978789892e-05,
+      "loss": 2.1624,
+      "step": 37120
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1703384829441e-05,
+      "loss": 2.1193,
+      "step": 37125
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1684412113849486e-05,
+      "loss": 2.0807,
+      "step": 37130
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1665453835116746e-05,
+      "loss": 2.082,
+      "step": 37135
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1646509996341914e-05,
+      "loss": 2.1154,
+      "step": 37140
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.162758060062168e-05,
+      "loss": 2.1292,
+      "step": 37145
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1608665651050466e-05,
+      "loss": 2.1528,
+      "step": 37150
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1589765150720222e-05,
+      "loss": 2.0689,
+      "step": 37155
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1570879102720645e-05,
+      "loss": 2.1032,
+      "step": 37160
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1552007510138974e-05,
+      "loss": 2.1426,
+      "step": 37165
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.1533150376060165e-05,
+      "loss": 2.1208,
+      "step": 37170
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.151430770356674e-05,
+      "loss": 2.0937,
+      "step": 37175
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1495479495738925e-05,
+      "loss": 2.1173,
+      "step": 37180
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1476665755654525e-05,
+      "loss": 2.1215,
+      "step": 37185
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1457866486389023e-05,
+      "loss": 2.0898,
+      "step": 37190
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1439081691015474e-05,
+      "loss": 2.0995,
+      "step": 37195
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1420311372604676e-05,
+      "loss": 2.0973,
+      "step": 37200
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1401555534224928e-05,
+      "loss": 2.1104,
+      "step": 37205
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.138281417894227e-05,
+      "loss": 2.1386,
+      "step": 37210
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1364087309820305e-05,
+      "loss": 2.0941,
+      "step": 37215
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1345374929920305e-05,
+      "loss": 2.109,
+      "step": 37220
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1326677042301193e-05,
+      "loss": 2.1039,
+      "step": 37225
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1307993650019443e-05,
+      "loss": 2.1726,
+      "step": 37230
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.128932475612926e-05,
+      "loss": 2.0756,
+      "step": 37235
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1270670363682378e-05,
+      "loss": 2.1117,
+      "step": 37240
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.1252030475728259e-05,
+      "loss": 2.1227,
+      "step": 37245
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1233405095313899e-05,
+      "loss": 2.0707,
+      "step": 37250
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1214794225484026e-05,
+      "loss": 2.0902,
+      "step": 37255
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1196197869280877e-05,
+      "loss": 2.1365,
+      "step": 37260
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1177616029744443e-05,
+      "loss": 2.1067,
+      "step": 37265
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1159048709912246e-05,
+      "loss": 2.127,
+      "step": 37270
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.114049591281946e-05,
+      "loss": 2.0752,
+      "step": 37275
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1121957641498892e-05,
+      "loss": 2.0989,
+      "step": 37280
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.1103433898980986e-05,
+      "loss": 2.1207,
+      "step": 37285
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1084924688293786e-05,
+      "loss": 2.1198,
+      "step": 37290
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1066430012463002e-05,
+      "loss": 2.1408,
+      "step": 37295
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1047949874511898e-05,
+      "loss": 2.1397,
+      "step": 37300
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1029484277461444e-05,
+      "loss": 2.0873,
+      "step": 37305
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1011033224330148e-05,
+      "loss": 2.0868,
+      "step": 37310
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.0992596718134241e-05,
+      "loss": 2.069,
+      "step": 37315
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0974174761887469e-05,
+      "loss": 2.0612,
+      "step": 37320
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.0955767358601287e-05,
+      "loss": 2.075,
+      "step": 37325
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.0937374511284703e-05,
+      "loss": 2.1004,
+      "step": 37330
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.09189962229444e-05,
+      "loss": 2.1061,
+      "step": 37335
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.090063249658464e-05,
+      "loss": 2.1084,
+      "step": 37340
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0882283335207356e-05,
+      "loss": 2.1042,
+      "step": 37345
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0863948741812013e-05,
+      "loss": 2.0378,
+      "step": 37350
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0845628719395805e-05,
+      "loss": 2.1191,
+      "step": 37355
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0827323270953438e-05,
+      "loss": 2.1443,
+      "step": 37360
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0809032399477337e-05,
+      "loss": 2.1008,
+      "step": 37365
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.0790756107957434e-05,
+      "loss": 2.0901,
+      "step": 37370
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0772494399381384e-05,
+      "loss": 2.0528,
+      "step": 37375
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0754247276734374e-05,
+      "loss": 2.0756,
+      "step": 37380
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0736014742999278e-05,
+      "loss": 2.1139,
+      "step": 37385
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.071779680115651e-05,
+      "loss": 2.0895,
+      "step": 37390
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.0699593454184177e-05,
+      "loss": 2.0696,
+      "step": 37395
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.068140470505793e-05,
+      "loss": 2.1434,
+      "step": 37400
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.0663230556751092e-05,
+      "loss": 2.1264,
+      "step": 37405
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0645071012234542e-05,
+      "loss": 2.1025,
+      "step": 37410
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.062692607447684e-05,
+      "loss": 2.0964,
+      "step": 37415
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0608795746444067e-05,
+      "loss": 2.1162,
+      "step": 37420
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0590680031100042e-05,
+      "loss": 2.0419,
+      "step": 37425
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0572578931406052e-05,
+      "loss": 2.0933,
+      "step": 37430
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.0554492450321107e-05,
+      "loss": 2.0593,
+      "step": 37435
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0536420590801765e-05,
+      "loss": 2.0875,
+      "step": 37440
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.051836335580224e-05,
+      "loss": 2.063,
+      "step": 37445
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0500320748274318e-05,
+      "loss": 2.1005,
+      "step": 37450
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.0482292771167368e-05,
+      "loss": 2.086,
+      "step": 37455
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.046427942742847e-05,
+      "loss": 2.1526,
+      "step": 37460
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0446280720002189e-05,
+      "loss": 2.0495,
+      "step": 37465
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.0428296651830805e-05,
+      "loss": 2.1223,
+      "step": 37470
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.041032722585411e-05,
+      "loss": 2.1731,
+      "step": 37475
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.039237244500958e-05,
+      "loss": 2.0788,
+      "step": 37480
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.037443231223224e-05,
+      "loss": 2.0921,
+      "step": 37485
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.0356506830454782e-05,
+      "loss": 2.1345,
+      "step": 37490
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0338596002607425e-05,
+      "loss": 2.0887,
+      "step": 37495
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0320699831618064e-05,
+      "loss": 2.1157,
+      "step": 37500
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.0302818320412133e-05,
+      "loss": 2.1119,
+      "step": 37505
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0284951471912751e-05,
+      "loss": 2.1458,
+      "step": 37510
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.026709928904055e-05,
+      "loss": 2.0622,
+      "step": 37515
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0249261774713847e-05,
+      "loss": 2.1445,
+      "step": 37520
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0231438931848481e-05,
+      "loss": 2.0865,
+      "step": 37525
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0213630763357973e-05,
+      "loss": 2.0592,
+      "step": 37530
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0195837272153374e-05,
+      "loss": 2.1246,
+      "step": 37535
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0178058461143392e-05,
+      "loss": 2.0638,
+      "step": 37540
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0160294333234288e-05,
+      "loss": 2.1164,
+      "step": 37545
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0142544891329975e-05,
+      "loss": 2.0502,
+      "step": 37550
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0124810138331907e-05,
+      "loss": 2.0837,
+      "step": 37555
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0107090077139192e-05,
+      "loss": 2.1013,
+      "step": 37560
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.0089384710648486e-05,
+      "loss": 2.1007,
+      "step": 37565
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0071694041754087e-05,
+      "loss": 2.1291,
+      "step": 37570
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0054018073347849e-05,
+      "loss": 2.0953,
+      "step": 37575
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.003635680831927e-05,
+      "loss": 2.1239,
+      "step": 37580
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.0018710249555397e-05,
+      "loss": 2.1552,
+      "step": 37585
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0001078399940922e-05,
+      "loss": 2.1038,
+      "step": 37590
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.983461262358063e-06,
+      "loss": 2.1258,
+      "step": 37595
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.965858839686714e-06,
+      "loss": 2.1501,
+      "step": 37600
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.948271134804299e-06,
+      "loss": 2.118,
+      "step": 37605
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.930698150585904e-06,
+      "loss": 2.0905,
+      "step": 37610
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.9131398899041e-06,
+      "loss": 2.0818,
+      "step": 37615
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.895596355629188e-06,
+      "loss": 2.0497,
+      "step": 37620
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.878067550628922e-06,
+      "loss": 2.1251,
+      "step": 37625
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.860553477768786e-06,
+      "loss": 2.0617,
+      "step": 37630
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.843054139911755e-06,
+      "loss": 2.1423,
+      "step": 37635
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.825569539918433e-06,
+      "loss": 2.108,
+      "step": 37640
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.808099680646976e-06,
+      "loss": 2.1287,
+      "step": 37645
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.7906445649532e-06,
+      "loss": 2.0799,
+      "step": 37650
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.773204195690489e-06,
+      "loss": 2.0501,
+      "step": 37655
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.755778575709752e-06,
+      "loss": 2.0785,
+      "step": 37660
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.738367707859597e-06,
+      "loss": 2.0956,
+      "step": 37665
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.720971594986106e-06,
+      "loss": 2.1278,
+      "step": 37670
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.703590239933047e-06,
+      "loss": 2.092,
+      "step": 37675
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.686223645541704e-06,
+      "loss": 2.083,
+      "step": 37680
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.668871814650993e-06,
+      "loss": 2.1038,
+      "step": 37685
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.651534750097379e-06,
+      "loss": 2.095,
+      "step": 37690
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.634212454714975e-06,
+      "loss": 2.0919,
+      "step": 37695
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.61690493133538e-06,
+      "loss": 2.1113,
+      "step": 37700
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.599612182787899e-06,
+      "loss": 2.1125,
+      "step": 37705
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.582334211899303e-06,
+      "loss": 2.1083,
+      "step": 37710
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.565071021494054e-06,
+      "loss": 2.1431,
+      "step": 37715
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.547822614394109e-06,
+      "loss": 2.1287,
+      "step": 37720
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.530588993419076e-06,
+      "loss": 2.108,
+      "step": 37725
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.513370161386093e-06,
+      "loss": 2.0868,
+      "step": 37730
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.496166121109939e-06,
+      "loss": 2.0574,
+      "step": 37735
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.4789768754029e-06,
+      "loss": 2.1209,
+      "step": 37740
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.461802427074917e-06,
+      "loss": 2.1025,
+      "step": 37745
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.444642778933444e-06,
+      "loss": 2.1484,
+      "step": 37750
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.427497933783603e-06,
+      "loss": 2.1551,
+      "step": 37755
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.410367894427985e-06,
+      "loss": 2.1268,
+      "step": 37760
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.393252663666874e-06,
+      "loss": 2.0937,
+      "step": 37765
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.376152244298041e-06,
+      "loss": 2.0697,
+      "step": 37770
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.359066639116898e-06,
+      "loss": 2.0637,
+      "step": 37775
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.341995850916386e-06,
+      "loss": 2.1107,
+      "step": 37780
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.324939882487083e-06,
+      "loss": 2.0499,
+      "step": 37785
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.307898736617082e-06,
+      "loss": 2.1245,
+      "step": 37790
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.2908724160921e-06,
+      "loss": 2.0916,
+      "step": 37795
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.273860923695387e-06,
+      "loss": 2.0763,
+      "step": 37800
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.256864262207832e-06,
+      "loss": 2.1127,
+      "step": 37805
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.23988243440782e-06,
+      "loss": 2.1346,
+      "step": 37810
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.222915443071412e-06,
+      "loss": 2.1135,
+      "step": 37815
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.20596329097212e-06,
+      "loss": 2.1657,
+      "step": 37820
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.189025980881127e-06,
+      "loss": 2.0825,
+      "step": 37825
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.172103515567144e-06,
+      "loss": 2.0832,
+      "step": 37830
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.155195897796487e-06,
+      "loss": 2.1049,
+      "step": 37835
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.138303130333004e-06,
+      "loss": 2.1261,
+      "step": 37840
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.121425215938174e-06,
+      "loss": 2.105,
+      "step": 37845
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.104562157370966e-06,
+      "loss": 2.1181,
+      "step": 37850
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.087713957388011e-06,
+      "loss": 2.0384,
+      "step": 37855
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.070880618743438e-06,
+      "loss": 2.103,
+      "step": 37860
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.05406214418899e-06,
+      "loss": 2.0871,
+      "step": 37865
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.037258536473947e-06,
+      "loss": 2.0953,
+      "step": 37870
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.020469798345222e-06,
+      "loss": 2.1083,
+      "step": 37875
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.003695932547196e-06,
+      "loss": 2.1418,
+      "step": 37880
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.986936941821911e-06,
+      "loss": 2.0939,
+      "step": 37885
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.97019282890893e-06,
+      "loss": 2.0983,
+      "step": 37890
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8.953463596545408e-06,
+      "loss": 2.0825,
+      "step": 37895
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.936749247466025e-06,
+      "loss": 2.0792,
+      "step": 37900
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.920049784403106e-06,
+      "loss": 2.1414,
+      "step": 37905
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.903365210086456e-06,
+      "loss": 2.107,
+      "step": 37910
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.886695527243505e-06,
+      "loss": 2.1018,
+      "step": 37915
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.870040738599217e-06,
+      "loss": 2.1068,
+      "step": 37920
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.853400846876147e-06,
+      "loss": 2.0802,
+      "step": 37925
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.836775854794378e-06,
+      "loss": 2.0655,
+      "step": 37930
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.820165765071608e-06,
+      "loss": 2.0459,
+      "step": 37935
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.803570580423048e-06,
+      "loss": 2.1002,
+      "step": 37940
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.786990303561526e-06,
+      "loss": 2.1262,
+      "step": 37945
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.770424937197364e-06,
+      "loss": 2.1494,
+      "step": 37950
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.753874484038526e-06,
+      "loss": 2.1398,
+      "step": 37955
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.737338946790453e-06,
+      "loss": 2.1198,
+      "step": 37960
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.720818328156233e-06,
+      "loss": 2.0625,
+      "step": 37965
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.704312630836443e-06,
+      "loss": 2.1101,
+      "step": 37970
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.687821857529287e-06,
+      "loss": 2.0639,
+      "step": 37975
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.671346010930459e-06,
+      "loss": 2.1333,
+      "step": 37980
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.654885093733255e-06,
+      "loss": 2.0847,
+      "step": 37985
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.63843910862856e-06,
+      "loss": 2.1558,
+      "step": 37990
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.622008058304754e-06,
+      "loss": 2.0944,
+      "step": 37995
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.605591945447811e-06,
+      "loss": 2.1405,
+      "step": 38000
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.589190772741229e-06,
+      "loss": 2.0528,
+      "step": 38005
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.572804542866131e-06,
+      "loss": 2.0573,
+      "step": 38010
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.556433258501128e-06,
+      "loss": 2.1535,
+      "step": 38015
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.540076922322437e-06,
+      "loss": 2.0956,
+      "step": 38020
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.523735537003796e-06,
+      "loss": 2.0512,
+      "step": 38025
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.507409105216534e-06,
+      "loss": 2.0848,
+      "step": 38030
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.491097629629486e-06,
+      "loss": 2.0492,
+      "step": 38035
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.474801112909104e-06,
+      "loss": 2.0183,
+      "step": 38040
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.458519557719336e-06,
+      "loss": 2.1081,
+      "step": 38045
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.442252966721743e-06,
+      "loss": 2.1117,
+      "step": 38050
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.426001342575374e-06,
+      "loss": 2.1464,
+      "step": 38055
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.409764687936894e-06,
+      "loss": 2.0336,
+      "step": 38060
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.393543005460469e-06,
+      "loss": 2.0838,
+      "step": 38065
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.377336297797855e-06,
+      "loss": 2.0768,
+      "step": 38070
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.361144567598334e-06,
+      "loss": 2.152,
+      "step": 38075
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.344967817508776e-06,
+      "loss": 2.0812,
+      "step": 38080
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.328806050173543e-06,
+      "loss": 2.1194,
+      "step": 38085
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.312659268234613e-06,
+      "loss": 2.0593,
+      "step": 38090
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.29652747433146e-06,
+      "loss": 2.0828,
+      "step": 38095
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.280410671101169e-06,
+      "loss": 2.1263,
+      "step": 38100
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.264308861178282e-06,
+      "loss": 2.1093,
+      "step": 38105
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.248222047194997e-06,
+      "loss": 2.0713,
+      "step": 38110
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.232150231780966e-06,
+      "loss": 2.1317,
+      "step": 38115
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.216093417563475e-06,
+      "loss": 2.1513,
+      "step": 38120
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.200051607167281e-06,
+      "loss": 2.111,
+      "step": 38125
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.18402480321474e-06,
+      "loss": 2.1364,
+      "step": 38130
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.168013008325725e-06,
+      "loss": 2.0714,
+      "step": 38135
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.152016225117688e-06,
+      "loss": 2.1396,
+      "step": 38140
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.136034456205577e-06,
+      "loss": 2.1131,
+      "step": 38145
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.12006770420195e-06,
+      "loss": 2.1632,
+      "step": 38150
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.104115971716841e-06,
+      "loss": 2.1251,
+      "step": 38155
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.088179261357898e-06,
+      "loss": 2.115,
+      "step": 38160
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.072257575730235e-06,
+      "loss": 2.0615,
+      "step": 38165
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.056350917436606e-06,
+      "loss": 2.1366,
+      "step": 38170
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.040459289077207e-06,
+      "loss": 2.08,
+      "step": 38175
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.02458269324987e-06,
+      "loss": 2.059,
+      "step": 38180
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.008721132549901e-06,
+      "loss": 2.1523,
+      "step": 38185
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.992874609570167e-06,
+      "loss": 2.1055,
+      "step": 38190
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.977043126901107e-06,
+      "loss": 2.0855,
+      "step": 38195
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.96122668713064e-06,
+      "loss": 2.0637,
+      "step": 38200
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.945425292844311e-06,
+      "loss": 2.1453,
+      "step": 38205
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.929638946625106e-06,
+      "loss": 2.0756,
+      "step": 38210
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.913867651053652e-06,
+      "loss": 2.0993,
+      "step": 38215
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.89811140870803e-06,
+      "loss": 2.0754,
+      "step": 38220
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.882370222163927e-06,
+      "loss": 2.1042,
+      "step": 38225
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.866644093994501e-06,
+      "loss": 2.1017,
+      "step": 38230
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.850933026770536e-06,
+      "loss": 2.1288,
+      "step": 38235
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.835237023060249e-06,
+      "loss": 2.104,
+      "step": 38240
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.819556085429492e-06,
+      "loss": 2.109,
+      "step": 38245
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.80389021644159e-06,
+      "loss": 2.1156,
+      "step": 38250
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.78823941865744e-06,
+      "loss": 2.0961,
+      "step": 38255
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.77260369463545e-06,
+      "loss": 2.0985,
+      "step": 38260
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.756983046931587e-06,
+      "loss": 2.157,
+      "step": 38265
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.741377478099321e-06,
+      "loss": 2.1452,
+      "step": 38270
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.72578699068971e-06,
+      "loss": 2.0955,
+      "step": 38275
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.710211587251281e-06,
+      "loss": 2.136,
+      "step": 38280
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.694651270330167e-06,
+      "loss": 2.14,
+      "step": 38285
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.679106042469963e-06,
+      "loss": 2.095,
+      "step": 38290
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.3671875,
+      "learning_rate": 7.663575906211862e-06,
+      "loss": 2.101,
+      "step": 38295
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.64806086409452e-06,
+      "loss": 2.0633,
+      "step": 38300
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.632560918654209e-06,
+      "loss": 2.1647,
+      "step": 38305
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.617076072424645e-06,
+      "loss": 2.1054,
+      "step": 38310
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.601606327937161e-06,
+      "loss": 2.16,
+      "step": 38315
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.586151687720555e-06,
+      "loss": 2.0816,
+      "step": 38320
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.5707121543011965e-06,
+      "loss": 2.1449,
+      "step": 38325
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.555287730202954e-06,
+      "loss": 2.0819,
+      "step": 38330
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.539878417947255e-06,
+      "loss": 2.1091,
+      "step": 38335
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.52448422005303e-06,
+      "loss": 2.1407,
+      "step": 38340
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.509105139036787e-06,
+      "loss": 2.0974,
+      "step": 38345
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.4937411774124725e-06,
+      "loss": 2.1369,
+      "step": 38350
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.478392337691675e-06,
+      "loss": 2.1308,
+      "step": 38355
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.463058622383423e-06,
+      "loss": 2.1273,
+      "step": 38360
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.4477400339943215e-06,
+      "loss": 2.1108,
+      "step": 38365
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.4324365750284345e-06,
+      "loss": 2.1231,
+      "step": 38370
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.365234375,
+      "learning_rate": 7.417148247987471e-06,
+      "loss": 2.093,
+      "step": 38375
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.401875055370544e-06,
+      "loss": 2.0988,
+      "step": 38380
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.386616999674389e-06,
+      "loss": 2.1231,
+      "step": 38385
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.371374083393179e-06,
+      "loss": 2.0941,
+      "step": 38390
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.3561463090187075e-06,
+      "loss": 2.0794,
+      "step": 38395
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.340933679040207e-06,
+      "loss": 2.0741,
+      "step": 38400
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.3257361959444774e-06,
+      "loss": 2.115,
+      "step": 38405
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.3105538622158635e-06,
+      "loss": 2.1346,
+      "step": 38410
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.295386680336158e-06,
+      "loss": 2.0543,
+      "step": 38415
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.280234652784779e-06,
+      "loss": 2.0998,
+      "step": 38420
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.265097782038554e-06,
+      "loss": 2.0857,
+      "step": 38425
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.249976070571951e-06,
+      "loss": 2.1094,
+      "step": 38430
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.234869520856846e-06,
+      "loss": 2.1336,
+      "step": 38435
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.219778135362743e-06,
+      "loss": 2.1089,
+      "step": 38440
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.204701916556566e-06,
+      "loss": 2.0988,
+      "step": 38445
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.189640866902847e-06,
+      "loss": 2.1618,
+      "step": 38450
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.17459498886357e-06,
+      "loss": 2.0992,
+      "step": 38455
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 7.159564284898313e-06,
+      "loss": 2.1298,
+      "step": 38460
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.1445487574640755e-06,
+      "loss": 2.1112,
+      "step": 38465
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.129548409015474e-06,
+      "loss": 2.138,
+      "step": 38470
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.114563242004568e-06,
+      "loss": 2.0544,
+      "step": 38475
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.099593258880999e-06,
+      "loss": 2.1369,
+      "step": 38480
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.084638462091864e-06,
+      "loss": 2.0749,
+      "step": 38485
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.06969885408183e-06,
+      "loss": 2.1047,
+      "step": 38490
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.054774437293044e-06,
+      "loss": 2.0734,
+      "step": 38495
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.039865214165209e-06,
+      "loss": 2.0761,
+      "step": 38500
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.0249711871354874e-06,
+      "loss": 2.0977,
+      "step": 38505
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.01009235863862e-06,
+      "loss": 2.0944,
+      "step": 38510
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.995228731106806e-06,
+      "loss": 2.0643,
+      "step": 38515
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.980380306969825e-06,
+      "loss": 2.0713,
+      "step": 38520
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.965547088654889e-06,
+      "loss": 2.121,
+      "step": 38525
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.950729078586815e-06,
+      "loss": 2.1221,
+      "step": 38530
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.935926279187843e-06,
+      "loss": 2.116,
+      "step": 38535
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.921138692877827e-06,
+      "loss": 2.111,
+      "step": 38540
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.90636632207401e-06,
+      "loss": 2.0729,
+      "step": 38545
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.891609169191271e-06,
+      "loss": 2.1228,
+      "step": 38550
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.876867236641904e-06,
+      "loss": 2.0563,
+      "step": 38555
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.862140526835803e-06,
+      "loss": 2.1449,
+      "step": 38560
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.847429042180287e-06,
+      "loss": 2.0903,
+      "step": 38565
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.8327327850802535e-06,
+      "loss": 2.1341,
+      "step": 38570
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.8180517579380706e-06,
+      "loss": 2.0992,
+      "step": 38575
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.803385963153653e-06,
+      "loss": 2.1252,
+      "step": 38580
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.788735403124369e-06,
+      "loss": 2.1428,
+      "step": 38585
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.7741000802451715e-06,
+      "loss": 2.1071,
+      "step": 38590
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.759479996908435e-06,
+      "loss": 2.1175,
+      "step": 38595
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.744875155504127e-06,
+      "loss": 2.1556,
+      "step": 38600
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.7302855584196685e-06,
+      "loss": 2.1092,
+      "step": 38605
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.7157112080400095e-06,
+      "loss": 2.1348,
+      "step": 38610
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.701152106747599e-06,
+      "loss": 2.0994,
+      "step": 38615
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.686608256922422e-06,
+      "loss": 2.0826,
+      "step": 38620
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.67207966094191e-06,
+      "loss": 2.0967,
+      "step": 38625
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.657566321181075e-06,
+      "loss": 2.1206,
+      "step": 38630
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.643068240012362e-06,
+      "loss": 2.0453,
+      "step": 38635
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.6285854198057885e-06,
+      "loss": 2.054,
+      "step": 38640
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.6141178629288255e-06,
+      "loss": 2.1089,
+      "step": 38645
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.599665571746483e-06,
+      "loss": 2.0641,
+      "step": 38650
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.585228548621247e-06,
+      "loss": 2.1117,
+      "step": 38655
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.570806795913142e-06,
+      "loss": 2.1135,
+      "step": 38660
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.5564003159796605e-06,
+      "loss": 2.0705,
+      "step": 38665
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.542009111175829e-06,
+      "loss": 2.0868,
+      "step": 38670
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.527633183854143e-06,
+      "loss": 2.0508,
+      "step": 38675
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.5132725363646454e-06,
+      "loss": 2.09,
+      "step": 38680
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.498927171054836e-06,
+      "loss": 2.0924,
+      "step": 38685
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.484597090269761e-06,
+      "loss": 2.1358,
+      "step": 38690
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.470282296351915e-06,
+      "loss": 2.0535,
+      "step": 38695
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.455982791641346e-06,
+      "loss": 2.0596,
+      "step": 38700
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.4416985784755525e-06,
+      "loss": 2.0944,
+      "step": 38705
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.427429659189599e-06,
+      "loss": 2.1158,
+      "step": 38710
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.4131760361159645e-06,
+      "loss": 2.0839,
+      "step": 38715
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.39893771158473e-06,
+      "loss": 2.1068,
+      "step": 38720
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.3847146879233765e-06,
+      "loss": 2.0975,
+      "step": 38725
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.370506967456935e-06,
+      "loss": 2.1193,
+      "step": 38730
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.356314552507936e-06,
+      "loss": 2.098,
+      "step": 38735
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.34213744539639e-06,
+      "loss": 2.1522,
+      "step": 38740
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.327975648439832e-06,
+      "loss": 2.0844,
+      "step": 38745
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.3138291639532446e-06,
+      "loss": 2.1293,
+      "step": 38750
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.299697994249176e-06,
+      "loss": 2.1045,
+      "step": 38755
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.285582141637591e-06,
+      "loss": 2.0964,
+      "step": 38760
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.271481608426044e-06,
+      "loss": 2.0747,
+      "step": 38765
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.257396396919501e-06,
+      "loss": 2.148,
+      "step": 38770
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.243326509420467e-06,
+      "loss": 2.0892,
+      "step": 38775
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.229271948228921e-06,
+      "loss": 2.1173,
+      "step": 38780
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.2152327156423716e-06,
+      "loss": 2.077,
+      "step": 38785
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.201208813955772e-06,
+      "loss": 2.1096,
+      "step": 38790
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.187200245461622e-06,
+      "loss": 2.1034,
+      "step": 38795
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.173207012449855e-06,
+      "loss": 2.1112,
+      "step": 38800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.159229117207965e-06,
+      "loss": 2.2008,
+      "step": 38805
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.145266562020868e-06,
+      "loss": 2.131,
+      "step": 38810
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.13131934917105e-06,
+      "loss": 2.1402,
+      "step": 38815
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.117387480938408e-06,
+      "loss": 2.0662,
+      "step": 38820
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.103470959600399e-06,
+      "loss": 2.1359,
+      "step": 38825
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.089569787431926e-06,
+      "loss": 2.1371,
+      "step": 38830
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.075683966705415e-06,
+      "loss": 2.1452,
+      "step": 38835
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.0618134996907385e-06,
+      "loss": 2.0767,
+      "step": 38840
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.047958388655317e-06,
+      "loss": 2.0288,
+      "step": 38845
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.034118635864017e-06,
+      "loss": 2.0939,
+      "step": 38850
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.0202942435792165e-06,
+      "loss": 2.1432,
+      "step": 38855
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.006485214060753e-06,
+      "loss": 2.1391,
+      "step": 38860
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.99269154956602e-06,
+      "loss": 2.0961,
+      "step": 38865
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.978913252349805e-06,
+      "loss": 2.1214,
+      "step": 38870
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.965150324664459e-06,
+      "loss": 2.1006,
+      "step": 38875
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.951402768759784e-06,
+      "loss": 2.1028,
+      "step": 38880
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.9376705868830816e-06,
+      "loss": 2.1236,
+      "step": 38885
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.923953781279135e-06,
+      "loss": 2.0988,
+      "step": 38890
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.910252354190237e-06,
+      "loss": 2.1119,
+      "step": 38895
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.8965663078561e-06,
+      "loss": 2.0584,
+      "step": 38900
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.882895644514019e-06,
+      "loss": 2.1367,
+      "step": 38905
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.869240366398709e-06,
+      "loss": 2.0924,
+      "step": 38910
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.855600475742351e-06,
+      "loss": 2.1231,
+      "step": 38915
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.841975974774682e-06,
+      "loss": 2.1042,
+      "step": 38920
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.828366865722868e-06,
+      "loss": 2.0445,
+      "step": 38925
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.814773150811581e-06,
+      "loss": 2.1124,
+      "step": 38930
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.8011948322629685e-06,
+      "loss": 2.0886,
+      "step": 38935
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.787631912296665e-06,
+      "loss": 2.1228,
+      "step": 38940
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.774084393129786e-06,
+      "loss": 2.0713,
+      "step": 38945
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.76055227697695e-06,
+      "loss": 2.0785,
+      "step": 38950
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.747035566050208e-06,
+      "loss": 2.1029,
+      "step": 38955
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.733534262559148e-06,
+      "loss": 2.101,
+      "step": 38960
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.720048368710784e-06,
+      "loss": 2.1309,
+      "step": 38965
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.706577886709674e-06,
+      "loss": 2.128,
+      "step": 38970
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.6931228187578016e-06,
+      "loss": 2.1276,
+      "step": 38975
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.679683167054672e-06,
+      "loss": 2.0714,
+      "step": 38980
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.66625893379723e-06,
+      "loss": 2.1045,
+      "step": 38985
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.652850121179942e-06,
+      "loss": 2.1182,
+      "step": 38990
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.63945673139471e-06,
+      "loss": 2.1592,
+      "step": 38995
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.6260787666309714e-06,
+      "loss": 2.1806,
+      "step": 39000
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.6127162290755655e-06,
+      "loss": 2.1409,
+      "step": 39005
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.59936912091289e-06,
+      "loss": 2.0718,
+      "step": 39010
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.586037444324754e-06,
+      "loss": 2.1277,
+      "step": 39015
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.572721201490505e-06,
+      "loss": 2.1198,
+      "step": 39020
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.5594203945869e-06,
+      "loss": 2.1269,
+      "step": 39025
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.546135025788235e-06,
+      "loss": 2.0813,
+      "step": 39030
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.532865097266238e-06,
+      "loss": 2.1088,
+      "step": 39035
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.519610611190151e-06,
+      "loss": 2.0725,
+      "step": 39040
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.5063715697266425e-06,
+      "loss": 2.1561,
+      "step": 39045
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.493147975039925e-06,
+      "loss": 2.1635,
+      "step": 39050
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.479939829291602e-06,
+      "loss": 2.164,
+      "step": 39055
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.466747134640837e-06,
+      "loss": 2.1007,
+      "step": 39060
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.4535698932441815e-06,
+      "loss": 2.0573,
+      "step": 39065
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.440408107255746e-06,
+      "loss": 2.068,
+      "step": 39070
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.427261778827053e-06,
+      "loss": 2.0851,
+      "step": 39075
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.414130910107129e-06,
+      "loss": 2.1132,
+      "step": 39080
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.401015503242468e-06,
+      "loss": 2.1059,
+      "step": 39085
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.38791556037701e-06,
+      "loss": 2.0939,
+      "step": 39090
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.374831083652198e-06,
+      "loss": 2.1131,
+      "step": 39095
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.361762075206955e-06,
+      "loss": 2.0774,
+      "step": 39100
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.348708537177627e-06,
+      "loss": 2.1142,
+      "step": 39105
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.3356704716981065e-06,
+      "loss": 2.1237,
+      "step": 39110
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.322647880899667e-06,
+      "loss": 2.0999,
+      "step": 39115
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.309640766911128e-06,
+      "loss": 2.0996,
+      "step": 39120
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.296649131858733e-06,
+      "loss": 2.1546,
+      "step": 39125
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.283672977866227e-06,
+      "loss": 2.1296,
+      "step": 39130
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.270712307054793e-06,
+      "loss": 2.0271,
+      "step": 39135
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.257767121543111e-06,
+      "loss": 2.056,
+      "step": 39140
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.244837423447302e-06,
+      "loss": 2.1371,
+      "step": 39145
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.231923214880996e-06,
+      "loss": 2.0964,
+      "step": 39150
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.219024497955239e-06,
+      "loss": 2.1136,
+      "step": 39155
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.206141274778586e-06,
+      "loss": 2.1644,
+      "step": 39160
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.193273547457034e-06,
+      "loss": 2.1183,
+      "step": 39165
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.180421318094053e-06,
+      "loss": 2.0503,
+      "step": 39170
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.1675845887906196e-06,
+      "loss": 2.0719,
+      "step": 39175
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.154763361645087e-06,
+      "loss": 2.0754,
+      "step": 39180
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.141957638753381e-06,
+      "loss": 2.1233,
+      "step": 39185
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.129167422208803e-06,
+      "loss": 2.1114,
+      "step": 39190
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.1163927141021804e-06,
+      "loss": 2.1103,
+      "step": 39195
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.103633516521755e-06,
+      "loss": 2.148,
+      "step": 39200
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.090889831553292e-06,
+      "loss": 2.1764,
+      "step": 39205
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.078161661279946e-06,
+      "loss": 2.1464,
+      "step": 39210
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.06544900778243e-06,
+      "loss": 2.1047,
+      "step": 39215
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.052751873138828e-06,
+      "loss": 2.1092,
+      "step": 39220
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.040070259424745e-06,
+      "loss": 2.0935,
+      "step": 39225
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.027404168713223e-06,
+      "loss": 2.1015,
+      "step": 39230
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.014753603074796e-06,
+      "loss": 2.0454,
+      "step": 39235
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.002118564577396e-06,
+      "loss": 2.1675,
+      "step": 39240
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.989499055286517e-06,
+      "loss": 2.1278,
+      "step": 39245
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.976895077264998e-06,
+      "loss": 2.137,
+      "step": 39250
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.964306632573257e-06,
+      "loss": 2.0694,
+      "step": 39255
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.951733723269058e-06,
+      "loss": 2.1046,
+      "step": 39260
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.9391763514077375e-06,
+      "loss": 2.0706,
+      "step": 39265
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.926634519041984e-06,
+      "loss": 2.1424,
+      "step": 39270
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.914108228222036e-06,
+      "loss": 2.0755,
+      "step": 39275
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.901597480995513e-06,
+      "loss": 2.0926,
+      "step": 39280
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.889102279407587e-06,
+      "loss": 2.1466,
+      "step": 39285
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.876622625500782e-06,
+      "loss": 2.0739,
+      "step": 39290
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.8641585213151875e-06,
+      "loss": 2.1177,
+      "step": 39295
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.8517099688882515e-06,
+      "loss": 2.0624,
+      "step": 39300
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.839276970254958e-06,
+      "loss": 2.0864,
+      "step": 39305
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.82685952744768e-06,
+      "loss": 2.1191,
+      "step": 39310
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.81445764249634e-06,
+      "loss": 2.1044,
+      "step": 39315
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.8020713174282024e-06,
+      "loss": 2.1168,
+      "step": 39320
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.789700554268095e-06,
+      "loss": 2.0977,
+      "step": 39325
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.777345355038221e-06,
+      "loss": 2.1062,
+      "step": 39330
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.765005721758298e-06,
+      "loss": 2.0609,
+      "step": 39335
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.752681656445435e-06,
+      "loss": 2.0796,
+      "step": 39340
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.740373161114287e-06,
+      "loss": 2.0867,
+      "step": 39345
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.728080237776855e-06,
+      "loss": 2.1231,
+      "step": 39350
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.715802888442699e-06,
+      "loss": 2.1933,
+      "step": 39355
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.703541115118759e-06,
+      "loss": 2.0374,
+      "step": 39360
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.691294919809464e-06,
+      "loss": 2.0676,
+      "step": 39365
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.679064304516667e-06,
+      "loss": 2.1021,
+      "step": 39370
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.666849271239737e-06,
+      "loss": 2.0846,
+      "step": 39375
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.654649821975399e-06,
+      "loss": 2.1169,
+      "step": 39380
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.6424659587179345e-06,
+      "loss": 2.1166,
+      "step": 39385
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.630297683458984e-06,
+      "loss": 2.1454,
+      "step": 39390
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.6181449981877125e-06,
+      "loss": 2.163,
+      "step": 39395
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.606007904890697e-06,
+      "loss": 2.1613,
+      "step": 39400
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.593886405551983e-06,
+      "loss": 2.1077,
+      "step": 39405
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.581780502153043e-06,
+      "loss": 2.1253,
+      "step": 39410
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.569690196672838e-06,
+      "loss": 2.0772,
+      "step": 39415
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.55761549108773e-06,
+      "loss": 2.1153,
+      "step": 39420
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.545556387371585e-06,
+      "loss": 2.0781,
+      "step": 39425
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.533512887495683e-06,
+      "loss": 2.0496,
+      "step": 39430
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.5214849934287705e-06,
+      "loss": 2.1181,
+      "step": 39435
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.509472707137008e-06,
+      "loss": 2.137,
+      "step": 39440
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.497476030584069e-06,
+      "loss": 2.1759,
+      "step": 39445
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.485494965731018e-06,
+      "loss": 2.0739,
+      "step": 39450
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.473529514536367e-06,
+      "loss": 2.1093,
+      "step": 39455
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.46157967895614e-06,
+      "loss": 2.1727,
+      "step": 39460
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.44964546094373e-06,
+      "loss": 2.0826,
+      "step": 39465
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.437726862450042e-06,
+      "loss": 2.0892,
+      "step": 39470
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.425823885423352e-06,
+      "loss": 2.0473,
+      "step": 39475
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.413936531809481e-06,
+      "loss": 2.1527,
+      "step": 39480
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.402064803551597e-06,
+      "loss": 2.1044,
+      "step": 39485
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.3902087025903815e-06,
+      "loss": 2.0722,
+      "step": 39490
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.3783682308639385e-06,
+      "loss": 2.1,
+      "step": 39495
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.36654339030782e-06,
+      "loss": 2.0687,
+      "step": 39500
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.354734182855002e-06,
+      "loss": 2.1425,
+      "step": 39505
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.342940610435953e-06,
+      "loss": 2.1243,
+      "step": 39510
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.331162674978528e-06,
+      "loss": 2.048,
+      "step": 39515
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.3194003784080784e-06,
+      "loss": 2.0655,
+      "step": 39520
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.307653722647353e-06,
+      "loss": 2.07,
+      "step": 39525
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.295922709616584e-06,
+      "loss": 2.1216,
+      "step": 39530
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.284207341233404e-06,
+      "loss": 2.0825,
+      "step": 39535
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.272507619412946e-06,
+      "loss": 2.0453,
+      "step": 39540
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.260823546067727e-06,
+      "loss": 2.0914,
+      "step": 39545
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.24915512310774e-06,
+      "loss": 2.1365,
+      "step": 39550
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.2375023524403925e-06,
+      "loss": 2.1419,
+      "step": 39555
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.225865235970583e-06,
+      "loss": 2.0742,
+      "step": 39560
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.214243775600591e-06,
+      "loss": 2.1959,
+      "step": 39565
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.2026379732301835e-06,
+      "loss": 2.1294,
+      "step": 39570
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.191047830756534e-06,
+      "loss": 2.0523,
+      "step": 39575
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.179473350074281e-06,
+      "loss": 2.1512,
+      "step": 39580
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.16791453307549e-06,
+      "loss": 2.0366,
+      "step": 39585
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.156371381649671e-06,
+      "loss": 2.1093,
+      "step": 39590
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.14484389768377e-06,
+      "loss": 2.0826,
+      "step": 39595
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.133332083062169e-06,
+      "loss": 2.1425,
+      "step": 39600
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.121835939666696e-06,
+      "loss": 2.0887,
+      "step": 39605
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.110355469376625e-06,
+      "loss": 2.14,
+      "step": 39610
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.098890674068623e-06,
+      "loss": 2.1293,
+      "step": 39615
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.087441555616878e-06,
+      "loss": 2.1067,
+      "step": 39620
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.076008115892915e-06,
+      "loss": 2.1122,
+      "step": 39625
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.064590356765785e-06,
+      "loss": 2.113,
+      "step": 39630
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.053188280101927e-06,
+      "loss": 2.0853,
+      "step": 39635
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.041801887765207e-06,
+      "loss": 2.1231,
+      "step": 39640
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.030431181616967e-06,
+      "loss": 2.1533,
+      "step": 39645
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.019076163515956e-06,
+      "loss": 2.1369,
+      "step": 39650
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.0077368353183875e-06,
+      "loss": 2.0521,
+      "step": 39655
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.9964131988778575e-06,
+      "loss": 2.0556,
+      "step": 39660
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.9851052560454516e-06,
+      "loss": 2.1186,
+      "step": 39665
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.973813008669658e-06,
+      "loss": 2.1468,
+      "step": 39670
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.962536458596422e-06,
+      "loss": 2.1418,
+      "step": 39675
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.9512756076690915e-06,
+      "loss": 2.0987,
+      "step": 39680
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.940030457728483e-06,
+      "loss": 2.0777,
+      "step": 39685
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.928801010612803e-06,
+      "loss": 2.1321,
+      "step": 39690
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.917587268157752e-06,
+      "loss": 2.1581,
+      "step": 39695
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.906389232196406e-06,
+      "loss": 2.1172,
+      "step": 39700
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.895206904559312e-06,
+      "loss": 2.1238,
+      "step": 39705
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.88404028707442e-06,
+      "loss": 2.0712,
+      "step": 39710
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.872889381567135e-06,
+      "loss": 2.1214,
+      "step": 39715
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.861754189860267e-06,
+      "loss": 2.1295,
+      "step": 39720
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.850634713774104e-06,
+      "loss": 2.1114,
+      "step": 39725
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.839530955126303e-06,
+      "loss": 2.1057,
+      "step": 39730
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.828442915732022e-06,
+      "loss": 2.1019,
+      "step": 39735
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.817370597403769e-06,
+      "loss": 2.0635,
+      "step": 39740
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.8063140019515498e-06,
+      "loss": 2.0941,
+      "step": 39745
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.795273131182764e-06,
+      "loss": 2.0762,
+      "step": 39750
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.7842479869022785e-06,
+      "loss": 2.076,
+      "step": 39755
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.7732385709123185e-06,
+      "loss": 2.1199,
+      "step": 39760
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.7622448850126223e-06,
+      "loss": 2.1094,
+      "step": 39765
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.7512669310002744e-06,
+      "loss": 2.0437,
+      "step": 39770
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.740304710669884e-06,
+      "loss": 2.1233,
+      "step": 39775
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.729358225813384e-06,
+      "loss": 2.1384,
+      "step": 39780
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.71842747822021e-06,
+      "loss": 2.0935,
+      "step": 39785
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.707512469677188e-06,
+      "loss": 2.0082,
+      "step": 39790
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.696613201968613e-06,
+      "loss": 2.0366,
+      "step": 39795
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.685729676876126e-06,
+      "loss": 2.1042,
+      "step": 39800
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.674861896178894e-06,
+      "loss": 2.1227,
+      "step": 39805
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.66400986165345e-06,
+      "loss": 2.0984,
+      "step": 39810
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.6531735750737427e-06,
+      "loss": 2.1553,
+      "step": 39815
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.642353038211188e-06,
+      "loss": 2.1273,
+      "step": 39820
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.6315482528346044e-06,
+      "loss": 2.1034,
+      "step": 39825
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.6207592207102236e-06,
+      "loss": 2.0568,
+      "step": 39830
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.609985943601757e-06,
+      "loss": 2.0722,
+      "step": 39835
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.599228423270251e-06,
+      "loss": 2.1127,
+      "step": 39840
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.5884866614742663e-06,
+      "loss": 2.1221,
+      "step": 39845
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.577760659969731e-06,
+      "loss": 2.1027,
+      "step": 39850
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.567050420510021e-06,
+      "loss": 2.0919,
+      "step": 39855
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.5563559448459125e-06,
+      "loss": 2.0829,
+      "step": 39860
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.5456772347256417e-06,
+      "loss": 2.093,
+      "step": 39865
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.5350142918948224e-06,
+      "loss": 2.0766,
+      "step": 39870
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.524367118096539e-06,
+      "loss": 2.0823,
+      "step": 39875
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.5137357150712537e-06,
+      "loss": 2.1271,
+      "step": 39880
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.503120084556899e-06,
+      "loss": 2.136,
+      "step": 39885
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.4925202282887647e-06,
+      "loss": 2.1396,
+      "step": 39890
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.4819361479996204e-06,
+      "loss": 2.1417,
+      "step": 39895
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.4713678454196375e-06,
+      "loss": 2.1076,
+      "step": 39900
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.4608153222764007e-06,
+      "loss": 2.0626,
+      "step": 39905
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.4502785802949077e-06,
+      "loss": 2.0553,
+      "step": 39910
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.4397576211976035e-06,
+      "loss": 2.1067,
+      "step": 39915
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.429252446704334e-06,
+      "loss": 2.1444,
+      "step": 39920
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.4187630585323706e-06,
+      "loss": 2.1252,
+      "step": 39925
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.4082894583964188e-06,
+      "loss": 2.1732,
+      "step": 39930
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.3978316480085427e-06,
+      "loss": 2.0991,
+      "step": 39935
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.38738962907833e-06,
+      "loss": 2.071,
+      "step": 39940
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.376963403312672e-06,
+      "loss": 2.0427,
+      "step": 39945
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.3665529724159704e-06,
+      "loss": 2.0627,
+      "step": 39950
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.3561583380899765e-06,
+      "loss": 2.1059,
+      "step": 39955
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.3457795020339188e-06,
+      "loss": 2.1264,
+      "step": 39960
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.3354164659444076e-06,
+      "loss": 2.1432,
+      "step": 39965
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.3250692315154653e-06,
+      "loss": 2.0812,
+      "step": 39970
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.314737800438561e-06,
+      "loss": 2.089,
+      "step": 39975
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.3044221744025548e-06,
+      "loss": 2.0844,
+      "step": 39980
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.2941223550937206e-06,
+      "loss": 2.0899,
+      "step": 39985
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.2838383441957777e-06,
+      "loss": 2.1471,
+      "step": 39990
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.273570143389837e-06,
+      "loss": 2.0974,
+      "step": 39995
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.2633177543544337e-06,
+      "loss": 2.0777,
+      "step": 40000
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.2530811787654937e-06,
+      "loss": 2.1023,
+      "step": 40005
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.2428604182964117e-06,
+      "loss": 2.1193,
+      "step": 40010
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.2326554746179406e-06,
+      "loss": 2.1033,
+      "step": 40015
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.22246634939829e-06,
+      "loss": 2.1167,
+      "step": 40020
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.2122930443030497e-06,
+      "loss": 2.0952,
+      "step": 40025
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.2021355609952566e-06,
+      "loss": 2.091,
+      "step": 40030
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.191993901135315e-06,
+      "loss": 2.0754,
+      "step": 40035
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1818680663811106e-06,
+      "loss": 2.1201,
+      "step": 40040
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.1717580583878636e-06,
+      "loss": 2.1242,
+      "step": 40045
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.161663878808285e-06,
+      "loss": 2.1529,
+      "step": 40050
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.1515855292924334e-06,
+      "loss": 2.0668,
+      "step": 40055
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.1415230114878237e-06,
+      "loss": 2.0744,
+      "step": 40060
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.13147632703934e-06,
+      "loss": 2.0534,
+      "step": 40065
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.1214454775893354e-06,
+      "loss": 2.1113,
+      "step": 40070
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.1114304647775093e-06,
+      "loss": 2.1459,
+      "step": 40075
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.1014312902410413e-06,
+      "loss": 2.1471,
+      "step": 40080
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.091447955614457e-06,
+      "loss": 2.0806,
+      "step": 40085
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.08148046252974e-06,
+      "loss": 2.1434,
+      "step": 40090
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.0715288126162534e-06,
+      "loss": 2.0682,
+      "step": 40095
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.0615930075007957e-06,
+      "loss": 2.1181,
+      "step": 40100
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.0516730488075464e-06,
+      "loss": 2.12,
+      "step": 40105
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.0417689381581406e-06,
+      "loss": 2.1139,
+      "step": 40110
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.0318806771715503e-06,
+      "loss": 2.1413,
+      "step": 40115
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.022008267464238e-06,
+      "loss": 2.0925,
+      "step": 40120
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.012151710650013e-06,
+      "loss": 2.116,
+      "step": 40125
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.0023110083401305e-06,
+      "loss": 2.1142,
+      "step": 40130
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9924861621432266e-06,
+      "loss": 2.0955,
+      "step": 40135
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.9826771736653714e-06,
+      "loss": 2.1361,
+      "step": 40140
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.972884044510016e-06,
+      "loss": 2.1177,
+      "step": 40145
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.9631067762780573e-06,
+      "loss": 2.1553,
+      "step": 40150
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.9533453705677507e-06,
+      "loss": 2.1462,
+      "step": 40155
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.9435998289747967e-06,
+      "loss": 2.1795,
+      "step": 40160
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.9338701530922775e-06,
+      "loss": 2.0652,
+      "step": 40165
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.924156344510709e-06,
+      "loss": 2.0874,
+      "step": 40170
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.914458404817999e-06,
+      "loss": 2.0766,
+      "step": 40175
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.904776335599435e-06,
+      "loss": 2.136,
+      "step": 40180
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.895110138437751e-06,
+      "loss": 2.103,
+      "step": 40185
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8854598149130717e-06,
+      "loss": 2.0899,
+      "step": 40190
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8758253666029244e-06,
+      "loss": 2.1001,
+      "step": 40195
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.866206795082238e-06,
+      "loss": 2.0868,
+      "step": 40200
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.8566041019233657e-06,
+      "loss": 2.1107,
+      "step": 40205
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.847017288696019e-06,
+      "loss": 2.1181,
+      "step": 40210
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.837446356967388e-06,
+      "loss": 2.1288,
+      "step": 40215
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.8278913083019777e-06,
+      "loss": 2.0949,
+      "step": 40220
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.818352144261771e-06,
+      "loss": 2.1215,
+      "step": 40225
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.8088288664061214e-06,
+      "loss": 2.1014,
+      "step": 40230
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.799321476291783e-06,
+      "loss": 2.0811,
+      "step": 40235
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.789829975472913e-06,
+      "loss": 2.1226,
+      "step": 40240
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7803543655011035e-06,
+      "loss": 2.0594,
+      "step": 40245
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.7708946479252816e-06,
+      "loss": 2.1052,
+      "step": 40250
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.761450824291867e-06,
+      "loss": 2.0942,
+      "step": 40255
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.7520228961445903e-06,
+      "loss": 2.0824,
+      "step": 40260
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.742610865024653e-06,
+      "loss": 2.0856,
+      "step": 40265
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.7332147324706014e-06,
+      "loss": 2.0921,
+      "step": 40270
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.7238345000184406e-06,
+      "loss": 2.0943,
+      "step": 40275
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.714470169201533e-06,
+      "loss": 2.1115,
+      "step": 40280
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.7051217415506647e-06,
+      "loss": 2.0383,
+      "step": 40285
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.6957892185939913e-06,
+      "loss": 2.1154,
+      "step": 40290
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.6864726018571263e-06,
+      "loss": 2.0869,
+      "step": 40295
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.6771718928630174e-06,
+      "loss": 2.0831,
+      "step": 40300
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.66788709313206e-06,
+      "loss": 2.0408,
+      "step": 40305
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.6586182041820174e-06,
+      "loss": 2.0638,
+      "step": 40310
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.6493652275280778e-06,
+      "loss": 2.1021,
+      "step": 40315
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.640128164682809e-06,
+      "loss": 2.1282,
+      "step": 40320
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.630907017156181e-06,
+      "loss": 2.1475,
+      "step": 40325
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.621701786455566e-06,
+      "loss": 2.1045,
+      "step": 40330
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.6125124740857487e-06,
+      "loss": 2.1263,
+      "step": 40335
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.603339081548872e-06,
+      "loss": 2.0611,
+      "step": 40340
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.594181610344526e-06,
+      "loss": 2.0843,
+      "step": 40345
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.585040061969657e-06,
+      "loss": 2.1205,
+      "step": 40350
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.5759144379186364e-06,
+      "loss": 2.0848,
+      "step": 40355
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.566804739683204e-06,
+      "loss": 2.1494,
+      "step": 40360
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.5577109687525135e-06,
+      "loss": 2.0784,
+      "step": 40365
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.5486331266131423e-06,
+      "loss": 2.076,
+      "step": 40370
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5395712147490035e-06,
+      "loss": 2.1029,
+      "step": 40375
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.5305252346414564e-06,
+      "loss": 2.1399,
+      "step": 40380
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.5214951877692185e-06,
+      "loss": 2.1285,
+      "step": 40385
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.5124810756084527e-06,
+      "loss": 2.1331,
+      "step": 40390
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.50348289963267e-06,
+      "loss": 2.0933,
+      "step": 40395
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.494500661312793e-06,
+      "loss": 2.0689,
+      "step": 40400
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4855343621171367e-06,
+      "loss": 2.0698,
+      "step": 40405
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4765840035114176e-06,
+      "loss": 2.1066,
+      "step": 40410
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.467649586958731e-06,
+      "loss": 2.1284,
+      "step": 40415
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.458731113919599e-06,
+      "loss": 2.1437,
+      "step": 40420
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.449828585851899e-06,
+      "loss": 2.0898,
+      "step": 40425
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.440942004210922e-06,
+      "loss": 2.0985,
+      "step": 40430
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.4320713704493405e-06,
+      "loss": 2.0949,
+      "step": 40435
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.423216686017249e-06,
+      "loss": 2.0852,
+      "step": 40440
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4143779523620792e-06,
+      "loss": 2.1007,
+      "step": 40445
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4055551709287307e-06,
+      "loss": 2.1166,
+      "step": 40450
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3967483431594273e-06,
+      "loss": 2.1185,
+      "step": 40455
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.387957470493818e-06,
+      "loss": 2.0591,
+      "step": 40460
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.3791825543689308e-06,
+      "loss": 2.0742,
+      "step": 40465
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.370423596219218e-06,
+      "loss": 2.111,
+      "step": 40470
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.361680597476468e-06,
+      "loss": 2.1133,
+      "step": 40475
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3529535595699036e-06,
+      "loss": 2.0955,
+      "step": 40480
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.344242483926129e-06,
+      "loss": 2.1011,
+      "step": 40485
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3355473719691267e-06,
+      "loss": 2.1056,
+      "step": 40490
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.326868225120271e-06,
+      "loss": 2.1366,
+      "step": 40495
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.3182050447983495e-06,
+      "loss": 2.1075,
+      "step": 40500
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.309557832419518e-06,
+      "loss": 2.1257,
+      "step": 40505
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.300926589397323e-06,
+      "loss": 2.1174,
+      "step": 40510
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2923113171427034e-06,
+      "loss": 2.1486,
+      "step": 40515
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.283712017063999e-06,
+      "loss": 2.0985,
+      "step": 40520
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.275128690566919e-06,
+      "loss": 2.136,
+      "step": 40525
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2665613390545846e-06,
+      "loss": 2.1483,
+      "step": 40530
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2580099639274656e-06,
+      "loss": 2.0779,
+      "step": 40535
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.249474566583487e-06,
+      "loss": 2.1031,
+      "step": 40540
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.240955148417867e-06,
+      "loss": 2.1506,
+      "step": 40545
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2324517108233247e-06,
+      "loss": 2.0811,
+      "step": 40550
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.223964255189859e-06,
+      "loss": 2.0801,
+      "step": 40555
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.2154927829049375e-06,
+      "loss": 2.1002,
+      "step": 40560
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.207037295353365e-06,
+      "loss": 2.1443,
+      "step": 40565
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.1985977939173675e-06,
+      "loss": 2.0622,
+      "step": 40570
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.1901742799765203e-06,
+      "loss": 2.0852,
+      "step": 40575
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.1817667549078325e-06,
+      "loss": 2.1063,
+      "step": 40580
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.1733752200856496e-06,
+      "loss": 2.0853,
+      "step": 40585
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.1649996768817402e-06,
+      "loss": 2.1045,
+      "step": 40590
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1566401266652435e-06,
+      "loss": 2.0708,
+      "step": 40595
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.148296570802699e-06,
+      "loss": 2.1696,
+      "step": 40600
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.375,
+      "learning_rate": 2.139969010657994e-06,
+      "loss": 2.1437,
+      "step": 40605
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.1316574475924387e-06,
+      "loss": 2.1301,
+      "step": 40610
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.123361882964714e-06,
+      "loss": 2.1039,
+      "step": 40615
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.115082318130901e-06,
+      "loss": 2.1321,
+      "step": 40620
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.1068187544444284e-06,
+      "loss": 2.071,
+      "step": 40625
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.0985711932561603e-06,
+      "loss": 2.093,
+      "step": 40630
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.0903396359142844e-06,
+      "loss": 2.102,
+      "step": 40635
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0821240837644364e-06,
+      "loss": 2.1461,
+      "step": 40640
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.073924538149574e-06,
+      "loss": 2.1015,
+      "step": 40645
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.065741000410104e-06,
+      "loss": 2.101,
+      "step": 40650
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0575734718837446e-06,
+      "loss": 2.0422,
+      "step": 40655
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0494219539056613e-06,
+      "loss": 2.1491,
+      "step": 40660
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.0412864478083438e-06,
+      "loss": 2.0779,
+      "step": 40665
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.033166954921739e-06,
+      "loss": 2.1221,
+      "step": 40670
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.025063476573086e-06,
+      "loss": 2.0954,
+      "step": 40675
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0169760140870906e-06,
+      "loss": 2.13,
+      "step": 40680
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0089045687857632e-06,
+      "loss": 2.0944,
+      "step": 40685
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.00084914198857e-06,
+      "loss": 2.0783,
+      "step": 40690
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.992809735012313e-06,
+      "loss": 2.0664,
+      "step": 40695
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9847863491711747e-06,
+      "loss": 2.0812,
+      "step": 40700
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9767789857767396e-06,
+      "loss": 2.1037,
+      "step": 40705
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9687876461379596e-06,
+      "loss": 2.113,
+      "step": 40710
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.960812331561179e-06,
+      "loss": 2.1316,
+      "step": 40715
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9528530433501113e-06,
+      "loss": 2.1535,
+      "step": 40720
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9449097828058484e-06,
+      "loss": 2.0868,
+      "step": 40725
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9369825512268513e-06,
+      "loss": 2.0936,
+      "step": 40730
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.929071349909006e-06,
+      "loss": 2.0992,
+      "step": 40735
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.921176180145534e-06,
+      "loss": 2.0313,
+      "step": 40740
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.913297043227047e-06,
+      "loss": 2.1215,
+      "step": 40745
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9054339404415478e-06,
+      "loss": 2.0912,
+      "step": 40750
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8975868730744084e-06,
+      "loss": 2.0807,
+      "step": 40755
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8897558424083806e-06,
+      "loss": 2.0793,
+      "step": 40760
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8819408497235957e-06,
+      "loss": 2.1036,
+      "step": 40765
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8741418962975543e-06,
+      "loss": 2.1331,
+      "step": 40770
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8663589834051475e-06,
+      "loss": 2.0823,
+      "step": 40775
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8585921123186467e-06,
+      "loss": 2.1054,
+      "step": 40780
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8508412843076917e-06,
+      "loss": 2.14,
+      "step": 40785
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8431065006392912e-06,
+      "loss": 2.1154,
+      "step": 40790
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.835387762577867e-06,
+      "loss": 2.1282,
+      "step": 40795
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8276850713851656e-06,
+      "loss": 2.0729,
+      "step": 40800
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8199984283203576e-06,
+      "loss": 2.1107,
+      "step": 40805
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8123278346399596e-06,
+      "loss": 2.1012,
+      "step": 40810
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8046732915978804e-06,
+      "loss": 2.0866,
+      "step": 40815
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7970348004453963e-06,
+      "loss": 2.0816,
+      "step": 40820
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7894123624311755e-06,
+      "loss": 2.0693,
+      "step": 40825
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.781805978801232e-06,
+      "loss": 2.0457,
+      "step": 40830
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7742156507989938e-06,
+      "loss": 2.1278,
+      "step": 40835
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7666413796652237e-06,
+      "loss": 2.1175,
+      "step": 40840
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.759083166638087e-06,
+      "loss": 2.1811,
+      "step": 40845
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7515410129531285e-06,
+      "loss": 2.1631,
+      "step": 40850
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.744014919843251e-06,
+      "loss": 2.1174,
+      "step": 40855
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7365048885387148e-06,
+      "loss": 2.1004,
+      "step": 40860
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7290109202672155e-06,
+      "loss": 2.0824,
+      "step": 40865
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7215330162537512e-06,
+      "loss": 2.1254,
+      "step": 40870
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7140711777207551e-06,
+      "loss": 2.1095,
+      "step": 40875
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.706625405887974e-06,
+      "loss": 2.0806,
+      "step": 40880
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6991957019726002e-06,
+      "loss": 2.0757,
+      "step": 40885
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6917820671891182e-06,
+      "loss": 2.0881,
+      "step": 40890
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6843845027494586e-06,
+      "loss": 2.1367,
+      "step": 40895
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6770030098628875e-06,
+      "loss": 2.1223,
+      "step": 40900
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6696375897360394e-06,
+      "loss": 2.0813,
+      "step": 40905
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6622882435729403e-06,
+      "loss": 2.1148,
+      "step": 40910
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6549549725749625e-06,
+      "loss": 2.0828,
+      "step": 40915
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6476377779409024e-06,
+      "loss": 2.0838,
+      "step": 40920
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6403366608668702e-06,
+      "loss": 2.1436,
+      "step": 40925
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6330516225463777e-06,
+      "loss": 2.0557,
+      "step": 40930
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6257826641702945e-06,
+      "loss": 2.1266,
+      "step": 40935
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6185297869268811e-06,
+      "loss": 2.092,
+      "step": 40940
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.611292992001756e-06,
+      "loss": 2.0808,
+      "step": 40945
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6040722805779063e-06,
+      "loss": 2.0923,
+      "step": 40950
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5968676538356986e-06,
+      "loss": 2.1113,
+      "step": 40955
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5896791129528687e-06,
+      "loss": 2.1001,
+      "step": 40960
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.58250665910451e-06,
+      "loss": 2.1236,
+      "step": 40965
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.575350293463107e-06,
+      "loss": 2.1048,
+      "step": 40970
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5682100171984904e-06,
+      "loss": 2.0379,
+      "step": 40975
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5610858314778932e-06,
+      "loss": 2.1166,
+      "step": 40980
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5539777374658727e-06,
+      "loss": 2.0669,
+      "step": 40985
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5468857363244105e-06,
+      "loss": 2.0885,
+      "step": 40990
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5398098292128016e-06,
+      "loss": 2.074,
+      "step": 40995
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.532750017287754e-06,
+      "loss": 2.0696,
+      "step": 41000
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5257063017033223e-06,
+      "loss": 2.141,
+      "step": 41005
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5186786836109524e-06,
+      "loss": 2.1251,
+      "step": 41010
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5116671641594137e-06,
+      "loss": 2.0928,
+      "step": 41015
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.50467174449489e-06,
+      "loss": 2.0931,
+      "step": 41020
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4976924257608992e-06,
+      "loss": 2.0846,
+      "step": 41025
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4907292090983737e-06,
+      "loss": 2.1236,
+      "step": 41030
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4837820956455474e-06,
+      "loss": 2.1082,
+      "step": 41035
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4768510865380891e-06,
+      "loss": 2.0994,
+      "step": 41040
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4699361829089708e-06,
+      "loss": 2.118,
+      "step": 41045
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4630373858885994e-06,
+      "loss": 2.1046,
+      "step": 41050
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4561546966046946e-06,
+      "loss": 2.1087,
+      "step": 41055
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.449288116182368e-06,
+      "loss": 2.0893,
+      "step": 41060
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.442437645744088e-06,
+      "loss": 2.1684,
+      "step": 41065
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4356032864097036e-06,
+      "loss": 2.0775,
+      "step": 41070
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4287850392963987e-06,
+      "loss": 2.0708,
+      "step": 41075
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4219829055187707e-06,
+      "loss": 2.0712,
+      "step": 41080
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4151968861887521e-06,
+      "loss": 2.0855,
+      "step": 41085
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4084269824156338e-06,
+      "loss": 2.0777,
+      "step": 41090
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4016731953060969e-06,
+      "loss": 2.1071,
+      "step": 41095
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3949355259641693e-06,
+      "loss": 2.1241,
+      "step": 41100
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3882139754912483e-06,
+      "loss": 2.0652,
+      "step": 41105
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3815085449860987e-06,
+      "loss": 2.0649,
+      "step": 41110
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3748192355448774e-06,
+      "loss": 2.1149,
+      "step": 41115
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.368146048261032e-06,
+      "loss": 2.1081,
+      "step": 41120
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3614889842254675e-06,
+      "loss": 2.1272,
+      "step": 41125
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.35484804452638e-06,
+      "loss": 2.1373,
+      "step": 41130
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.348223230249357e-06,
+      "loss": 2.0812,
+      "step": 41135
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3416145424773651e-06,
+      "loss": 2.0914,
+      "step": 41140
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3350219822907073e-06,
+      "loss": 2.0965,
+      "step": 41145
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3284455507670657e-06,
+      "loss": 2.1031,
+      "step": 41150
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3218852489814914e-06,
+      "loss": 2.1346,
+      "step": 41155
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3153410780063712e-06,
+      "loss": 2.0701,
+      "step": 41160
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3088130389114938e-06,
+      "loss": 2.1111,
+      "step": 41165
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3023011327639834e-06,
+      "loss": 2.1057,
+      "step": 41170
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2958053606283328e-06,
+      "loss": 2.1905,
+      "step": 41175
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2893257235664036e-06,
+      "loss": 2.0997,
+      "step": 41180
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2828622226374044e-06,
+      "loss": 2.0726,
+      "step": 41185
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2764148588979342e-06,
+      "loss": 2.1466,
+      "step": 41190
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2699836334019277e-06,
+      "loss": 2.1601,
+      "step": 41195
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2635685472006776e-06,
+      "loss": 2.0736,
+      "step": 41200
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.257169601342878e-06,
+      "loss": 2.1493,
+      "step": 41205
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2507867968745369e-06,
+      "loss": 2.0756,
+      "step": 41210
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.244420134839064e-06,
+      "loss": 2.107,
+      "step": 41215
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2380696162771822e-06,
+      "loss": 2.0868,
+      "step": 41220
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2317352422270389e-06,
+      "loss": 2.0738,
+      "step": 41225
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.225417013724084e-06,
+      "loss": 2.0912,
+      "step": 41230
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2191149318011574e-06,
+      "loss": 2.1277,
+      "step": 41235
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2128289974884465e-06,
+      "loss": 2.1205,
+      "step": 41240
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2065592118135294e-06,
+      "loss": 2.0518,
+      "step": 41245
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2003055758013082e-06,
+      "loss": 2.0637,
+      "step": 41250
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.1940680904740543e-06,
+      "loss": 2.0682,
+      "step": 41255
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1878467568513963e-06,
+      "loss": 2.0648,
+      "step": 41260
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.181641575950354e-06,
+      "loss": 2.1368,
+      "step": 41265
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1754525487852608e-06,
+      "loss": 2.0983,
+      "step": 41270
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1692796763678405e-06,
+      "loss": 2.066,
+      "step": 41275
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1631229597071525e-06,
+      "loss": 2.1483,
+      "step": 41280
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1569823998096473e-06,
+      "loss": 2.1095,
+      "step": 41285
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1508579976791113e-06,
+      "loss": 2.1014,
+      "step": 41290
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1447497543166875e-06,
+      "loss": 2.114,
+      "step": 41295
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1386576707208773e-06,
+      "loss": 2.1325,
+      "step": 41300
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1325817478875734e-06,
+      "loss": 2.114,
+      "step": 41305
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1265219868099696e-06,
+      "loss": 2.1781,
+      "step": 41310
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.120478388478674e-06,
+      "loss": 2.1019,
+      "step": 41315
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1144509538816183e-06,
+      "loss": 2.1264,
+      "step": 41320
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1084396840040922e-06,
+      "loss": 2.024,
+      "step": 41325
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1024445798287653e-06,
+      "loss": 2.0758,
+      "step": 41330
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0964656423356534e-06,
+      "loss": 2.0628,
+      "step": 41335
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0905028725021082e-06,
+      "loss": 2.0717,
+      "step": 41340
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0845562713028723e-06,
+      "loss": 2.0738,
+      "step": 41345
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.0786258397100234e-06,
+      "loss": 2.0836,
+      "step": 41350
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0727115786930197e-06,
+      "loss": 2.112,
+      "step": 41355
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0668134892186321e-06,
+      "loss": 2.1625,
+      "step": 41360
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.060931572251045e-06,
+      "loss": 2.1466,
+      "step": 41365
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0550658287517335e-06,
+      "loss": 2.1302,
+      "step": 41370
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.0492162596796085e-06,
+      "loss": 2.1224,
+      "step": 41375
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0433828659908496e-06,
+      "loss": 2.1347,
+      "step": 41380
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0375656486390717e-06,
+      "loss": 2.134,
+      "step": 41385
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0317646085751809e-06,
+      "loss": 2.089,
+      "step": 41390
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0259797467474962e-06,
+      "loss": 2.1184,
+      "step": 41395
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0202110641016282e-06,
+      "loss": 2.1118,
+      "step": 41400
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0144585615806113e-06,
+      "loss": 2.0763,
+      "step": 41405
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0087222401247709e-06,
+      "loss": 2.0873,
+      "step": 41410
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.003002100671846e-06,
+      "loss": 2.1011,
+      "step": 41415
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.972981441568885e-07,
+      "loss": 2.1205,
+      "step": 41420
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.916103715123192e-07,
+      "loss": 2.1189,
+      "step": 41425
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.859387836679057e-07,
+      "loss": 2.1119,
+      "step": 41430
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.80283381550784e-07,
+      "loss": 2.0929,
+      "step": 41435
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.746441660854367e-07,
+      "loss": 2.098,
+      "step": 41440
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.690211381937041e-07,
+      "loss": 2.0813,
+      "step": 41445
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.634142987947737e-07,
+      "loss": 2.0998,
+      "step": 41450
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.578236488051785e-07,
+      "loss": 2.1691,
+      "step": 41455
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.522491891388319e-07,
+      "loss": 2.0905,
+      "step": 41460
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.466909207069719e-07,
+      "loss": 2.0956,
+      "step": 41465
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.41148844418227e-07,
+      "loss": 2.0552,
+      "step": 41470
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.356229611785172e-07,
+      "loss": 2.1013,
+      "step": 41475
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.301132718911975e-07,
+      "loss": 2.1273,
+      "step": 41480
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.246197774568921e-07,
+      "loss": 2.0946,
+      "step": 41485
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.19142478773638e-07,
+      "loss": 2.1052,
+      "step": 41490
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.136813767367969e-07,
+      "loss": 2.0899,
+      "step": 41495
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.08236472239088e-07,
+      "loss": 2.1028,
+      "step": 41500
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.028077661705881e-07,
+      "loss": 2.0727,
+      "step": 41505
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.973952594187318e-07,
+      "loss": 2.1403,
+      "step": 41510
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.919989528682671e-07,
+      "loss": 2.1336,
+      "step": 41515
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.866188474013548e-07,
+      "loss": 2.1977,
+      "step": 41520
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.812549438974583e-07,
+      "loss": 2.0894,
+      "step": 41525
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.759072432334203e-07,
+      "loss": 2.105,
+      "step": 41530
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.705757462833974e-07,
+      "loss": 2.0949,
+      "step": 41535
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.652604539189591e-07,
+      "loss": 2.1089,
+      "step": 41540
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.599613670089657e-07,
+      "loss": 2.1626,
+      "step": 41545
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.546784864196689e-07,
+      "loss": 2.1212,
+      "step": 41550
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.494118130146445e-07,
+      "loss": 2.071,
+      "step": 41555
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.441613476548482e-07,
+      "loss": 2.1067,
+      "step": 41560
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.389270911985381e-07,
+      "loss": 2.102,
+      "step": 41565
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.33709044501374e-07,
+      "loss": 2.0953,
+      "step": 41570
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.285072084163403e-07,
+      "loss": 2.0596,
+      "step": 41575
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.233215837937791e-07,
+      "loss": 2.0785,
+      "step": 41580
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.181521714813789e-07,
+      "loss": 2.0802,
+      "step": 41585
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 8.129989723241749e-07,
+      "loss": 2.1127,
+      "step": 41590
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 8.078619871645487e-07,
+      "loss": 2.0674,
+      "step": 41595
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.02741216842251e-07,
+      "loss": 2.1239,
+      "step": 41600
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.976366621943565e-07,
+      "loss": 2.1109,
+      "step": 41605
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.925483240553089e-07,
+      "loss": 2.0568,
+      "step": 41610
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.874762032568873e-07,
+      "loss": 2.147,
+      "step": 41615
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.824203006282504e-07,
+      "loss": 2.1027,
+      "step": 41620
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.773806169958486e-07,
+      "loss": 2.0976,
+      "step": 41625
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.723571531835338e-07,
+      "loss": 2.111,
+      "step": 41630
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.673499100124714e-07,
+      "loss": 2.092,
+      "step": 41635
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.623588883012068e-07,
+      "loss": 2.1382,
+      "step": 41640
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.573840888656091e-07,
+      "loss": 2.1469,
+      "step": 41645
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.52425512518895e-07,
+      "loss": 2.0791,
+      "step": 41650
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.474831600716603e-07,
+      "loss": 2.0862,
+      "step": 41655
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.425570323318032e-07,
+      "loss": 2.1237,
+      "step": 41660
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.376471301046128e-07,
+      "loss": 2.1051,
+      "step": 41665
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.327534541926917e-07,
+      "loss": 2.1249,
+      "step": 41670
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.278760053959999e-07,
+      "loss": 2.0723,
+      "step": 41675
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.230147845118662e-07,
+      "loss": 2.1182,
+      "step": 41680
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.18169792334944e-07,
+      "loss": 2.0723,
+      "step": 41685
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.133410296572218e-07,
+      "loss": 2.1223,
+      "step": 41690
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.085284972680795e-07,
+      "loss": 2.1077,
+      "step": 41695
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.037321959541987e-07,
+      "loss": 2.0886,
+      "step": 41700
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.989521264996302e-07,
+      "loss": 2.1536,
+      "step": 41705
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.941882896857599e-07,
+      "loss": 2.1287,
+      "step": 41710
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.894406862913538e-07,
+      "loss": 2.1483,
+      "step": 41715
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.847093170924579e-07,
+      "loss": 2.0968,
+      "step": 41720
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.35546875,
+      "learning_rate": 6.799941828625422e-07,
+      "loss": 2.1137,
+      "step": 41725
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.75295284372357e-07,
+      "loss": 2.1067,
+      "step": 41730
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.706126223900433e-07,
+      "loss": 2.1079,
+      "step": 41735
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.659461976810666e-07,
+      "loss": 2.138,
+      "step": 41740
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.612960110082389e-07,
+      "loss": 2.1511,
+      "step": 41745
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.566620631317188e-07,
+      "loss": 2.0883,
+      "step": 41750
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.520443548090338e-07,
+      "loss": 2.1152,
+      "step": 41755
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.474428867950022e-07,
+      "loss": 2.094,
+      "step": 41760
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.428576598418556e-07,
+      "loss": 2.0816,
+      "step": 41765
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.382886746991057e-07,
+      "loss": 2.1122,
+      "step": 41770
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.337359321136549e-07,
+      "loss": 2.0832,
+      "step": 41775
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.291994328297413e-07,
+      "loss": 2.05,
+      "step": 41780
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.246791775889383e-07,
+      "loss": 2.0752,
+      "step": 41785
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.201751671301548e-07,
+      "loss": 2.0818,
+      "step": 41790
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.156874021896686e-07,
+      "loss": 2.098,
+      "step": 41795
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.112158835010928e-07,
+      "loss": 2.0762,
+      "step": 41800
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.067606117953761e-07,
+      "loss": 2.114,
+      "step": 41805
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.023215878008137e-07,
+      "loss": 2.1302,
+      "step": 41810
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.978988122430584e-07,
+      "loss": 2.1162,
+      "step": 41815
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.934922858450875e-07,
+      "loss": 2.0708,
+      "step": 41820
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.891020093272359e-07,
+      "loss": 2.0507,
+      "step": 41825
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.84727983407174e-07,
+      "loss": 2.0617,
+      "step": 41830
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.803702087999297e-07,
+      "loss": 2.097,
+      "step": 41835
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.760286862178443e-07,
+      "loss": 2.1139,
+      "step": 41840
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.717034163706391e-07,
+      "loss": 2.095,
+      "step": 41845
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.673943999653486e-07,
+      "loss": 2.1573,
+      "step": 41850
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.631016377063869e-07,
+      "loss": 2.0912,
+      "step": 41855
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.588251302954594e-07,
+      "loss": 2.1124,
+      "step": 41860
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.545648784316626e-07,
+      "loss": 2.0665,
+      "step": 41865
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.503208828114059e-07,
+      "loss": 2.0922,
+      "step": 41870
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.460931441284456e-07,
+      "loss": 2.1622,
+      "step": 41875
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.418816630738954e-07,
+      "loss": 2.1377,
+      "step": 41880
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.376864403362047e-07,
+      "loss": 2.0708,
+      "step": 41885
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.335074766011472e-07,
+      "loss": 2.1352,
+      "step": 41890
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.293447725518652e-07,
+      "loss": 2.1353,
+      "step": 41895
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.251983288688367e-07,
+      "loss": 2.1378,
+      "step": 41900
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.210681462298639e-07,
+      "loss": 2.1398,
+      "step": 41905
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.169542253101179e-07,
+      "loss": 2.0958,
+      "step": 41910
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.128565667820829e-07,
+      "loss": 2.1676,
+      "step": 41915
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.087751713156008e-07,
+      "loss": 2.1175,
+      "step": 41920
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.047100395778604e-07,
+      "loss": 2.04,
+      "step": 41925
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.006611722333854e-07,
+      "loss": 2.089,
+      "step": 41930
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.966285699440354e-07,
+      "loss": 2.1074,
+      "step": 41935
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.926122333690164e-07,
+      "loss": 2.1237,
+      "step": 41940
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.886121631648588e-07,
+      "loss": 2.0405,
+      "step": 41945
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.846283599854839e-07,
+      "loss": 2.1525,
+      "step": 41950
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.806608244820931e-07,
+      "loss": 2.0755,
+      "step": 41955
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.767095573032676e-07,
+      "loss": 2.1363,
+      "step": 41960
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.727745590949129e-07,
+      "loss": 2.0727,
+      "step": 41965
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.6885583050028105e-07,
+      "loss": 2.1273,
+      "step": 41970
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.6495337215995973e-07,
+      "loss": 2.0529,
+      "step": 41975
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.6106718471188306e-07,
+      "loss": 2.1406,
+      "step": 41980
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.571972687913206e-07,
+      "loss": 2.09,
+      "step": 41985
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.5334362503088867e-07,
+      "loss": 2.0677,
+      "step": 41990
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.495062540605166e-07,
+      "loss": 2.1236,
+      "step": 41995
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.4568515650752485e-07,
+      "loss": 2.0776,
+      "step": 42000
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.4188033299652486e-07,
+      "loss": 2.0652,
+      "step": 42005
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.380917841494969e-07,
+      "loss": 2.0268,
+      "step": 42010
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.343195105857567e-07,
+      "loss": 2.079,
+      "step": 42015
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.305635129219332e-07,
+      "loss": 2.0772,
+      "step": 42020
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.2682379177203527e-07,
+      "loss": 2.1241,
+      "step": 42025
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.231003477473738e-07,
+      "loss": 2.113,
+      "step": 42030
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.1939318145663985e-07,
+      "loss": 2.1181,
+      "step": 42035
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.1570229350581526e-07,
+      "loss": 2.1172,
+      "step": 42040
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.1202768449826176e-07,
+      "loss": 2.0777,
+      "step": 42045
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.0836935503466566e-07,
+      "loss": 2.1261,
+      "step": 42050
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.047273057130263e-07,
+      "loss": 2.0739,
+      "step": 42055
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.011015371287452e-07,
+      "loss": 2.143,
+      "step": 42060
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.9749204987448165e-07,
+      "loss": 2.1519,
+      "step": 42065
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.93898844540308e-07,
+      "loss": 2.1228,
+      "step": 42070
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.903219217135767e-07,
+      "loss": 2.1858,
+      "step": 42075
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.8676128197903117e-07,
+      "loss": 2.1111,
+      "step": 42080
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.832169259186946e-07,
+      "loss": 2.1016,
+      "step": 42085
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.796888541119814e-07,
+      "loss": 2.1135,
+      "step": 42090
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.7617706713561905e-07,
+      "loss": 2.1465,
+      "step": 42095
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.7268156556365954e-07,
+      "loss": 2.0519,
+      "step": 42100
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.6920234996752343e-07,
+      "loss": 2.142,
+      "step": 42105
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.657394209159559e-07,
+      "loss": 2.0964,
+      "step": 42110
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.6229277897503743e-07,
+      "loss": 2.0909,
+      "step": 42115
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.58862424708184e-07,
+      "loss": 2.1339,
+      "step": 42120
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.55448358676147e-07,
+      "loss": 2.1075,
+      "step": 42125
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.520505814370245e-07,
+      "loss": 2.153,
+      "step": 42130
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.4866909354625e-07,
+      "loss": 2.1865,
+      "step": 42135
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.453038955565924e-07,
+      "loss": 2.1025,
+      "step": 42140
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.4195498801815605e-07,
+      "loss": 2.079,
+      "step": 42145
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.3862237147838094e-07,
+      "loss": 2.1049,
+      "step": 42150
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.3530604648205344e-07,
+      "loss": 2.1377,
+      "step": 42155
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.320060135712733e-07,
+      "loss": 2.0981,
+      "step": 42160
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.2872227328550885e-07,
+      "loss": 2.1062,
+      "step": 42165
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.254548261615531e-07,
+      "loss": 2.1149,
+      "step": 42170
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.2220367273352317e-07,
+      "loss": 2.1414,
+      "step": 42175
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.1896881353288276e-07,
+      "loss": 2.0905,
+      "step": 42180
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.157502490884312e-07,
+      "loss": 2.0992,
+      "step": 42185
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.1254797992629205e-07,
+      "loss": 2.0679,
+      "step": 42190
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.093620065699687e-07,
+      "loss": 2.0735,
+      "step": 42195
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.061923295402336e-07,
+      "loss": 2.0591,
+      "step": 42200
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.03038949355261e-07,
+      "loss": 2.1115,
+      "step": 42205
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9990186653050534e-07,
+      "loss": 2.0975,
+      "step": 42210
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.967810815788008e-07,
+      "loss": 2.1297,
+      "step": 42215
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.936765950102838e-07,
+      "loss": 2.1147,
+      "step": 42220
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.905884073324594e-07,
+      "loss": 2.1197,
+      "step": 42225
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.87516519050135e-07,
+      "loss": 2.0359,
+      "step": 42230
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.8446093066547554e-07,
+      "loss": 2.1326,
+      "step": 42235
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.814216426779703e-07,
+      "loss": 2.1145,
+      "step": 42240
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.783986555844553e-07,
+      "loss": 2.1036,
+      "step": 42245
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.7539196987907967e-07,
+      "loss": 2.0853,
+      "step": 42250
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.724015860533724e-07,
+      "loss": 2.0954,
+      "step": 42255
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.694275045961425e-07,
+      "loss": 2.1064,
+      "step": 42260
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.664697259935789e-07,
+      "loss": 2.0427,
+      "step": 42265
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.635282507291614e-07,
+      "loss": 2.1031,
+      "step": 42270
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.6060307928376106e-07,
+      "loss": 2.0832,
+      "step": 42275
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.5769421213552856e-07,
+      "loss": 2.1291,
+      "step": 42280
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.5480164975998367e-07,
+      "loss": 2.0812,
+      "step": 42285
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.519253926299703e-07,
+      "loss": 2.1397,
+      "step": 42290
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.490654412156568e-07,
+      "loss": 2.1,
+      "step": 42295
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.462217959845803e-07,
+      "loss": 2.1051,
+      "step": 42300
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.43394457401569e-07,
+      "loss": 2.1561,
+      "step": 42305
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4058342592879755e-07,
+      "loss": 2.1061,
+      "step": 42310
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.3778870202580961e-07,
+      "loss": 2.0563,
+      "step": 42315
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.3501028614942854e-07,
+      "loss": 2.1121,
+      "step": 42320
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.3224817875385775e-07,
+      "loss": 2.1598,
+      "step": 42325
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.2950238029061376e-07,
+      "loss": 2.1018,
+      "step": 42330
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2677289120854871e-07,
+      "loss": 2.1329,
+      "step": 42335
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.240597119538279e-07,
+      "loss": 2.0589,
+      "step": 42340
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2136284297000766e-07,
+      "loss": 2.1146,
+      "step": 42345
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.1868228469791308e-07,
+      "loss": 2.0666,
+      "step": 42350
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.1601803757576035e-07,
+      "loss": 2.053,
+      "step": 42355
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.133701020390455e-07,
+      "loss": 2.1121,
+      "step": 42360
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1073847852062235e-07,
+      "loss": 2.0893,
+      "step": 42365
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.0812316745070225e-07,
+      "loss": 2.0534,
+      "step": 42370
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.0552416925679886e-07,
+      "loss": 2.0995,
+      "step": 42375
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0294148436376116e-07,
+      "loss": 2.0641,
+      "step": 42380
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.0037511319378476e-07,
+      "loss": 2.0737,
+      "step": 42385
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9782505616637858e-07,
+      "loss": 2.0549,
+      "step": 42390
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9529131369842024e-07,
+      "loss": 2.103,
+      "step": 42395
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9277388620407844e-07,
+      "loss": 2.1028,
+      "step": 42400
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9027277409487955e-07,
+      "loss": 2.1157,
+      "step": 42405
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8778797777968537e-07,
+      "loss": 2.119,
+      "step": 42410
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8531949766467104e-07,
+      "loss": 2.1419,
+      "step": 42415
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8286733415336933e-07,
+      "loss": 2.0527,
+      "step": 42420
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.804314876466262e-07,
+      "loss": 2.0753,
+      "step": 42425
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.780119585426232e-07,
+      "loss": 2.0913,
+      "step": 42430
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7560874723688836e-07,
+      "loss": 2.0653,
+      "step": 42435
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.73221854122263e-07,
+      "loss": 2.1147,
+      "step": 42440
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.70851279588935e-07,
+      "loss": 2.1163,
+      "step": 42445
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6849702402441658e-07,
+      "loss": 2.1101,
+      "step": 42450
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.661590878135555e-07,
+      "loss": 2.0577,
+      "step": 42455
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6383747133853488e-07,
+      "loss": 2.0609,
+      "step": 42460
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.615321749788623e-07,
+      "loss": 2.1043,
+      "step": 42465
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5924319911139185e-07,
+      "loss": 2.07,
+      "step": 42470
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5697054411029089e-07,
+      "loss": 2.1858,
+      "step": 42475
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5471421034707335e-07,
+      "loss": 2.0693,
+      "step": 42480
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.524741981905775e-07,
+      "loss": 2.1156,
+      "step": 42485
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5025050800696605e-07,
+      "loss": 2.1102,
+      "step": 42490
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4804314015975928e-07,
+      "loss": 2.1058,
+      "step": 42495
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.458520950097908e-07,
+      "loss": 2.1266,
+      "step": 42500
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4367737291521854e-07,
+      "loss": 2.148,
+      "step": 42505
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4151897423154704e-07,
+      "loss": 2.0686,
+      "step": 42510
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3937689931161624e-07,
+      "loss": 2.1096,
+      "step": 42515
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.372511485055794e-07,
+      "loss": 2.1346,
+      "step": 42520
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3514172216092525e-07,
+      "loss": 2.1459,
+      "step": 42525
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3304862062248903e-07,
+      "loss": 2.0891,
+      "step": 42530
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3097184423243037e-07,
+      "loss": 2.1695,
+      "step": 42535
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2891139333023327e-07,
+      "loss": 2.1163,
+      "step": 42540
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2686726825271722e-07,
+      "loss": 2.108,
+      "step": 42545
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.248394693340371e-07,
+      "loss": 2.0776,
+      "step": 42550
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2282799690566116e-07,
+      "loss": 2.1242,
+      "step": 42555
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.208328512964263e-07,
+      "loss": 2.1454,
+      "step": 42560
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.1885403283244945e-07,
+      "loss": 2.1274,
+      "step": 42565
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.1689154183722739e-07,
+      "loss": 2.0632,
+      "step": 42570
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.1494537863155908e-07,
+      "loss": 2.1431,
+      "step": 42575
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1301554353359e-07,
+      "loss": 2.1169,
+      "step": 42580
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1110203685876786e-07,
+      "loss": 2.0899,
+      "step": 42585
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0920485891990906e-07,
+      "loss": 2.0715,
+      "step": 42590
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0732401002714331e-07,
+      "loss": 2.1451,
+      "step": 42595
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.054594904879247e-07,
+      "loss": 2.1242,
+      "step": 42600
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.0361130060704272e-07,
+      "loss": 2.1104,
+      "step": 42605
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.0177944068662237e-07,
+      "loss": 2.104,
+      "step": 42610
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.996391102612413e-08,
+      "loss": 2.1098,
+      "step": 42615
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.816471192231058e-08,
+      "loss": 2.121,
+      "step": 42620
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.638184366931314e-08,
+      "loss": 2.0857,
+      "step": 42625
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.461530655857642e-08,
+      "loss": 2.1077,
+      "step": 42630
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.286510087885835e-08,
+      "loss": 2.0955,
+      "step": 42635
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.113122691627451e-08,
+      "loss": 2.0897,
+      "step": 42640
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.941368495425373e-08,
+      "loss": 2.1352,
+      "step": 42645
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.77124752735714e-08,
+      "loss": 2.0865,
+      "step": 42650
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.60275981523051e-08,
+      "loss": 2.1064,
+      "step": 42655
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.353515625,
+      "learning_rate": 8.435905386590115e-08,
+      "loss": 2.0949,
+      "step": 42660
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 8.270684268709694e-08,
+      "loss": 2.0894,
+      "step": 42665
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.107096488598753e-08,
+      "loss": 2.1164,
+      "step": 42670
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.945142072999235e-08,
+      "loss": 2.1107,
+      "step": 42675
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.784821048383295e-08,
+      "loss": 2.0992,
+      "step": 42680
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.626133440962191e-08,
+      "loss": 2.1064,
+      "step": 42685
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.469079276672953e-08,
+      "loss": 2.0834,
+      "step": 42690
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.313658581190597e-08,
+      "loss": 2.1015,
+      "step": 42695
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.159871379921468e-08,
+      "loss": 2.1725,
+      "step": 42700
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.007717698005456e-08,
+      "loss": 2.0533,
+      "step": 42705
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.857197560313777e-08,
+      "loss": 2.0727,
+      "step": 42710
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.708310991453415e-08,
+      "loss": 2.0838,
+      "step": 42715
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.561058015761568e-08,
+      "loss": 2.079,
+      "step": 42720
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.415438657310091e-08,
+      "loss": 2.1235,
+      "step": 42725
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 6.271452939902167e-08,
+      "loss": 2.1116,
+      "step": 42730
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.33984375,
+      "learning_rate": 6.129100887075633e-08,
+      "loss": 2.0784,
+      "step": 42735
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.988382522100766e-08,
+      "loss": 2.1196,
+      "step": 42740
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.8492978679813846e-08,
+      "loss": 2.1215,
+      "step": 42745
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.711846947452637e-08,
+      "loss": 2.1091,
+      "step": 42750
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.5760297829832166e-08,
+      "loss": 2.0976,
+      "step": 42755
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.441846396775363e-08,
+      "loss": 2.0772,
+      "step": 42760
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.309296810764863e-08,
+      "loss": 2.1197,
+      "step": 42765
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.17838104661772e-08,
+      "loss": 2.1327,
+      "step": 42770
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.049099125735701e-08,
+      "loss": 2.0868,
+      "step": 42775
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.9214510692519034e-08,
+      "loss": 2.1434,
+      "step": 42780
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.795436898032968e-08,
+      "loss": 2.1554,
+      "step": 42785
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.671056632677973e-08,
+      "loss": 2.0868,
+      "step": 42790
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.548310293520652e-08,
+      "loss": 2.0958,
+      "step": 42795
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.427197900624957e-08,
+      "loss": 2.133,
+      "step": 42800
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.307719473788385e-08,
+      "loss": 2.1455,
+      "step": 42805
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.189875032544199e-08,
+      "loss": 2.1372,
+      "step": 42810
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.0736645961536594e-08,
+      "loss": 2.1386,
+      "step": 42815
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.959088183614901e-08,
+      "loss": 2.0857,
+      "step": 42820
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.846145813658497e-08,
+      "loss": 2.0807,
+      "step": 42825
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.734837504745237e-08,
+      "loss": 2.135,
+      "step": 42830
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.625163275072785e-08,
+      "loss": 2.0901,
+      "step": 42835
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.5171231425668025e-08,
+      "loss": 2.1023,
+      "step": 42840
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.410717124889828e-08,
+      "loss": 2.1784,
+      "step": 42845
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.305945239435726e-08,
+      "loss": 2.0792,
+      "step": 42850
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.2028075033330166e-08,
+      "loss": 2.1082,
+      "step": 42855
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.1013039334382155e-08,
+      "loss": 2.0985,
+      "step": 42860
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.001434546346937e-08,
+      "loss": 2.1064,
+      "step": 42865
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.903199358383901e-08,
+      "loss": 2.1423,
+      "step": 42870
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.806598385607373e-08,
+      "loss": 2.1284,
+      "step": 42875
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.7116316438080548e-08,
+      "loss": 2.1236,
+      "step": 42880
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.618299148511305e-08,
+      "loss": 2.0762,
+      "step": 42885
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.526600914972699e-08,
+      "loss": 2.105,
+      "step": 42890
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4365369581835774e-08,
+      "loss": 2.1047,
+      "step": 42895
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.3481072928654978e-08,
+      "loss": 2.0806,
+      "step": 42900
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.2613119334735645e-08,
+      "loss": 2.1328,
+      "step": 42905
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.1761508941975373e-08,
+      "loss": 2.0866,
+      "step": 42910
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.092624188957393e-08,
+      "loss": 2.0848,
+      "step": 42915
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0107318314077638e-08,
+      "loss": 2.1586,
+      "step": 42920
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9304738349346095e-08,
+      "loss": 2.1444,
+      "step": 42925
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.851850212659656e-08,
+      "loss": 2.0704,
+      "step": 42930
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.774860977432624e-08,
+      "loss": 2.1268,
+      "step": 42935
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6995061418412227e-08,
+      "loss": 2.1413,
+      "step": 42940
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6257857182033766e-08,
+      "loss": 2.0728,
+      "step": 42945
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.553699718569446e-08,
+      "loss": 2.128,
+      "step": 42950
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4832481547222277e-08,
+      "loss": 2.0909,
+      "step": 42955
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.414431038180286e-08,
+      "loss": 2.0835,
+      "step": 42960
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3472483801924007e-08,
+      "loss": 2.0876,
+      "step": 42965
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2817001917397875e-08,
+      "loss": 2.1116,
+      "step": 42970
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2177864835394292e-08,
+      "loss": 2.0757,
+      "step": 42975
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1555072660374144e-08,
+      "loss": 2.0838,
+      "step": 42980
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.0948625494155985e-08,
+      "loss": 2.1574,
+      "step": 42985
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0358523435882728e-08,
+      "loss": 2.0629,
+      "step": 42990
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.784766581988347e-09,
+      "loss": 2.0999,
+      "step": 42995
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.22735502629779e-09,
+      "loss": 2.1121,
+      "step": 43000
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.68628885991596e-09,
+      "loss": 2.1299,
+      "step": 43005
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.161568171283218e-09,
+      "loss": 2.1468,
+      "step": 43010
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.653193046186502e-09,
+      "loss": 2.0775,
+      "step": 43015
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.161163567726004e-09,
+      "loss": 2.1125,
+      "step": 43020
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.685479816326279e-09,
+      "loss": 2.115,
+      "step": 43025
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.226141869758451e-09,
+      "loss": 2.1427,
+      "step": 43030
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.7831498030958e-09,
+      "loss": 2.0686,
+      "step": 43035
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.356503688769276e-09,
+      "loss": 2.0748,
+      "step": 43040
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.9462035965230915e-09,
+      "loss": 2.104,
+      "step": 43045
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.552249593414714e-09,
+      "loss": 2.0743,
+      "step": 43050
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.174641743848184e-09,
+      "loss": 2.1023,
+      "step": 43055
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.813380109551901e-09,
+      "loss": 2.1142,
+      "step": 43060
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.468464749578626e-09,
+      "loss": 2.0618,
+      "step": 43065
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.139895720316588e-09,
+      "loss": 2.069,
+      "step": 43070
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.8276730754783765e-09,
+      "loss": 2.1337,
+      "step": 43075
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.5317968661009438e-09,
+      "loss": 2.0719,
+      "step": 43080
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.2522671405345032e-09,
+      "loss": 2.1156,
+      "step": 43085
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.989083944498038e-09,
+      "loss": 2.0763,
+      "step": 43090
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7422473210015889e-09,
+      "loss": 2.1628,
+      "step": 43095
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5117573104017624e-09,
+      "loss": 2.0898,
+      "step": 43100
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2976139503573237e-09,
+      "loss": 2.1309,
+      "step": 43105
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.099817275895809e-09,
+      "loss": 2.1241,
+      "step": 43110
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.183673193469134e-10,
+      "loss": 2.0856,
+      "step": 43115
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.532641103646931e-10,
+      "loss": 2.0962,
+      "step": 43120
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.045076759497725e-10,
+      "loss": 2.1257,
+      "step": 43125
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 4.720980404049336e-10,
+      "loss": 2.0845,
+      "step": 43130
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.5603522537952517e-10,
+      "loss": 2.1411,
+      "step": 43135
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.5631924984725884e-10,
+      "loss": 2.1441,
+      "step": 43140
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7295013011731088e-10,
+      "loss": 2.1098,
+      "step": 43145
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.0592787980101548e-10,
+      "loss": 2.1205,
+      "step": 43150
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.525250987847841e-11,
+      "loss": 2.1061,
+      "step": 43155
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.092402860975895e-11,
+      "loss": 2.124,
+      "step": 43160
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.9424416125856114e-12,
+      "loss": 2.0543,
+      "step": 43165
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 2.14786434173584,
+      "eval_runtime": 188.8048,
+      "eval_samples_per_second": 25.714,
+      "eval_steps_per_second": 3.215,
+      "step": 43168
+    },
+    {
+      "epoch": 4.0,
+      "step": 43168,
+      "total_flos": 1.2258755156891402e+18,
+      "train_loss": 2.0942285917837413,
+      "train_runtime": 51337.8051,
+      "train_samples_per_second": 6.727,
+      "train_steps_per_second": 0.841
     }
   ],
   "logging_steps": 5,
-  "max_steps": 10806,
+  "max_steps": 43168,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 4,
   "save_steps": 100,
-  "total_flos": 6.137045068818678e+17,
-  "train_batch_size": 8,
+  "total_flos": 1.2258755156891402e+18,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }