diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,3831 +3,15178 @@
   "best_model_checkpoint": null,
   "epoch": 2.0,
   "eval_steps": 500,
-  "global_step": 2702,
+  "global_step": 10806,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.0,
-      "grad_norm": 0.34765625,
-      "learning_rate": 1.4760147601476015e-06,
-      "loss": 2.6485,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.7002775208140615e-07,
+      "loss": 2.5996,
       "step": 1
     },
     {
       "epoch": 0.0,
-      "grad_norm": 0.36328125,
-      "learning_rate": 7.380073800738008e-06,
-      "loss": 2.6486,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8501387604070305e-06,
+      "loss": 2.6552,
       "step": 5
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.37890625,
-      "learning_rate": 1.4760147601476015e-05,
-      "loss": 2.6449,
+      "epoch": 0.0,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.700277520814061e-06,
+      "loss": 2.6426,
       "step": 10
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.400390625,
-      "learning_rate": 2.2140221402214025e-05,
-      "loss": 2.6688,
+      "epoch": 0.0,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.550416281221091e-06,
+      "loss": 2.6381,
       "step": 15
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 0.34375,
-      "learning_rate": 2.952029520295203e-05,
-      "loss": 2.6268,
+      "epoch": 0.0,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.400555041628122e-06,
+      "loss": 2.6589,
       "step": 20
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.271484375,
-      "learning_rate": 3.690036900369004e-05,
-      "loss": 2.6019,
+      "epoch": 0.0,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.250693802035152e-06,
+      "loss": 2.6584,
       "step": 25
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 0.25,
-      "learning_rate": 4.428044280442805e-05,
-      "loss": 2.5532,
+      "epoch": 0.01,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1100832562442182e-05,
+      "loss": 2.6289,
       "step": 30
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.2578125,
-      "learning_rate": 5.166051660516605e-05,
-      "loss": 2.508,
+      "epoch": 0.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.2950971322849214e-05,
+      "loss": 2.6038,
       "step": 35
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.19140625,
-      "learning_rate": 5.904059040590406e-05,
-      "loss": 2.4591,
+      "epoch": 0.01,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4801110083256244e-05,
+      "loss": 2.5957,
       "step": 40
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 0.1708984375,
-      "learning_rate": 6.642066420664207e-05,
-      "loss": 2.4235,
+      "epoch": 0.01,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6651248843663276e-05,
+      "loss": 2.63,
       "step": 45
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1787109375,
-      "learning_rate": 7.380073800738008e-05,
-      "loss": 2.3654,
+      "epoch": 0.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8501387604070305e-05,
+      "loss": 2.6358,
       "step": 50
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.1337890625,
-      "learning_rate": 8.118081180811809e-05,
-      "loss": 2.3604,
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.0351526364477336e-05,
+      "loss": 2.5969,
       "step": 55
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 0.142578125,
-      "learning_rate": 8.85608856088561e-05,
-      "loss": 2.3221,
+      "epoch": 0.01,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.2201665124884365e-05,
+      "loss": 2.5457,
       "step": 60
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.11767578125,
-      "learning_rate": 9.59409594095941e-05,
-      "loss": 2.3041,
+      "epoch": 0.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.40518038852914e-05,
+      "loss": 2.5137,
       "step": 65
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.0001033210332103321,
-      "loss": 2.3051,
+      "epoch": 0.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.590194264569843e-05,
+      "loss": 2.5039,
       "step": 70
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.00011070110701107013,
-      "loss": 2.2795,
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.7752081406105457e-05,
+      "loss": 2.4904,
       "step": 75
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00011808118081180812,
-      "loss": 2.2926,
+      "epoch": 0.01,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.960222016651249e-05,
+      "loss": 2.4536,
       "step": 80
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00012546125461254613,
-      "loss": 2.2594,
+      "epoch": 0.02,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.145235892691952e-05,
+      "loss": 2.4519,
       "step": 85
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00013284132841328414,
-      "loss": 2.2639,
+      "epoch": 0.02,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.330249768732655e-05,
+      "loss": 2.4403,
       "step": 90
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.00014022140221402215,
-      "loss": 2.2496,
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.5152636447733584e-05,
+      "loss": 2.4031,
       "step": 95
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00014760147601476016,
-      "loss": 2.2428,
+      "epoch": 0.02,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.700277520814061e-05,
+      "loss": 2.3909,
       "step": 100
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.00015498154981549817,
-      "loss": 2.2572,
+      "epoch": 0.02,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.885291396854764e-05,
+      "loss": 2.3601,
       "step": 105
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00016236162361623618,
-      "loss": 2.2447,
+      "epoch": 0.02,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.070305272895467e-05,
+      "loss": 2.3702,
       "step": 110
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.0001697416974169742,
-      "loss": 2.2258,
+      "epoch": 0.02,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 2.3449,
       "step": 115
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.0001771217712177122,
-      "loss": 2.2284,
+      "epoch": 0.02,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.440333024976873e-05,
+      "loss": 2.3317,
       "step": 120
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.0001845018450184502,
-      "loss": 2.2346,
+      "epoch": 0.02,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.625346901017577e-05,
+      "loss": 2.3406,
       "step": 125
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.0001918819188191882,
-      "loss": 2.2365,
+      "epoch": 0.02,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.81036077705828e-05,
+      "loss": 2.3098,
       "step": 130
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.00019926199261992622,
-      "loss": 2.2068,
+      "epoch": 0.02,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.9953746530989825e-05,
+      "loss": 2.319,
       "step": 135
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.0002066420664206642,
-      "loss": 2.2316,
+      "epoch": 0.03,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.180388529139686e-05,
+      "loss": 2.2914,
       "step": 140
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 0.00021402214022140222,
-      "loss": 2.2427,
+      "epoch": 0.03,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.365402405180389e-05,
+      "loss": 2.315,
       "step": 145
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00022140221402214025,
-      "loss": 2.2405,
+      "epoch": 0.03,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.5504162812210914e-05,
+      "loss": 2.3091,
       "step": 150
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.115234375,
-      "learning_rate": 0.00022878228782287826,
-      "loss": 2.2197,
+      "epoch": 0.03,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.735430157261795e-05,
+      "loss": 2.2459,
       "step": 155
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.10693359375,
-      "learning_rate": 0.00023616236162361624,
-      "loss": 2.2411,
+      "epoch": 0.03,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.920444033302498e-05,
+      "loss": 2.3152,
       "step": 160
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.1181640625,
-      "learning_rate": 0.00024354243542435425,
-      "loss": 2.2195,
+      "epoch": 0.03,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.105457909343201e-05,
+      "loss": 2.2955,
       "step": 165
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.10693359375,
-      "learning_rate": 0.00025092250922509226,
-      "loss": 2.2147,
+      "epoch": 0.03,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.290471785383904e-05,
+      "loss": 2.2946,
       "step": 170
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.00025830258302583027,
-      "loss": 2.2341,
+      "epoch": 0.03,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.475485661424607e-05,
+      "loss": 2.262,
       "step": 175
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002656826568265683,
-      "loss": 2.2116,
+      "epoch": 0.03,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.66049953746531e-05,
+      "loss": 2.2851,
       "step": 180
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002730627306273063,
-      "loss": 2.1973,
+      "epoch": 0.03,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.845513413506012e-05,
+      "loss": 2.235,
       "step": 185
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.10693359375,
-      "learning_rate": 0.0002804428044280443,
-      "loss": 2.2137,
+      "epoch": 0.04,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.030527289546717e-05,
+      "loss": 2.2502,
       "step": 190
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0002878228782287823,
-      "loss": 2.2388,
+      "epoch": 0.04,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.21554116558742e-05,
+      "loss": 2.2395,
       "step": 195
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.0002952029520295203,
-      "loss": 2.2111,
+      "epoch": 0.04,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.400555041628122e-05,
+      "loss": 2.2683,
       "step": 200
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00030258302583025833,
-      "loss": 2.2361,
+      "epoch": 0.04,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.585568917668826e-05,
+      "loss": 2.2791,
       "step": 205
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 0.00030996309963099634,
-      "loss": 2.1882,
+      "epoch": 0.04,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.770582793709528e-05,
+      "loss": 2.2839,
       "step": 210
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.00031734317343173435,
-      "loss": 2.2399,
+      "epoch": 0.04,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.955596669750231e-05,
+      "loss": 2.282,
       "step": 215
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.10546875,
-      "learning_rate": 0.00032472324723247236,
-      "loss": 2.2062,
+      "epoch": 0.04,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.140610545790935e-05,
+      "loss": 2.2102,
       "step": 220
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.10546875,
-      "learning_rate": 0.0003321033210332103,
-      "loss": 2.2068,
+      "epoch": 0.04,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.325624421831638e-05,
+      "loss": 2.2292,
       "step": 225
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.10498046875,
-      "learning_rate": 0.0003394833948339484,
-      "loss": 2.2058,
+      "epoch": 0.04,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.510638297872341e-05,
+      "loss": 2.2687,
       "step": 230
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 0.0003468634686346864,
-      "loss": 2.2137,
+      "epoch": 0.04,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.695652173913044e-05,
+      "loss": 2.2431,
       "step": 235
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0003542435424354244,
-      "loss": 2.2127,
+      "epoch": 0.04,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.880666049953746e-05,
+      "loss": 2.2333,
       "step": 240
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.09716796875,
-      "learning_rate": 0.00036162361623616235,
-      "loss": 2.2093,
+      "epoch": 0.05,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.06567992599445e-05,
+      "loss": 2.2432,
       "step": 245
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0003690036900369004,
-      "loss": 2.2063,
+      "epoch": 0.05,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.250693802035154e-05,
+      "loss": 2.2293,
       "step": 250
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003763837638376384,
-      "loss": 2.2347,
+      "epoch": 0.05,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.435707678075855e-05,
+      "loss": 2.28,
       "step": 255
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0003837638376383764,
-      "loss": 2.1907,
+      "epoch": 0.05,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.62072155411656e-05,
+      "loss": 2.2034,
       "step": 260
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003911439114391144,
-      "loss": 2.2137,
+      "epoch": 0.05,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.805735430157262e-05,
+      "loss": 2.2733,
       "step": 265
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00039852398523985245,
-      "loss": 2.1975,
+      "epoch": 0.05,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.990749306197965e-05,
+      "loss": 2.223,
       "step": 270
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.00039999732792377247,
-      "loss": 2.203,
+      "epoch": 0.05,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001017576318223867,
+      "loss": 2.2366,
       "step": 275
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00039998647273646793,
-      "loss": 2.2062,
+      "epoch": 0.05,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010360777058279371,
+      "loss": 2.2828,
       "step": 280
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00039996726788618994,
-      "loss": 2.2064,
+      "epoch": 0.05,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010545790934320075,
+      "loss": 2.2162,
       "step": 285
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.00039993971417476293,
-      "loss": 2.1835,
+      "epoch": 0.05,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00010730804810360778,
+      "loss": 2.229,
       "step": 290
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003999038127525856,
-      "loss": 2.1989,
+      "epoch": 0.05,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001091581868640148,
+      "loss": 2.2279,
       "step": 295
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.09716796875,
-      "learning_rate": 0.00039985956511858335,
-      "loss": 2.2137,
+      "epoch": 0.06,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00011100832562442183,
+      "loss": 2.2605,
       "step": 300
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.00039980697312014523,
-      "loss": 2.1926,
+      "epoch": 0.06,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00011285846438482887,
+      "loss": 2.2645,
       "step": 305
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.00039974603895304704,
-      "loss": 2.1829,
+      "epoch": 0.06,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001147086031452359,
+      "loss": 2.2575,
       "step": 310
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00039967676516135974,
-      "loss": 2.1946,
+      "epoch": 0.06,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011655874190564294,
+      "loss": 2.2618,
       "step": 315
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00039959915463734285,
-      "loss": 2.1814,
+      "epoch": 0.06,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00011840888066604995,
+      "loss": 2.2295,
       "step": 320
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00039951321062132425,
-      "loss": 2.187,
+      "epoch": 0.06,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012025901942645699,
+      "loss": 2.219,
       "step": 325
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00039941893670156453,
-      "loss": 2.2039,
+      "epoch": 0.06,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012210915818686402,
+      "loss": 2.2416,
       "step": 330
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.09423828125,
-      "learning_rate": 0.0003993163368141071,
-      "loss": 2.1839,
+      "epoch": 0.06,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00012395929694727104,
+      "loss": 2.1879,
       "step": 335
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.0003992054152426141,
-      "loss": 2.2013,
+      "epoch": 0.06,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00012580943570767808,
+      "loss": 2.2485,
       "step": 340
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.0003990861766181874,
-      "loss": 2.1924,
+      "epoch": 0.06,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00012765957446808513,
+      "loss": 2.2153,
       "step": 345
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.0003989586259191755,
-      "loss": 2.1955,
+      "epoch": 0.06,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00012950971322849215,
+      "loss": 2.2566,
       "step": 350
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.0003988227684709653,
-      "loss": 2.1965,
+      "epoch": 0.07,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013135985198889916,
+      "loss": 2.2282,
       "step": 355
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.00039867860994575994,
-      "loss": 2.1946,
+      "epoch": 0.07,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001332099907493062,
+      "loss": 2.2299,
       "step": 360
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.000398526156362342,
-      "loss": 2.1955,
+      "epoch": 0.07,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00013506012950971323,
+      "loss": 2.2566,
       "step": 365
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003983654140858221,
-      "loss": 2.1897,
+      "epoch": 0.07,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00013691026827012025,
+      "loss": 2.2137,
       "step": 370
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00039819638982737353,
-      "loss": 2.189,
+      "epoch": 0.07,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013876040703052732,
+      "loss": 2.2106,
       "step": 375
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.09423828125,
-      "learning_rate": 0.0003980190906439514,
-      "loss": 2.1812,
+      "epoch": 0.07,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00014061054579093434,
+      "loss": 2.2085,
       "step": 380
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00039783352393799856,
-      "loss": 2.188,
+      "epoch": 0.07,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00014246068455134135,
+      "loss": 2.195,
       "step": 385
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.0927734375,
-      "learning_rate": 0.00039763969745713635,
-      "loss": 2.1859,
+      "epoch": 0.07,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001443108233117484,
+      "loss": 2.2238,
       "step": 390
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.0003974376192938411,
-      "loss": 2.1697,
+      "epoch": 0.07,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00014616096207215542,
+      "loss": 2.2245,
       "step": 395
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003972272978851061,
-      "loss": 2.2024,
+      "epoch": 0.07,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00014801110083256244,
+      "loss": 2.2254,
       "step": 400
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.09375,
-      "learning_rate": 0.00039700874201208976,
-      "loss": 2.1922,
+      "epoch": 0.07,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00014986123959296948,
+      "loss": 2.252,
       "step": 405
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.00039678196079974865,
-      "loss": 2.1752,
+      "epoch": 0.08,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015171137835337653,
+      "loss": 2.2084,
       "step": 410
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.09375,
-      "learning_rate": 0.00039654696371645663,
-      "loss": 2.1947,
+      "epoch": 0.08,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00015356151711378355,
+      "loss": 2.2621,
       "step": 415
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.0003963037605736096,
-      "loss": 2.194,
+      "epoch": 0.08,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00015541165587419056,
+      "loss": 2.2136,
       "step": 420
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.09375,
-      "learning_rate": 0.0003960523615252156,
-      "loss": 2.2065,
+      "epoch": 0.08,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001572617946345976,
+      "loss": 2.2319,
       "step": 425
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.00039579277706747125,
-      "loss": 2.1707,
+      "epoch": 0.08,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00015911193339500463,
+      "loss": 2.2309,
       "step": 430
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.0947265625,
-      "learning_rate": 0.00039552501803832336,
-      "loss": 2.1889,
+      "epoch": 0.08,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00016096207215541165,
+      "loss": 2.2266,
       "step": 435
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.09521484375,
-      "learning_rate": 0.0003952490956170161,
-      "loss": 2.1876,
+      "epoch": 0.08,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001628122109158187,
+      "loss": 2.2047,
       "step": 440
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.09375,
-      "learning_rate": 0.00039496502132362494,
-      "loss": 2.1877,
+      "epoch": 0.08,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00016466234967622574,
+      "loss": 2.2206,
       "step": 445
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.000394672807018575,
-      "loss": 2.177,
+      "epoch": 0.08,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00016651248843663275,
+      "loss": 2.1969,
       "step": 450
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003943724649021464,
-      "loss": 2.2009,
+      "epoch": 0.08,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001683626271970398,
+      "loss": 2.219,
       "step": 455
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.09423828125,
-      "learning_rate": 0.00039406400751396445,
-      "loss": 2.1802,
+      "epoch": 0.09,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00017021276595744682,
+      "loss": 2.191,
       "step": 460
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0003937474477324764,
-      "loss": 2.1789,
+      "epoch": 0.09,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017206290471785384,
+      "loss": 2.2323,
       "step": 465
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.09521484375,
-      "learning_rate": 0.00039342279877441357,
-      "loss": 2.1829,
+      "epoch": 0.09,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00017391304347826088,
+      "loss": 2.2011,
       "step": 470
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00039309007419423964,
-      "loss": 2.1871,
+      "epoch": 0.09,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001757631822386679,
+      "loss": 2.1978,
       "step": 475
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.09716796875,
-      "learning_rate": 0.00039274928788358477,
-      "loss": 2.2183,
+      "epoch": 0.09,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017761332099907492,
+      "loss": 2.2146,
       "step": 480
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.00039240045407066556,
-      "loss": 2.1888,
+      "epoch": 0.09,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000179463459759482,
+      "loss": 2.2108,
       "step": 485
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 0.09423828125,
-      "learning_rate": 0.00039204358731969083,
-      "loss": 2.1962,
+      "epoch": 0.09,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000181313598519889,
+      "loss": 2.2392,
       "step": 490
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003916787025302538,
-      "loss": 2.1746,
+      "epoch": 0.09,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018316373728029603,
+      "loss": 2.231,
       "step": 495
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.0003913058149367101,
-      "loss": 2.1784,
+      "epoch": 0.09,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018501387604070307,
+      "loss": 2.19,
       "step": 500
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.0003909249401075413,
-      "loss": 2.1903,
+      "epoch": 0.09,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001868640148011101,
+      "loss": 2.2252,
       "step": 505
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.09423828125,
-      "learning_rate": 0.0003905360939447052,
-      "loss": 2.1772,
+      "epoch": 0.09,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001887141535615171,
+      "loss": 2.2244,
       "step": 510
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00039013929268297195,
-      "loss": 2.2023,
+      "epoch": 0.1,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019056429232192415,
+      "loss": 2.1966,
       "step": 515
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00038973455288924614,
-      "loss": 2.1887,
+      "epoch": 0.1,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001924144310823312,
+      "loss": 2.2371,
       "step": 520
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003893218914618749,
-      "loss": 2.1764,
+      "epoch": 0.1,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019426456984273822,
+      "loss": 2.1947,
       "step": 525
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00038890132562994286,
-      "loss": 2.1911,
+      "epoch": 0.1,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019611470860314524,
+      "loss": 2.2009,
       "step": 530
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003884728729525525,
-      "loss": 2.2008,
+      "epoch": 0.1,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019796484736355228,
+      "loss": 2.1746,
       "step": 535
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.0003880365513180908,
-      "loss": 2.1832,
+      "epoch": 0.1,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001998149861239593,
+      "loss": 2.1965,
       "step": 540
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.09619140625,
-      "learning_rate": 0.00038759237894348306,
-      "loss": 2.1775,
+      "epoch": 0.1,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00020166512488436634,
+      "loss": 2.1786,
       "step": 545
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003871403743734316,
-      "loss": 2.1768,
+      "epoch": 0.1,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002035152636447734,
+      "loss": 2.2412,
       "step": 550
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.0003866805564796421,
-      "loss": 2.1825,
+      "epoch": 0.1,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002053654024051804,
+      "loss": 2.2405,
       "step": 555
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 0.09521484375,
-      "learning_rate": 0.000386212944460035,
-      "loss": 2.191,
+      "epoch": 0.1,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020721554116558743,
+      "loss": 2.2099,
       "step": 560
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.0003857375578379449,
-      "loss": 2.168,
+      "epoch": 0.1,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020906567992599447,
+      "loss": 2.2487,
       "step": 565
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003852544164613043,
-      "loss": 2.1844,
+      "epoch": 0.11,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002109158186864015,
+      "loss": 2.2261,
       "step": 570
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003847635405018162,
-      "loss": 2.1926,
+      "epoch": 0.11,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002127659574468085,
+      "loss": 2.2253,
       "step": 575
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00038426495045411064,
-      "loss": 2.1935,
+      "epoch": 0.11,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00021461609620721555,
+      "loss": 2.2131,
       "step": 580
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.0003837586671348901,
-      "loss": 2.1654,
+      "epoch": 0.11,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00021646623496762257,
+      "loss": 2.2409,
       "step": 585
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00038324471168205945,
-      "loss": 2.1816,
+      "epoch": 0.11,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002183163737280296,
+      "loss": 2.2165,
       "step": 590
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.0003827231055538443,
-      "loss": 2.1697,
+      "epoch": 0.11,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00022016651248843664,
+      "loss": 2.2068,
       "step": 595
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.0003821938705278944,
-      "loss": 2.1581,
+      "epoch": 0.11,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00022201665124884365,
+      "loss": 2.2435,
       "step": 600
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.00038165702870037485,
-      "loss": 2.1588,
+      "epoch": 0.11,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00022386679000925073,
+      "loss": 2.1959,
       "step": 605
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.0003811126024850432,
-      "loss": 2.1842,
+      "epoch": 0.11,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00022571692876965774,
+      "loss": 2.2377,
       "step": 610
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.0003805606146123139,
-      "loss": 2.169,
+      "epoch": 0.11,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002275670675300648,
+      "loss": 2.2028,
       "step": 615
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003800010881283093,
-      "loss": 2.17,
+      "epoch": 0.11,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002294172062904718,
+      "loss": 2.1899,
       "step": 620
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003794340463938972,
-      "loss": 2.1613,
+      "epoch": 0.12,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00023126734505087883,
+      "loss": 2.2237,
       "step": 625
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.0003788595130837157,
-      "loss": 2.1806,
+      "epoch": 0.12,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00023311748381128587,
+      "loss": 2.2715,
       "step": 630
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00037827751218518494,
-      "loss": 2.1841,
+      "epoch": 0.12,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002349676225716929,
+      "loss": 2.2158,
       "step": 635
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.000377688067997505,
-      "loss": 2.1827,
+      "epoch": 0.12,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002368177613320999,
+      "loss": 2.2045,
       "step": 640
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.000377091205130642,
-      "loss": 2.1846,
+      "epoch": 0.12,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00023866790009250695,
+      "loss": 2.19,
       "step": 645
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003764869485043003,
-      "loss": 2.1877,
+      "epoch": 0.12,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00024051803885291397,
+      "loss": 2.175,
       "step": 650
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003758753233468823,
-      "loss": 2.1702,
+      "epoch": 0.12,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.000242368177613321,
+      "loss": 2.2398,
       "step": 655
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.0966796875,
-      "learning_rate": 0.00037525635519443466,
-      "loss": 2.1755,
+      "epoch": 0.12,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00024421831637372804,
+      "loss": 2.2246,
       "step": 660
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00037463006988958266,
-      "loss": 2.1691,
+      "epoch": 0.12,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002460684551341351,
+      "loss": 2.2097,
       "step": 665
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003739964935804509,
-      "loss": 2.1713,
+      "epoch": 0.12,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00024791859389454207,
+      "loss": 2.1865,
       "step": 670
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.095703125,
-      "learning_rate": 0.0003733556527195719,
-      "loss": 2.1749,
+      "epoch": 0.12,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002497687326549491,
+      "loss": 2.1929,
       "step": 675
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00037270757406278126,
-      "loss": 2.1685,
+      "epoch": 0.13,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00025161887141535616,
+      "loss": 2.223,
       "step": 680
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00037205228466810094,
-      "loss": 2.1818,
+      "epoch": 0.13,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002534690101757632,
+      "loss": 2.1875,
       "step": 685
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.00037138981189460945,
-      "loss": 2.1661,
+      "epoch": 0.13,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00025531914893617025,
+      "loss": 2.2754,
       "step": 690
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.00037072018340129936,
-      "loss": 2.1733,
+      "epoch": 0.13,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00025716928769657724,
+      "loss": 2.2107,
       "step": 695
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003700434271459229,
-      "loss": 2.1749,
+      "epoch": 0.13,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002590194264569843,
+      "loss": 2.2189,
       "step": 700
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003693595713838243,
-      "loss": 2.1902,
+      "epoch": 0.13,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00026086956521739134,
+      "loss": 2.2028,
       "step": 705
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.0003686686446667605,
-      "loss": 2.1807,
+      "epoch": 0.13,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002627197039777983,
+      "loss": 2.2138,
       "step": 710
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.0003679706758417087,
-      "loss": 2.1834,
+      "epoch": 0.13,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00026456984273820537,
+      "loss": 2.2098,
       "step": 715
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.0003672656940496621,
-      "loss": 2.1786,
+      "epoch": 0.13,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002664199814986124,
+      "loss": 2.1761,
       "step": 720
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.0003665537287244134,
-      "loss": 2.1742,
+      "epoch": 0.13,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002682701202590194,
+      "loss": 2.1912,
       "step": 725
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00036583480959132567,
-      "loss": 2.1676,
+      "epoch": 0.14,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00027012025901942645,
+      "loss": 2.1687,
       "step": 730
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003651089666660914,
-      "loss": 2.1633,
+      "epoch": 0.14,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002719703977798335,
+      "loss": 2.2049,
       "step": 735
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.0003643762302534792,
-      "loss": 2.1699,
+      "epoch": 0.14,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002738205365402405,
+      "loss": 2.1749,
       "step": 740
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003636366309460688,
-      "loss": 2.1772,
+      "epoch": 0.14,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00027567067530064754,
+      "loss": 2.1944,
       "step": 745
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00036289019962297347,
-      "loss": 2.1602,
+      "epoch": 0.14,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00027752081406105464,
+      "loss": 2.206,
       "step": 750
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.000362136967448551,
-      "loss": 2.1592,
+      "epoch": 0.14,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002793709528214616,
+      "loss": 2.1873,
       "step": 755
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00036137696587110234,
-      "loss": 2.1497,
+      "epoch": 0.14,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00028122109158186867,
+      "loss": 2.2215,
       "step": 760
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003606102266215589,
-      "loss": 2.1816,
+      "epoch": 0.14,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002830712303422757,
+      "loss": 2.2269,
       "step": 765
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003598367817121574,
-      "loss": 2.1681,
+      "epoch": 0.14,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002849213691026827,
+      "loss": 2.2421,
       "step": 770
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 0.0986328125,
-      "learning_rate": 0.0003590566634351036,
-      "loss": 2.179,
+      "epoch": 0.14,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00028677150786308975,
+      "loss": 2.2153,
       "step": 775
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.000358269904361224,
-      "loss": 2.186,
+      "epoch": 0.14,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002886216466234968,
+      "loss": 2.2253,
       "step": 780
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00035747653733860576,
-      "loss": 2.1635,
+      "epoch": 0.15,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002904717853839038,
+      "loss": 2.1929,
       "step": 785
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003566765954912256,
-      "loss": 2.1863,
+      "epoch": 0.15,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00029232192414431084,
+      "loss": 2.2065,
       "step": 790
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.0003558701122175666,
-      "loss": 2.1639,
+      "epoch": 0.15,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002941720629047179,
+      "loss": 2.2065,
       "step": 795
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0003550571211892238,
-      "loss": 2.1795,
+      "epoch": 0.15,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00029602220166512487,
+      "loss": 2.1999,
       "step": 800
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.00035423765634949844,
-      "loss": 2.1634,
+      "epoch": 0.15,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002978723404255319,
+      "loss": 2.2407,
       "step": 805
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.0003534117519119807,
-      "loss": 2.1673,
+      "epoch": 0.15,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00029972247918593896,
+      "loss": 2.2366,
       "step": 810
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003525794423591214,
-      "loss": 2.1722,
+      "epoch": 0.15,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00030157261794634595,
+      "loss": 2.2105,
       "step": 815
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00035174076244079216,
-      "loss": 2.1687,
+      "epoch": 0.15,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00030342275670675305,
+      "loss": 2.2145,
       "step": 820
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.10498046875,
-      "learning_rate": 0.00035089574717283466,
-      "loss": 2.1736,
+      "epoch": 0.15,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00030527289546716004,
+      "loss": 2.1843,
       "step": 825
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003500444318355986,
-      "loss": 2.1784,
+      "epoch": 0.15,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003071230342275671,
+      "loss": 2.1932,
       "step": 830
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003491868519724688,
-      "loss": 2.1783,
+      "epoch": 0.15,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00030897317298797414,
+      "loss": 2.1702,
       "step": 835
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.000348323043388381,
-      "loss": 2.1674,
+      "epoch": 0.16,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0003108233117483811,
+      "loss": 2.1664,
       "step": 840
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.00034745304214832726,
-      "loss": 2.1791,
+      "epoch": 0.16,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00031267345050878817,
+      "loss": 2.2214,
       "step": 845
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.0003465768845758502,
-      "loss": 2.1573,
+      "epoch": 0.16,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0003145235892691952,
+      "loss": 2.2186,
       "step": 850
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00034569460725152615,
-      "loss": 2.1842,
+      "epoch": 0.16,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0003163737280296022,
+      "loss": 2.2511,
       "step": 855
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.09765625,
-      "learning_rate": 0.00034480624701143807,
-      "loss": 2.1624,
+      "epoch": 0.16,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00031822386679000925,
+      "loss": 2.225,
       "step": 860
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00034391184094563764,
-      "loss": 2.1892,
+      "epoch": 0.16,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0003200740055504163,
+      "loss": 2.1627,
       "step": 865
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 0.09814453125,
-      "learning_rate": 0.00034301142639659663,
-      "loss": 2.1788,
+      "epoch": 0.16,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003219241443108233,
+      "loss": 2.1866,
       "step": 870
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003421050409576478,
-      "loss": 2.174,
+      "epoch": 0.16,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00032377428307123034,
+      "loss": 2.2332,
       "step": 875
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00034119272247141536,
-      "loss": 2.171,
+      "epoch": 0.16,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0003256244218316374,
+      "loss": 2.2043,
       "step": 880
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0003402745090282351,
-      "loss": 2.1571,
+      "epoch": 0.16,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00032747456059204437,
+      "loss": 2.1818,
       "step": 885
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.10498046875,
-      "learning_rate": 0.00033935043896456384,
-      "loss": 2.1716,
+      "epoch": 0.16,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00032932469935245147,
+      "loss": 2.1872,
       "step": 890
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.000338420550861379,
-      "loss": 2.1631,
+      "epoch": 0.17,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003311748381128585,
+      "loss": 2.1946,
       "step": 895
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.0003374848835425679,
-      "loss": 2.1652,
+      "epoch": 0.17,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003330249768732655,
+      "loss": 2.2237,
       "step": 900
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00033654347607330656,
-      "loss": 2.1661,
-      "step": 905
+      "epoch": 0.17,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00033487511563367255,
+      "loss": 2.2096,
+      "step": 905
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0003355963677584288,
-      "loss": 2.1697,
+      "epoch": 0.17,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003367252543940796,
+      "loss": 2.1838,
       "step": 910
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00033464359814078536,
-      "loss": 2.1937,
+      "epoch": 0.17,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003385753931544866,
+      "loss": 2.2126,
       "step": 915
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0003336852069995927,
-      "loss": 2.1646,
+      "epoch": 0.17,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00034042553191489364,
+      "loss": 2.1749,
       "step": 920
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003327212343487725,
-      "loss": 2.1635,
+      "epoch": 0.17,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0003422756706753007,
+      "loss": 2.1816,
       "step": 925
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.0003317517204352804,
-      "loss": 2.149,
+      "epoch": 0.17,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00034412580943570767,
+      "loss": 2.2064,
       "step": 930
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0003307767057374266,
-      "loss": 2.1678,
+      "epoch": 0.17,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003459759481961147,
+      "loss": 2.1977,
       "step": 935
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.00032979623096318487,
-      "loss": 2.1643,
+      "epoch": 0.17,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00034782608695652176,
+      "loss": 2.2267,
       "step": 940
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00032881033704849357,
-      "loss": 2.1748,
+      "epoch": 0.17,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00034967622571692875,
+      "loss": 2.1538,
       "step": 945
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00032781906515554646,
-      "loss": 2.1617,
+      "epoch": 0.18,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003515263644773358,
+      "loss": 2.2294,
       "step": 950
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0003268224566710738,
-      "loss": 2.1529,
+      "epoch": 0.18,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00035337650323774284,
+      "loss": 2.2237,
       "step": 955
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00032582055320461465,
-      "loss": 2.1781,
+      "epoch": 0.18,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00035522664199814984,
+      "loss": 2.2067,
       "step": 960
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 0.10546875,
-      "learning_rate": 0.0003248133965867798,
-      "loss": 2.1429,
+      "epoch": 0.18,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00035707678075855694,
+      "loss": 2.18,
       "step": 965
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.00032380102886750493,
-      "loss": 2.1629,
+      "epoch": 0.18,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000358926919518964,
+      "loss": 2.2009,
       "step": 970
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.000322783492314295,
-      "loss": 2.1655,
+      "epoch": 0.18,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00036077705827937097,
+      "loss": 2.2513,
       "step": 975
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0003217608294104601,
-      "loss": 2.1758,
+      "epoch": 0.18,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000362627197039778,
+      "loss": 2.1711,
       "step": 980
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.09912109375,
-      "learning_rate": 0.00032073308285334085,
-      "loss": 2.1565,
+      "epoch": 0.18,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00036447733580018506,
+      "loss": 2.2227,
       "step": 985
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.0003197002955525264,
-      "loss": 2.1806,
+      "epoch": 0.18,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00036632747456059205,
+      "loss": 2.181,
       "step": 990
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.099609375,
-      "learning_rate": 0.00031866251062806267,
-      "loss": 2.1701,
+      "epoch": 0.18,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003681776133209991,
+      "loss": 2.2341,
       "step": 995
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00031761977140865207,
-      "loss": 2.1675,
+      "epoch": 0.19,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00037002775208140614,
+      "loss": 2.1537,
       "step": 1000
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.0003165721214298444,
-      "loss": 2.1671,
+      "epoch": 0.19,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00037187789084181314,
+      "loss": 2.2413,
       "step": 1005
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0003155196044322193,
-      "loss": 2.1575,
+      "epoch": 0.19,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003737280296022202,
+      "loss": 2.2119,
       "step": 1010
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.00031446226435956,
-      "loss": 2.1381,
+      "epoch": 0.19,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003755781683626272,
+      "loss": 2.2427,
       "step": 1015
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0003134001453570186,
-      "loss": 2.1626,
+      "epoch": 0.19,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0003774283071230342,
+      "loss": 2.2112,
       "step": 1020
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00031233329176927295,
-      "loss": 2.1648,
+      "epoch": 0.19,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00037927844588344126,
+      "loss": 2.1871,
       "step": 1025
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0003112617481386752,
-      "loss": 2.1671,
+      "epoch": 0.19,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003811285846438483,
+      "loss": 2.1889,
       "step": 1030
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0003101855592033922,
-      "loss": 2.1648,
+      "epoch": 0.19,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00038297872340425535,
+      "loss": 2.1728,
       "step": 1035
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.0003091047698955375,
-      "loss": 2.1746,
+      "epoch": 0.19,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003848288621646624,
+      "loss": 2.1774,
       "step": 1040
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.00030801942533929545,
-      "loss": 2.1603,
+      "epoch": 0.19,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0003866790009250694,
+      "loss": 2.2112,
       "step": 1045
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00030692957084903726,
-      "loss": 2.172,
+      "epoch": 0.19,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00038852913968547644,
+      "loss": 2.2181,
       "step": 1050
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00030583525192742897,
-      "loss": 2.1643,
+      "epoch": 0.2,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003903792784458835,
+      "loss": 2.1747,
       "step": 1055
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 0.1005859375,
-      "learning_rate": 0.00030473651426353167,
-      "loss": 2.1756,
+      "epoch": 0.2,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00039222941720629047,
+      "loss": 2.2193,
       "step": 1060
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00030363340373089413,
-      "loss": 2.1637,
+      "epoch": 0.2,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003940795559666975,
+      "loss": 2.1631,
       "step": 1065
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.00030252596638563714,
-      "loss": 2.1718,
+      "epoch": 0.2,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00039592969472710456,
+      "loss": 2.1777,
       "step": 1070
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.000301414248464531,
-      "loss": 2.1552,
+      "epoch": 0.2,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00039777983348751155,
+      "loss": 2.2187,
       "step": 1075
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.000300298296383065,
-      "loss": 2.187,
+      "epoch": 0.2,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003996299722479186,
+      "loss": 2.1994,
       "step": 1080
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00029917815673350935,
-      "loss": 2.1747,
+      "epoch": 0.2,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003999998330292309,
+      "loss": 2.226,
       "step": 1085
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0002980538762829698,
-      "loss": 2.1677,
+      "epoch": 0.2,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039999915471095927,
+      "loss": 2.2093,
       "step": 1090
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00029692550197143563,
-      "loss": 2.1476,
+      "epoch": 0.2,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003999979546112803,
+      "loss": 2.1598,
       "step": 1095
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.00029579308090981913,
-      "loss": 2.163,
+      "epoch": 0.2,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039999623273332497,
+      "loss": 2.1882,
       "step": 1100
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00029465666037798935,
-      "loss": 2.1673,
+      "epoch": 0.2,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039999398908158553,
+      "loss": 2.1787,
       "step": 1105
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002935162878227975,
-      "loss": 2.1737,
+      "epoch": 0.21,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00039999122366191547,
+      "loss": 2.2214,
       "step": 1110
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0002923720108560964,
-      "loss": 2.1412,
+      "epoch": 0.21,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039998793648152956,
+      "loss": 2.2272,
       "step": 1115
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.00029122387725275244,
-      "loss": 2.1455,
+      "epoch": 0.21,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003999841275490038,
+      "loss": 2.1692,
       "step": 1120
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00029007193494865103,
-      "loss": 2.1477,
+      "epoch": 0.21,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039997979687427535,
+      "loss": 2.1905,
       "step": 1125
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00028891623203869523,
-      "loss": 2.1486,
+      "epoch": 0.21,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039997494446864255,
+      "loss": 2.2032,
       "step": 1130
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0002877568167747975,
-      "loss": 2.1703,
+      "epoch": 0.21,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00039996957034476503,
+      "loss": 2.203,
       "step": 1135
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.0002865937375638654,
-      "loss": 2.1819,
+      "epoch": 0.21,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039996367451666337,
+      "loss": 2.1985,
       "step": 1140
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.0002854270429657805,
-      "loss": 2.1564,
+      "epoch": 0.21,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039995725699971935,
+      "loss": 2.1622,
       "step": 1145
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0002842567816913708,
-      "loss": 2.1723,
+      "epoch": 0.21,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00039995031781067566,
+      "loss": 2.1439,
       "step": 1150
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 0.10498046875,
-      "learning_rate": 0.0002830830026003773,
-      "loss": 2.1585,
+      "epoch": 0.21,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039994285696763615,
+      "loss": 2.2177,
       "step": 1155
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.0002819057546994135,
-      "loss": 2.1758,
+      "epoch": 0.21,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003999348744900655,
+      "loss": 2.183,
       "step": 1160
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.00028072508713992007,
-      "loss": 2.1637,
+      "epoch": 0.22,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003999263703987894,
+      "loss": 2.1907,
       "step": 1165
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00027954104921611194,
-      "loss": 2.1478,
+      "epoch": 0.22,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003999173447159943,
+      "loss": 2.1811,
       "step": 1170
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.00027835369036292087,
-      "loss": 2.1585,
+      "epoch": 0.22,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00039990779746522744,
+      "loss": 2.2091,
       "step": 1175
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.000277163060153931,
-      "loss": 2.1481,
+      "epoch": 0.22,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00039989772867139685,
+      "loss": 2.1883,
       "step": 1180
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0002759692082993095,
-      "loss": 2.1636,
+      "epoch": 0.22,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039988713836077117,
+      "loss": 2.1967,
       "step": 1185
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00027477218464373076,
-      "loss": 2.1594,
+      "epoch": 0.22,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039987602656097967,
+      "loss": 2.2445,
       "step": 1190
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.0002735720391642956,
-      "loss": 2.153,
+      "epoch": 0.22,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039986439330101203,
+      "loss": 2.1954,
       "step": 1195
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.0002723688219684443,
-      "loss": 2.1666,
+      "epoch": 0.22,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039985223861121864,
+      "loss": 2.1925,
       "step": 1200
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.00027116258329186514,
-      "loss": 2.1728,
+      "epoch": 0.22,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00039983956252331,
+      "loss": 2.1998,
       "step": 1205
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002699533734963964,
-      "loss": 2.158,
+      "epoch": 0.22,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003998263650703569,
+      "loss": 2.1904,
       "step": 1210
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.10546875,
-      "learning_rate": 0.000268741243067924,
-      "loss": 2.1384,
+      "epoch": 0.22,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003998126462867906,
+      "loss": 2.1962,
       "step": 1215
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 0.10009765625,
-      "learning_rate": 0.00026752624261427375,
-      "loss": 2.1691,
+      "epoch": 0.23,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039979840620840213,
+      "loss": 2.1577,
       "step": 1220
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0002663084228630982,
-      "loss": 2.1531,
+      "epoch": 0.23,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003997836448723427,
+      "loss": 2.1957,
       "step": 1225
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.0002650878346597586,
-      "loss": 2.1576,
+      "epoch": 0.23,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0003997683623171236,
+      "loss": 2.2043,
       "step": 1230
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.0002638645289652025,
-      "loss": 2.1653,
+      "epoch": 0.23,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003997525585826155,
+      "loss": 2.1525,
       "step": 1235
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.0002626385568538358,
-      "loss": 2.15,
+      "epoch": 0.23,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039973623371004923,
+      "loss": 2.1551,
       "step": 1240
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002614099695113901,
-      "loss": 2.1738,
+      "epoch": 0.23,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.000399719387742015,
+      "loss": 2.1806,
       "step": 1245
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00026017881823278607,
-      "loss": 2.1608,
+      "epoch": 0.23,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039970202072246263,
+      "loss": 2.1651,
       "step": 1250
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00025894515441999156,
-      "loss": 2.1914,
+      "epoch": 0.23,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003996841326967011,
+      "loss": 2.1923,
       "step": 1255
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.00025770902957987556,
-      "loss": 2.1659,
+      "epoch": 0.23,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039966572371139886,
+      "loss": 2.2101,
       "step": 1260
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0002564704953220578,
-      "loss": 2.1719,
+      "epoch": 0.23,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003996467938145835,
+      "loss": 2.1676,
       "step": 1265
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002552296033567541,
-      "loss": 2.1676,
+      "epoch": 0.24,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003996273430556415,
+      "loss": 2.1938,
       "step": 1270
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002539864054926169,
-      "loss": 2.165,
+      "epoch": 0.24,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003996073714853184,
+      "loss": 2.1765,
       "step": 1275
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.0002527409536345728,
-      "loss": 2.1561,
+      "epoch": 0.24,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039958687915571823,
+      "loss": 2.1636,
       "step": 1280
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00025149329978165516,
-      "loss": 2.1635,
+      "epoch": 0.24,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000399565866120304,
+      "loss": 2.208,
       "step": 1285
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 0.1044921875,
-      "learning_rate": 0.0002502434960248331,
-      "loss": 2.1513,
+      "epoch": 0.24,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039954433243389676,
+      "loss": 2.1786,
       "step": 1290
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.10205078125,
-      "learning_rate": 0.00024899159454483665,
-      "loss": 2.1599,
+      "epoch": 0.24,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003995222781526764,
+      "loss": 2.1826,
       "step": 1295
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.0002477376476099784,
-      "loss": 2.1674,
+      "epoch": 0.24,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003994997033341805,
+      "loss": 2.1526,
       "step": 1300
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00024648170757397055,
-      "loss": 2.14,
+      "epoch": 0.24,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.000399476608037305,
+      "loss": 2.1645,
       "step": 1305
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00024522382687374,
-      "loss": 2.1519,
+      "epoch": 0.24,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003994529923223036,
+      "loss": 2.2188,
       "step": 1310
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.0002439640580272384,
-      "loss": 2.1622,
+      "epoch": 0.24,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003994288562507879,
+      "loss": 2.201,
       "step": 1315
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1015625,
-      "learning_rate": 0.0002427024536312496,
-      "loss": 2.1535,
+      "epoch": 0.24,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039940419988572685,
+      "loss": 2.2056,
       "step": 1320
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.10302734375,
-      "learning_rate": 0.00024143906635919383,
-      "loss": 2.1383,
+      "epoch": 0.25,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003993790232914468,
+      "loss": 2.1827,
       "step": 1325
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 0.1025390625,
-      "learning_rate": 0.00024017394895892838,
-      "loss": 2.1282,
+      "epoch": 0.25,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039935332653363154,
+      "loss": 2.1834,
       "step": 1330
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.00023890715425054545,
-      "loss": 2.1554,
+      "epoch": 0.25,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039932710967932177,
+      "loss": 2.1927,
       "step": 1335
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 0.10400390625,
-      "learning_rate": 0.0002376387351241666,
-      "loss": 2.157,
+      "epoch": 0.25,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003993003727969151,
+      "loss": 2.1535,
       "step": 1340
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.10107421875,
-      "learning_rate": 0.00023636874453773475,
-      "loss": 2.1471,
+      "epoch": 0.25,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00039927311595616585,
+      "loss": 2.1653,
       "step": 1345
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.10546875,
-      "learning_rate": 0.00023509723551480325,
-      "loss": 2.1383,
+      "epoch": 0.25,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039924533922818493,
+      "loss": 2.2489,
       "step": 1350
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 2.1541337966918945,
-      "eval_runtime": 188.5948,
-      "eval_samples_per_second": 25.785,
-      "eval_steps_per_second": 3.224,
-      "step": 1351
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.103515625,
-      "learning_rate": 0.00023382426114232162,
-      "loss": 2.1347,
+      "epoch": 0.25,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039921704268543954,
+      "loss": 2.1557,
       "step": 1355
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.00023254987456841956,
-      "loss": 2.1403,
+      "epoch": 0.25,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000399188226401753,
+      "loss": 2.2094,
       "step": 1360
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.107421875,
-      "learning_rate": 0.00023127412900018782,
-      "loss": 2.1468,
+      "epoch": 0.25,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003991588904523047,
+      "loss": 2.1912,
       "step": 1365
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 0.10498046875,
-      "learning_rate": 0.00022999707770145653,
-      "loss": 2.1267,
+      "epoch": 0.25,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0003991290349136297,
+      "loss": 2.1823,
       "step": 1370
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.1064453125,
-      "learning_rate": 0.0002287187739905717,
-      "loss": 2.1406,
+      "epoch": 0.25,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003990986598636186,
+      "loss": 2.1842,
       "step": 1375
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 0.10595703125,
-      "learning_rate": 0.00022743927123816899,
-      "loss": 2.1304,
+      "epoch": 0.26,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039906776538151746,
+      "loss": 2.1882,
       "step": 1380
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 0.00022615862286494537,
-      "loss": 2.1199,
+      "epoch": 0.26,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039903635154792745,
+      "loss": 2.1994,
       "step": 1385
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.00022487688233942862,
-      "loss": 2.1216,
+      "epoch": 0.26,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003990044184448047,
+      "loss": 2.1443,
       "step": 1390
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 0.10693359375,
-      "learning_rate": 0.00022359410317574548,
-      "loss": 2.1188,
+      "epoch": 0.26,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003989719661554601,
+      "loss": 2.1881,
       "step": 1395
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 0.00022231033893138668,
-      "loss": 2.123,
+      "epoch": 0.26,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039893899476455886,
+      "loss": 2.2267,
       "step": 1400
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.1083984375,
-      "learning_rate": 0.0002210256432049714,
-      "loss": 2.1401,
+      "epoch": 0.26,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039890550435812077,
+      "loss": 2.184,
       "step": 1405
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.0002197400696340091,
-      "loss": 2.1212,
+      "epoch": 0.26,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039887149502351954,
+      "loss": 2.2021,
       "step": 1410
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 0.00021845367189266042,
-      "loss": 2.1297,
+      "epoch": 0.26,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003988369668494827,
+      "loss": 2.189,
       "step": 1415
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.107421875,
-      "learning_rate": 0.0002171665036894959,
-      "loss": 2.1237,
+      "epoch": 0.26,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003988019199260915,
+      "loss": 2.1912,
       "step": 1420
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00021587861876525377,
-      "loss": 2.1281,
+      "epoch": 0.26,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039876635434478046,
+      "loss": 2.1916,
       "step": 1425
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.1083984375,
-      "learning_rate": 0.00021459007089059625,
-      "loss": 2.0898,
+      "epoch": 0.26,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039873027019833725,
+      "loss": 2.177,
       "step": 1430
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.0002133009138638645,
-      "loss": 2.1212,
+      "epoch": 0.27,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003986936675809024,
+      "loss": 2.2042,
       "step": 1435
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.10693359375,
-      "learning_rate": 0.00021201120150883234,
-      "loss": 2.1285,
+      "epoch": 0.27,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003986565465879692,
+      "loss": 2.1839,
       "step": 1440
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.00021072098767245932,
-      "loss": 2.1302,
+      "epoch": 0.27,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0003986189073163833,
+      "loss": 2.187,
       "step": 1445
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 0.10888671875,
-      "learning_rate": 0.00020943032622264238,
-      "loss": 2.1079,
+      "epoch": 0.27,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003985807498643424,
+      "loss": 2.1903,
       "step": 1450
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00020813927104596666,
-      "loss": 2.1353,
+      "epoch": 0.27,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039854207433139623,
+      "loss": 2.2007,
       "step": 1455
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 0.0002068478760454562,
-      "loss": 2.1402,
+      "epoch": 0.27,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039850288081844614,
+      "loss": 2.1817,
       "step": 1460
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 0.1064453125,
-      "learning_rate": 0.0002055561951383227,
-      "loss": 2.1276,
+      "epoch": 0.27,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003984631694277446,
+      "loss": 2.1768,
       "step": 1465
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.00020426428225371496,
-      "loss": 2.1241,
+      "epoch": 0.27,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003984229402628956,
+      "loss": 2.1591,
       "step": 1470
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 0.10791015625,
-      "learning_rate": 0.00020297219133046714,
-      "loss": 2.1255,
+      "epoch": 0.27,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003983821934288535,
+      "loss": 2.2008,
       "step": 1475
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.0002016799763148467,
-      "loss": 2.1191,
+      "epoch": 0.27,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003983409290319237,
+      "loss": 2.202,
       "step": 1480
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.1083984375,
-      "learning_rate": 0.00020038769115830198,
-      "loss": 2.1308,
+      "epoch": 0.27,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039829914717976156,
+      "loss": 2.1906,
       "step": 1485
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00019909538981521002,
-      "loss": 2.1422,
+      "epoch": 0.28,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003982568479813724,
+      "loss": 2.1852,
       "step": 1490
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.1083984375,
-      "learning_rate": 0.00019780312624062326,
-      "loss": 2.1358,
+      "epoch": 0.28,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039821403154711136,
+      "loss": 2.1686,
       "step": 1495
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.00019651095438801775,
-      "loss": 2.1292,
+      "epoch": 0.28,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0003981706979886831,
+      "loss": 2.1896,
       "step": 1500
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.00019521892820703975,
-      "loss": 2.1308,
+      "epoch": 0.28,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039812684741914137,
+      "loss": 2.1638,
       "step": 1505
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.0001939271016412536,
-      "loss": 2.1339,
+      "epoch": 0.28,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003980824799528886,
+      "loss": 2.1888,
       "step": 1510
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00019263552862588948,
-      "loss": 2.126,
+      "epoch": 0.28,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003980375957056759,
+      "loss": 2.1733,
       "step": 1515
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.1103515625,
-      "learning_rate": 0.00019134426308559162,
-      "loss": 2.1325,
+      "epoch": 0.28,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0003979921947946026,
+      "loss": 2.1795,
       "step": 1520
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 0.00019005335893216665,
-      "loss": 2.1302,
+      "epoch": 0.28,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039794627733811595,
+      "loss": 2.1668,
       "step": 1525
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.0001887628700623332,
-      "loss": 2.125,
+      "epoch": 0.28,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003978998434560109,
+      "loss": 2.2003,
       "step": 1530
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.000187472850355471,
-      "loss": 2.1217,
+      "epoch": 0.28,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039785289326942953,
+      "loss": 2.18,
       "step": 1535
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.1103515625,
-      "learning_rate": 0.00018618335367137195,
-      "loss": 2.147,
+      "epoch": 0.29,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003978054269008611,
+      "loss": 2.1842,
       "step": 1540
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.0001848944338479909,
-      "loss": 2.1437,
+      "epoch": 0.29,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0003977574444741415,
+      "loss": 2.1539,
       "step": 1545
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 0.00018360614469919835,
-      "loss": 2.117,
+      "epoch": 0.29,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039770894611445293,
+      "loss": 2.182,
       "step": 1550
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 0.0001823185400125333,
-      "loss": 2.1119,
+      "epoch": 0.29,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003976599319483236,
+      "loss": 2.1939,
       "step": 1555
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.00018103167354695756,
-      "loss": 2.1242,
+      "epoch": 0.29,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003976104021036275,
+      "loss": 2.191,
       "step": 1560
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00017974559903061172,
-      "loss": 2.115,
+      "epoch": 0.29,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039756035670958384,
+      "loss": 2.1724,
       "step": 1565
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 0.00017846037015857127,
-      "loss": 2.1332,
+      "epoch": 0.29,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003975097958967571,
+      "loss": 2.1847,
       "step": 1570
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00017717604059060518,
-      "loss": 2.1049,
+      "epoch": 0.29,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00039745871979705616,
+      "loss": 2.1144,
       "step": 1575
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.109375,
-      "learning_rate": 0.0001758926639489354,
-      "loss": 2.1246,
+      "epoch": 0.29,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003974071285437345,
+      "loss": 2.1888,
       "step": 1580
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00017461029381599832,
-      "loss": 2.1229,
+      "epoch": 0.29,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003973550222713894,
+      "loss": 2.1943,
       "step": 1585
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 0.00017332898373220707,
-      "loss": 2.1136,
+      "epoch": 0.29,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003973024011159618,
+      "loss": 2.2067,
       "step": 1590
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.1103515625,
-      "learning_rate": 0.000172048787193717,
-      "loss": 2.1204,
+      "epoch": 0.3,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003972492652147361,
+      "loss": 2.1436,
       "step": 1595
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00017076975765019134,
-      "loss": 2.1239,
+      "epoch": 0.3,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003971956147063395,
+      "loss": 2.2443,
       "step": 1600
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.1103515625,
-      "learning_rate": 0.00016949194850257002,
-      "loss": 2.1308,
+      "epoch": 0.3,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003971414497307417,
+      "loss": 2.1936,
       "step": 1605
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00016821541310084006,
-      "loss": 2.1219,
+      "epoch": 0.3,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039708677042925473,
+      "loss": 2.173,
       "step": 1610
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00016694020474180814,
-      "loss": 2.1299,
+      "epoch": 0.3,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039703157694453246,
+      "loss": 2.1894,
       "step": 1615
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.115234375,
-      "learning_rate": 0.00016566637666687547,
-      "loss": 2.127,
+      "epoch": 0.3,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003969758694205701,
+      "loss": 2.1947,
       "step": 1620
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00016439398205981472,
-      "loss": 2.1255,
+      "epoch": 0.3,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000396919648002704,
+      "loss": 2.1641,
       "step": 1625
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00016312307404454967,
-      "loss": 2.1149,
+      "epoch": 0.3,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039686291283761127,
+      "loss": 2.1706,
       "step": 1630
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.0001618537056829373,
-      "loss": 2.1105,
+      "epoch": 0.3,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003968056640733092,
+      "loss": 2.1612,
       "step": 1635
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00016058592997255215,
-      "loss": 2.1477,
+      "epoch": 0.3,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039674790185915536,
+      "loss": 2.1836,
       "step": 1640
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00015931979984447385,
-      "loss": 2.13,
+      "epoch": 0.3,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003966896263458465,
+      "loss": 2.1739,
       "step": 1645
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00015805536816107703,
-      "loss": 2.1188,
+      "epoch": 0.31,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003966308376854186,
+      "loss": 2.2084,
       "step": 1650
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00015679268771382428,
-      "loss": 2.1194,
+      "epoch": 0.31,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003965715360312466,
+      "loss": 2.1563,
       "step": 1655
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00015553181122106234,
-      "loss": 2.1449,
+      "epoch": 0.31,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003965117215380435,
+      "loss": 2.2207,
       "step": 1660
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00015427279132582055,
-      "loss": 2.122,
+      "epoch": 0.31,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039645139436186047,
+      "loss": 2.1985,
       "step": 1665
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00015301568059361323,
-      "loss": 2.117,
+      "epoch": 0.31,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003963905546600862,
+      "loss": 2.201,
       "step": 1670
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.1103515625,
-      "learning_rate": 0.00015176053151024502,
-      "loss": 2.1244,
+      "epoch": 0.31,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00039632920259144645,
+      "loss": 2.185,
       "step": 1675
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.0001505073964796194,
-      "loss": 2.1206,
+      "epoch": 0.31,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003962673383160038,
+      "loss": 2.1761,
       "step": 1680
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.000149256327821551,
-      "loss": 2.1345,
+      "epoch": 0.31,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003962049619951571,
+      "loss": 2.1932,
       "step": 1685
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.00014800737776958097,
-      "loss": 2.1259,
+      "epoch": 0.31,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039614207379164086,
+      "loss": 2.2029,
       "step": 1690
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00014676059846879615,
-      "loss": 2.125,
+      "epoch": 0.31,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003960786738695255,
+      "loss": 2.2166,
       "step": 1695
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 0.00014551604197365222,
-      "loss": 2.124,
+      "epoch": 0.31,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039601476239421604,
+      "loss": 2.1965,
       "step": 1700
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.0001442737602458001,
-      "loss": 2.1176,
+      "epoch": 0.32,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003959503395324524,
+      "loss": 2.1923,
       "step": 1705
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.0001430338051519165,
-      "loss": 2.1208,
+      "epoch": 0.32,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00039588540545230845,
+      "loss": 2.1747,
       "step": 1710
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 0.00014179622846153872,
-      "loss": 2.1309,
+      "epoch": 0.32,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039581996032319195,
+      "loss": 2.1549,
       "step": 1715
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.0001405610818449027,
-      "loss": 2.1126,
+      "epoch": 0.32,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003957540043158439,
+      "loss": 2.1445,
       "step": 1720
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.0001393284168707864,
-      "loss": 2.1075,
+      "epoch": 0.32,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039568753760233806,
+      "loss": 2.1719,
       "step": 1725
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.0001380982850043562,
-      "loss": 2.1107,
+      "epoch": 0.32,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00039562056035608067,
+      "loss": 2.1836,
       "step": 1730
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00013687073760501828,
-      "loss": 2.1447,
+      "epoch": 0.32,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003955530727518099,
+      "loss": 2.1984,
       "step": 1735
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00013564582592427444,
-      "loss": 2.1151,
+      "epoch": 0.32,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003954850749655953,
+      "loss": 2.1842,
       "step": 1740
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 0.11669921875,
-      "learning_rate": 0.00013442360110358224,
-      "loss": 2.1212,
+      "epoch": 0.32,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039541656717483753,
+      "loss": 2.1859,
       "step": 1745
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00013320411417221974,
-      "loss": 2.1249,
+      "epoch": 0.32,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003953475495582679,
+      "loss": 2.1949,
       "step": 1750
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.0001319874160451551,
-      "loss": 2.1384,
+      "epoch": 0.32,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039527802229594763,
+      "loss": 2.1892,
       "step": 1755
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00013077355752092061,
-      "loss": 2.1235,
+      "epoch": 0.33,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039520798556926776,
+      "loss": 2.1656,
       "step": 1760
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00012956258927949196,
-      "loss": 2.1457,
+      "epoch": 0.33,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003951374395609483,
+      "loss": 2.1715,
       "step": 1765
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00012835456188017226,
-      "loss": 2.1232,
+      "epoch": 0.33,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039506638445503786,
+      "loss": 2.219,
       "step": 1770
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00012714952575948102,
-      "loss": 2.1135,
+      "epoch": 0.33,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00039499482043691363,
+      "loss": 2.1862,
       "step": 1775
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.00012594753122904858,
-      "loss": 2.1268,
+      "epoch": 0.33,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039492274769328014,
+      "loss": 2.1613,
       "step": 1780
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00012474862847351527,
-      "loss": 2.1204,
+      "epoch": 0.33,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00039485016641216934,
+      "loss": 2.181,
       "step": 1785
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00012355286754843654,
-      "loss": 2.1317,
+      "epoch": 0.33,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003947770767829398,
+      "loss": 2.1521,
       "step": 1790
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00012236029837819264,
-      "loss": 2.1233,
+      "epoch": 0.33,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00039470347899627634,
+      "loss": 2.1948,
       "step": 1795
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00012117097075390449,
-      "loss": 2.1264,
+      "epoch": 0.33,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003946293732441896,
+      "loss": 2.1656,
       "step": 1800
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00011998493433135474,
-      "loss": 2.1372,
+      "epoch": 0.33,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039455475972001544,
+      "loss": 2.1913,
       "step": 1805
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.00011880223862891462,
-      "loss": 2.1296,
+      "epoch": 0.33,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003944796386184144,
+      "loss": 2.198,
       "step": 1810
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.00011762293302547649,
-      "loss": 2.1252,
+      "epoch": 0.34,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003944040101353713,
+      "loss": 2.1771,
       "step": 1815
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 0.00011644706675839232,
-      "loss": 2.1107,
+      "epoch": 0.34,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003943278744681945,
+      "loss": 2.2192,
       "step": 1820
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 0.00011527468892141785,
-      "loss": 2.1107,
+      "epoch": 0.34,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000394251231815516,
+      "loss": 2.196,
       "step": 1825
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 0.00011410584846266266,
-      "loss": 2.1142,
+      "epoch": 0.34,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039417408237728995,
+      "loss": 2.1658,
       "step": 1830
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.0001129405941825471,
-      "loss": 2.1143,
+      "epoch": 0.34,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003940964263547931,
+      "loss": 2.1856,
       "step": 1835
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 0.00011177897473176413,
-      "loss": 2.1129,
+      "epoch": 0.34,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00039401826395062356,
+      "loss": 2.1597,
       "step": 1840
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.111328125,
-      "learning_rate": 0.00011062103860924873,
-      "loss": 2.1186,
+      "epoch": 0.34,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003939395953687006,
+      "loss": 2.1543,
       "step": 1845
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 0.00010946683416015264,
-      "loss": 2.1213,
+      "epoch": 0.34,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003938604208142642,
+      "loss": 2.213,
       "step": 1850
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00010831640957382601,
-      "loss": 2.1192,
+      "epoch": 0.34,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00039378074049387416,
+      "loss": 2.1782,
       "step": 1855
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 0.00010716981288180526,
-      "loss": 2.1373,
+      "epoch": 0.34,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039370055461541,
+      "loss": 2.1538,
       "step": 1860
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 0.000106027091955808,
-      "loss": 2.1416,
+      "epoch": 0.35,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039361986338807013,
+      "loss": 2.1781,
       "step": 1865
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00010488829450573434,
-      "loss": 2.1381,
+      "epoch": 0.35,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00039353866702237123,
+      "loss": 2.1624,
       "step": 1870
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 0.0001037534680776744,
-      "loss": 2.1228,
+      "epoch": 0.35,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003934569657301482,
+      "loss": 2.1507,
       "step": 1875
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 0.00010262266005192399,
-      "loss": 2.1245,
+      "epoch": 0.35,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039337475972455274,
+      "loss": 2.2238,
       "step": 1880
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 0.00010149591764100586,
-      "loss": 2.1272,
+      "epoch": 0.35,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003932920492200537,
+      "loss": 2.1408,
       "step": 1885
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.11328125,
-      "learning_rate": 0.00010037328788769884,
-      "loss": 2.1265,
+      "epoch": 0.35,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039320883443243605,
+      "loss": 2.1932,
       "step": 1890
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 9.925481766307341e-05,
-      "loss": 2.1364,
+      "epoch": 0.35,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00039312511557880025,
+      "loss": 2.193,
       "step": 1895
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 9.814055366453523e-05,
-      "loss": 2.1193,
+      "epoch": 0.35,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003930408928775621,
+      "loss": 2.2054,
       "step": 1900
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 9.703054241387499e-05,
-      "loss": 2.1262,
+      "epoch": 0.35,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00039295616654845154,
+      "loss": 2.1553,
       "step": 1905
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 9.592483025532652e-05,
-      "loss": 2.144,
+      "epoch": 0.35,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00039287093681251273,
+      "loss": 2.2483,
       "step": 1910
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 9.48234633536316e-05,
-      "loss": 2.1294,
+      "epoch": 0.35,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039278520389210306,
+      "loss": 2.2321,
       "step": 1915
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 9.372648769211258e-05,
-      "loss": 2.1113,
+      "epoch": 0.36,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039269896801089266,
+      "loss": 2.2265,
       "step": 1920
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 9.263394907075244e-05,
-      "loss": 2.1247,
+      "epoch": 0.36,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003926122293938639,
+      "loss": 2.1578,
       "step": 1925
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 9.154589310428288e-05,
-      "loss": 2.1161,
+      "epoch": 0.36,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003925249882673107,
+      "loss": 2.1928,
       "step": 1930
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 9.046236522027938e-05,
-      "loss": 2.13,
+      "epoch": 0.36,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000392437244858838,
+      "loss": 2.1682,
       "step": 1935
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 8.938341065726508e-05,
-      "loss": 2.1218,
+      "epoch": 0.36,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003923489993973612,
+      "loss": 2.219,
       "step": 1940
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.111328125,
-      "learning_rate": 8.830907446282162e-05,
-      "loss": 2.1196,
+      "epoch": 0.36,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003922602521131055,
+      "loss": 2.2151,
       "step": 1945
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 0.115234375,
-      "learning_rate": 8.723940149170853e-05,
-      "loss": 2.132,
+      "epoch": 0.36,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00039217100323760525,
+      "loss": 2.1969,
       "step": 1950
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.1171875,
-      "learning_rate": 8.617443640399056e-05,
-      "loss": 2.12,
+      "epoch": 0.36,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00039208125300370345,
+      "loss": 2.1983,
       "step": 1955
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 8.51142236631727e-05,
-      "loss": 2.1178,
+      "epoch": 0.36,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039199100164555116,
+      "loss": 2.164,
       "step": 1960
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 8.405880753434434e-05,
-      "loss": 2.1231,
+      "epoch": 0.36,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039190024939860673,
+      "loss": 2.1929,
       "step": 1965
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 8.300823208233062e-05,
-      "loss": 2.1259,
+      "epoch": 0.36,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00039180899649963536,
+      "loss": 2.1742,
       "step": 1970
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 8.196254116985303e-05,
-      "loss": 2.1061,
+      "epoch": 0.37,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003917172431867083,
+      "loss": 2.1718,
       "step": 1975
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 8.0921778455698e-05,
-      "loss": 2.1365,
+      "epoch": 0.37,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00039162498969920256,
+      "loss": 2.144,
       "step": 1980
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.988598739289408e-05,
-      "loss": 2.1183,
+      "epoch": 0.37,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003915322362777999,
+      "loss": 2.1415,
       "step": 1985
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.885521122689753e-05,
-      "loss": 2.1135,
+      "epoch": 0.37,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00039143898316448627,
+      "loss": 2.1537,
       "step": 1990
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 7.782949299378724e-05,
-      "loss": 2.1114,
+      "epoch": 0.37,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00039134523060255157,
+      "loss": 2.212,
       "step": 1995
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.68088755184673e-05,
-      "loss": 2.1381,
+      "epoch": 0.37,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003912509788365885,
+      "loss": 2.1902,
       "step": 2000
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 7.579340141287965e-05,
-      "loss": 2.1033,
+      "epoch": 0.37,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003911562281124921,
+      "loss": 2.1683,
       "step": 2005
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 7.478311307422456e-05,
-      "loss": 2.1169,
+      "epoch": 0.37,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003910609786774593,
+      "loss": 2.1913,
       "step": 2010
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.377805268319076e-05,
-      "loss": 2.1155,
+      "epoch": 0.37,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00039096523077998817,
+      "loss": 2.183,
       "step": 2015
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 7.27782622021939e-05,
-      "loss": 2.1077,
+      "epoch": 0.37,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00039086898466987705,
+      "loss": 2.2021,
       "step": 2020
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.178378337362519e-05,
-      "loss": 2.1162,
+      "epoch": 0.37,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00039077224059822413,
+      "loss": 2.1907,
       "step": 2025
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 7.079465771810828e-05,
-      "loss": 2.1294,
+      "epoch": 0.38,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00039067499881742675,
+      "loss": 2.1648,
       "step": 2030
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.11083984375,
-      "learning_rate": 6.981092653276547e-05,
-      "loss": 2.1204,
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00039057725958118085,
+      "loss": 2.1839,
       "step": 2035
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 6.88326308894941e-05,
-      "loss": 2.1197,
+      "epoch": 0.38,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00039047902314448,
+      "loss": 2.1565,
       "step": 2040
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 6.78598116332513e-05,
-      "loss": 2.1183,
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000390380289763615,
+      "loss": 2.1839,
       "step": 2045
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 6.68925093803489e-05,
-      "loss": 2.1259,
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00039028105969617327,
+      "loss": 2.2042,
       "step": 2050
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 6.593076451675734e-05,
-      "loss": 2.1203,
+      "epoch": 0.38,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003901813332010378,
+      "loss": 2.2105,
       "step": 2055
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 6.497461719642003e-05,
-      "loss": 2.1051,
+      "epoch": 0.38,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00039008111053838695,
+      "loss": 2.1962,
       "step": 2060
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.11328125,
-      "learning_rate": 6.402410733957627e-05,
-      "loss": 2.1393,
+      "epoch": 0.38,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00038998039196969336,
+      "loss": 2.1674,
       "step": 2065
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 6.307927463109504e-05,
-      "loss": 2.131,
+      "epoch": 0.38,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0003898791777577235,
+      "loss": 2.2015,
       "step": 2070
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.115234375,
-      "learning_rate": 6.214015851881788e-05,
-      "loss": 2.1138,
+      "epoch": 0.38,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000389777468166537,
+      "loss": 2.1528,
       "step": 2075
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 6.120679821191193e-05,
-      "loss": 2.1361,
+      "epoch": 0.38,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003896752634614859,
+      "loss": 2.2207,
       "step": 2080
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 0.115234375,
-      "learning_rate": 6.027923267923279e-05,
-      "loss": 2.1242,
+      "epoch": 0.39,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00038957256390921384,
+      "loss": 2.1882,
       "step": 2085
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 5.9357500647697786e-05,
-      "loss": 2.1396,
+      "epoch": 0.39,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003894693697776555,
+      "loss": 2.1701,
       "step": 2090
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 5.8441640600668924e-05,
-      "loss": 2.1037,
+      "epoch": 0.39,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00038936568133603595,
+      "loss": 2.1576,
       "step": 2095
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 5.75316907763459e-05,
-      "loss": 2.1335,
+      "epoch": 0.39,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003892614988548699,
+      "loss": 2.1723,
       "step": 2100
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.111328125,
-      "learning_rate": 5.6627689166170364e-05,
-      "loss": 2.1064,
+      "epoch": 0.39,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003891568226059608,
+      "loss": 2.2056,
       "step": 2105
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 0.111328125,
-      "learning_rate": 5.5729673513238814e-05,
-      "loss": 2.1039,
+      "epoch": 0.39,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003890516528624005,
+      "loss": 2.1687,
       "step": 2110
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 5.4837681310727464e-05,
-      "loss": 2.1427,
+      "epoch": 0.39,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003889459898985682,
+      "loss": 2.178,
       "step": 2115
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 5.395174980032645e-05,
-      "loss": 2.1386,
+      "epoch": 0.39,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003888398339901299,
+      "loss": 2.1991,
       "step": 2120
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 5.307191597068531e-05,
-      "loss": 2.124,
+      "epoch": 0.39,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00038873318541403774,
+      "loss": 2.1967,
       "step": 2125
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 5.2198216555868206e-05,
-      "loss": 2.1275,
+      "epoch": 0.39,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000388626044448529,
+      "loss": 2.2305,
       "step": 2130
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 5.133068803382073e-05,
-      "loss": 2.1182,
+      "epoch": 0.4,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003885184113731258,
+      "loss": 2.1956,
       "step": 2135
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 5.046936662484658e-05,
-      "loss": 2.1199,
+      "epoch": 0.4,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003884102864686339,
+      "loss": 2.1677,
       "step": 2140
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 4.9614288290095467e-05,
-      "loss": 2.1488,
+      "epoch": 0.4,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038830167001714243,
+      "loss": 2.194,
       "step": 2145
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 4.8765488730061485e-05,
-      "loss": 2.1171,
+      "epoch": 0.4,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00038819256230202275,
+      "loss": 2.1536,
       "step": 2150
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 4.792300338309288e-05,
-      "loss": 2.1305,
+      "epoch": 0.4,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00038808296360792787,
+      "loss": 2.1888,
       "step": 2155
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 4.70868674239124e-05,
-      "loss": 2.1236,
+      "epoch": 0.4,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00038797287422079194,
+      "loss": 2.1827,
       "step": 2160
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.115234375,
-      "learning_rate": 4.625711576214831e-05,
-      "loss": 2.1145,
+      "epoch": 0.4,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003878622944278291,
+      "loss": 2.1551,
       "step": 2165
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 4.543378304087746e-05,
-      "loss": 2.1362,
+      "epoch": 0.4,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000387751224517533,
+      "loss": 2.1649,
       "step": 2170
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 4.461690363517857e-05,
-      "loss": 2.1243,
+      "epoch": 0.4,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003876396647796759,
+      "loss": 2.1716,
       "step": 2175
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 4.380651165069707e-05,
-      "loss": 2.1126,
+      "epoch": 0.4,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000387527615505308,
+      "loss": 2.2026,
       "step": 2180
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.11328125,
-      "learning_rate": 4.3002640922221084e-05,
-      "loss": 2.1095,
+      "epoch": 0.4,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00038741507698675665,
+      "loss": 2.1967,
       "step": 2185
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.115234375,
-      "learning_rate": 4.220532501226902e-05,
-      "loss": 2.1351,
+      "epoch": 0.41,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00038730204951762574,
+      "loss": 2.1593,
       "step": 2190
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 4.141459720968792e-05,
-      "loss": 2.1225,
+      "epoch": 0.41,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00038718853339279457,
+      "loss": 2.1712,
       "step": 2195
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.115234375,
-      "learning_rate": 4.0630490528264196e-05,
-      "loss": 2.1168,
+      "epoch": 0.41,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00038707452890841744,
+      "loss": 2.1664,
       "step": 2200
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 3.985303770534459e-05,
-      "loss": 2.1339,
+      "epoch": 0.41,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003869600363619227,
+      "loss": 2.1739,
       "step": 2205
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 3.908227120046983e-05,
-      "loss": 2.1018,
+      "epoch": 0.41,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00038684505605201207,
+      "loss": 2.1687,
       "step": 2210
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.11328125,
-      "learning_rate": 3.831822319401916e-05,
-      "loss": 2.1128,
+      "epoch": 0.41,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00038672958827865966,
+      "loss": 2.1776,
       "step": 2215
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 3.756092558586694e-05,
-      "loss": 2.1124,
+      "epoch": 0.41,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003866136333431116,
+      "loss": 2.1977,
       "step": 2220
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.11328125,
-      "learning_rate": 3.681040999405079e-05,
-      "loss": 2.1334,
+      "epoch": 0.41,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003864971915478846,
+      "loss": 2.2004,
       "step": 2225
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 3.606670775345116e-05,
-      "loss": 2.1247,
+      "epoch": 0.41,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003863802631967659,
+      "loss": 2.2117,
       "step": 2230
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 0.11328125,
-      "learning_rate": 3.532984991448356e-05,
-      "loss": 2.1211,
+      "epoch": 0.41,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00038626284859481194,
+      "loss": 2.1687,
       "step": 2235
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 3.459986724180189e-05,
-      "loss": 2.1455,
+      "epoch": 0.41,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038614494804834783,
+      "loss": 2.1712,
       "step": 2240
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 0.115234375,
-      "learning_rate": 3.387679021301406e-05,
-      "loss": 2.1305,
+      "epoch": 0.42,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003860265618649664,
+      "loss": 2.161,
       "step": 2245
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.11328125,
-      "learning_rate": 3.316064901740934e-05,
-      "loss": 2.1421,
+      "epoch": 0.42,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00038590769035352755,
+      "loss": 2.1636,
       "step": 2250
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 3.245147355469822e-05,
-      "loss": 2.1317,
+      "epoch": 0.42,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003857883338241573,
+      "loss": 2.1801,
       "step": 2255
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 3.174929343376374e-05,
-      "loss": 2.1087,
+      "epoch": 0.42,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000385668492588247,
+      "loss": 2.1542,
       "step": 2260
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.119140625,
-      "learning_rate": 3.105413797142576e-05,
-      "loss": 2.1278,
+      "epoch": 0.42,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003855481669584527,
+      "loss": 2.1636,
       "step": 2265
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 3.0366036191216274e-05,
-      "loss": 2.1105,
+      "epoch": 0.42,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000385427357248694,
+      "loss": 2.2045,
       "step": 2270
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 2.9685016822168287e-05,
-      "loss": 2.1215,
+      "epoch": 0.42,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00038530606377415363,
+      "loss": 2.2136,
       "step": 2275
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.11328125,
-      "learning_rate": 2.9011108297615908e-05,
-      "loss": 2.1419,
+      "epoch": 0.42,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003851842868512763,
+      "loss": 2.1449,
       "step": 2280
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 2.834433875400755e-05,
-      "loss": 2.1307,
+      "epoch": 0.42,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.000385062026797768,
+      "loss": 2.173,
       "step": 2285
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 2.768473602973083e-05,
-      "loss": 2.1006,
+      "epoch": 0.42,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003849392839325952,
+      "loss": 2.1765,
       "step": 2290
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 2.7032327663950675e-05,
-      "loss": 2.1167,
+      "epoch": 0.42,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00038481605857598397,
+      "loss": 2.1903,
       "step": 2295
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 0.11328125,
-      "learning_rate": 2.6387140895459284e-05,
-      "loss": 2.1192,
+      "epoch": 0.43,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00038469235104941925,
+      "loss": 2.2198,
       "step": 2300
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 2.5749202661538972e-05,
-      "loss": 2.1193,
+      "epoch": 0.43,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003845681616756437,
+      "loss": 2.1736,
       "step": 2305
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 2.511853959683752e-05,
-      "loss": 2.119,
+      "epoch": 0.43,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003844434907786574,
+      "loss": 2.1704,
       "step": 2310
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 0.11328125,
-      "learning_rate": 2.4495178032255918e-05,
-      "loss": 2.1334,
+      "epoch": 0.43,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003843183386837163,
+      "loss": 2.2177,
       "step": 2315
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.11328125,
-      "learning_rate": 2.3879143993849474e-05,
-      "loss": 2.1016,
+      "epoch": 0.43,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00038419270571733217,
+      "loss": 2.2014,
       "step": 2320
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 2.3270463201740665e-05,
-      "loss": 2.1279,
+      "epoch": 0.43,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003840665922072709,
+      "loss": 2.1716,
       "step": 2325
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 2.2669161069045863e-05,
-      "loss": 2.1051,
+      "epoch": 0.43,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00038393999848255247,
+      "loss": 2.1645,
       "step": 2330
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 2.2075262700813747e-05,
-      "loss": 2.1403,
+      "epoch": 0.43,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003838129248734495,
+      "loss": 2.1691,
       "step": 2335
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 2.148879289297756e-05,
-      "loss": 2.1177,
+      "epoch": 0.43,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003836853717114866,
+      "loss": 2.143,
       "step": 2340
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.1181640625,
-      "learning_rate": 2.0909776131319548e-05,
-      "loss": 2.1135,
+      "epoch": 0.43,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0003835573393294396,
+      "loss": 2.2016,
       "step": 2345
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 2.0338236590448978e-05,
-      "loss": 2.1412,
+      "epoch": 0.43,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00038342882806133435,
+      "loss": 2.1799,
       "step": 2350
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 1.9774198132792353e-05,
-      "loss": 2.1297,
+      "epoch": 0.44,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00038329983824244624,
+      "loss": 2.1505,
       "step": 2355
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 1.9217684307597806e-05,
-      "loss": 2.1109,
+      "epoch": 0.44,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003831703702092993,
+      "loss": 2.1802,
       "step": 2360
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 1.866871834995112e-05,
-      "loss": 2.1333,
+      "epoch": 0.44,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038304042429966487,
+      "loss": 2.1758,
       "step": 2365
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 1.8127323179806234e-05,
-      "loss": 2.1076,
+      "epoch": 0.44,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00038291000085256116,
+      "loss": 2.1753,
       "step": 2370
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 1.7593521401027967e-05,
-      "loss": 2.1145,
+      "epoch": 0.44,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003827791002082524,
+      "loss": 2.1467,
       "step": 2375
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 1.7067335300448506e-05,
-      "loss": 2.1214,
+      "epoch": 0.44,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00038264772270824753,
+      "loss": 2.1684,
       "step": 2380
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 1.654878684693677e-05,
-      "loss": 2.1398,
+      "epoch": 0.44,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003825158686952997,
+      "loss": 2.1402,
       "step": 2385
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 1.6037897690481075e-05,
-      "loss": 2.0997,
+      "epoch": 0.44,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00038238353851340534,
+      "loss": 2.1557,
       "step": 2390
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 1.5534689161285643e-05,
-      "loss": 2.1291,
+      "epoch": 0.44,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00038225073250780284,
+      "loss": 2.1839,
       "step": 2395
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.115234375,
-      "learning_rate": 1.5039182268879504e-05,
-      "loss": 2.1358,
+      "epoch": 0.44,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003821174510249724,
+      "loss": 2.1392,
       "step": 2400
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 1.4551397701239721e-05,
-      "loss": 2.1091,
+      "epoch": 0.45,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003819836944126343,
+      "loss": 2.1385,
       "step": 2405
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 1.4071355823927424e-05,
-      "loss": 2.1098,
+      "epoch": 0.45,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00038184946301974876,
+      "loss": 2.1517,
       "step": 2410
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.115234375,
-      "learning_rate": 1.3599076679237676e-05,
-      "loss": 2.1321,
+      "epoch": 0.45,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003817147571965143,
+      "loss": 2.1759,
       "step": 2415
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 1.3134579985362517e-05,
-      "loss": 2.1004,
+      "epoch": 0.45,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038157957729436753,
+      "loss": 2.1571,
       "step": 2420
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.2677885135567979e-05,
-      "loss": 2.1175,
+      "epoch": 0.45,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00038144392366598167,
+      "loss": 2.1535,
       "step": 2425
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.2229011197384021e-05,
-      "loss": 2.1235,
+      "epoch": 0.45,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.000381307796665266,
+      "loss": 2.1696,
       "step": 2430
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 0.115234375,
-      "learning_rate": 1.1787976911808773e-05,
-      "loss": 2.1341,
+      "epoch": 0.45,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038117119664736454,
+      "loss": 2.2011,
       "step": 2435
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.1354800692525835e-05,
-      "loss": 2.1321,
+      "epoch": 0.45,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00038103412396865575,
+      "loss": 2.2036,
       "step": 2440
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 1.092950062513567e-05,
-      "loss": 2.0928,
+      "epoch": 0.45,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00038089657898675083,
+      "loss": 2.1503,
       "step": 2445
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.0512094466400402e-05,
-      "loss": 2.1177,
+      "epoch": 0.45,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003807585620604935,
+      "loss": 2.165,
       "step": 2450
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 1.0102599643502508e-05,
-      "loss": 2.1335,
+      "epoch": 0.45,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003806200735499585,
+      "loss": 2.1836,
       "step": 2455
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 9.70103325331717e-06,
-      "loss": 2.111,
+      "epoch": 0.46,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00038048111381645117,
+      "loss": 2.1602,
       "step": 2460
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 0.11328125,
-      "learning_rate": 9.307412061698428e-06,
-      "loss": 2.1119,
+      "epoch": 0.46,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00038034168322250594,
+      "loss": 2.1654,
       "step": 2465
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 8.92175250277929e-06,
-      "loss": 2.1151,
+      "epoch": 0.46,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003802017821318858,
+      "loss": 2.1579,
       "step": 2470
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 8.54407067828551e-06,
-      "loss": 2.1228,
+      "epoch": 0.46,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00038006141090958127,
+      "loss": 2.1749,
       "step": 2475
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.115234375,
-      "learning_rate": 8.174382356863365e-06,
-      "loss": 2.1171,
+      "epoch": 0.46,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003799205699218094,
+      "loss": 2.1696,
       "step": 2480
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 7.812702973421182e-06,
-      "loss": 2.1102,
+      "epoch": 0.46,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003797792595360127,
+      "loss": 2.1507,
       "step": 2485
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 7.4590476284852165e-06,
-      "loss": 2.1014,
+      "epoch": 0.46,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003796374801208585,
+      "loss": 2.1427,
       "step": 2490
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.11328125,
-      "learning_rate": 7.113431087568745e-06,
-      "loss": 2.1109,
+      "epoch": 0.46,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003794952320462376,
+      "loss": 2.189,
       "step": 2495
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 6.775867780555989e-06,
-      "loss": 2.1115,
+      "epoch": 0.46,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00037935251568326357,
+      "loss": 2.1501,
       "step": 2500
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 6.446371801099371e-06,
-      "loss": 2.1103,
+      "epoch": 0.46,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003792093314042718,
+      "loss": 2.2141,
       "step": 2505
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.11328125,
-      "learning_rate": 6.124956906031276e-06,
-      "loss": 2.1203,
+      "epoch": 0.46,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003790656795828182,
+      "loss": 2.1319,
       "step": 2510
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 0.1181640625,
-      "learning_rate": 5.811636514789598e-06,
-      "loss": 2.1227,
-      "step": 2515
+      "epoch": 0.47,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003789215605936787,
+      "loss": 2.1639,
+      "step": 2515
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.11328125,
-      "learning_rate": 5.506423708857455e-06,
-      "loss": 2.1251,
+      "epoch": 0.47,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00037877697481284784,
+      "loss": 2.2032,
       "step": 2520
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 5.209331231217052e-06,
-      "loss": 2.1305,
+      "epoch": 0.47,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003786319226175382,
+      "loss": 2.1946,
       "step": 2525
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 0.11328125,
-      "learning_rate": 4.920371485817632e-06,
-      "loss": 2.1124,
+      "epoch": 0.47,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000378486404386179,
+      "loss": 2.1725,
       "step": 2530
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 4.639556537057677e-06,
-      "loss": 2.11,
+      "epoch": 0.47,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003783404204984154,
+      "loss": 2.1877,
       "step": 2535
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 4.3668981092810365e-06,
-      "loss": 2.1314,
+      "epoch": 0.47,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003781939713351074,
+      "loss": 2.171,
       "step": 2540
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 4.102407586287571e-06,
-      "loss": 2.1307,
+      "epoch": 0.47,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003780470572783288,
+      "loss": 2.1502,
       "step": 2545
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.11328125,
-      "learning_rate": 3.846096010857791e-06,
-      "loss": 2.1042,
+      "epoch": 0.47,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00037789967871136647,
+      "loss": 2.1711,
       "step": 2550
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 3.5979740842918995e-06,
-      "loss": 2.1353,
+      "epoch": 0.47,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00037775183601871894,
+      "loss": 2.1952,
       "step": 2555
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 3.3580521659628106e-06,
-      "loss": 2.1164,
+      "epoch": 0.47,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00037760352958609564,
+      "loss": 2.1996,
       "step": 2560
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 3.126340272883899e-06,
-      "loss": 2.1268,
+      "epoch": 0.47,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000377454759800416,
+      "loss": 2.1831,
       "step": 2565
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 0.11376953125,
-      "learning_rate": 2.902848079290488e-06,
-      "loss": 2.119,
+      "epoch": 0.48,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037730552704980813,
+      "loss": 2.1548,
       "step": 2570
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 2.687584916236241e-06,
-      "loss": 2.1196,
+      "epoch": 0.48,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037715583172360807,
+      "loss": 2.2046,
       "step": 2575
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.1162109375,
-      "learning_rate": 2.4805597712032946e-06,
-      "loss": 2.1185,
+      "epoch": 0.48,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003770056742123586,
+      "loss": 2.187,
       "step": 2580
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 2.281781287727247e-06,
-      "loss": 2.1372,
+      "epoch": 0.48,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003768550549078085,
+      "loss": 2.1945,
       "step": 2585
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 2.091257765036092e-06,
-      "loss": 2.1092,
+      "epoch": 0.48,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003767039742029111,
+      "loss": 2.1916,
       "step": 2590
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.908997157703918e-06,
-      "loss": 2.1357,
+      "epoch": 0.48,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00037655243249182355,
+      "loss": 2.1468,
       "step": 2595
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 0.11181640625,
-      "learning_rate": 1.7350070753186176e-06,
-      "loss": 2.1145,
+      "epoch": 0.48,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00037640043016990587,
+      "loss": 2.2071,
       "step": 2600
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.115234375,
-      "learning_rate": 1.5692947821642324e-06,
-      "loss": 2.114,
+      "epoch": 0.48,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00037624796763371953,
+      "loss": 2.1319,
       "step": 2605
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 1.4118671969177265e-06,
-      "loss": 2.1391,
+      "epoch": 0.48,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00037609504528102693,
+      "loss": 2.2019,
       "step": 2610
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 1.2627308923600644e-06,
-      "loss": 2.1302,
+      "epoch": 0.48,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00037594166351078985,
+      "loss": 2.1725,
       "step": 2615
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.11328125,
-      "learning_rate": 1.1218920951018064e-06,
-      "loss": 2.135,
+      "epoch": 0.48,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00037578782272316877,
+      "loss": 2.1626,
       "step": 2620
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 0.115234375,
-      "learning_rate": 9.893566853230951e-07,
-      "loss": 2.1385,
+      "epoch": 0.49,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003756335233195218,
+      "loss": 2.1622,
       "step": 2625
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.111328125,
-      "learning_rate": 8.651301965282077e-07,
-      "loss": 2.1302,
+      "epoch": 0.49,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003754787657024033,
+      "loss": 2.1923,
       "step": 2630
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.11669921875,
-      "learning_rate": 7.492178153145402e-07,
-      "loss": 2.125,
+      "epoch": 0.49,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00037532355027556334,
+      "loss": 2.176,
       "step": 2635
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 0.11572265625,
-      "learning_rate": 6.416243811559808e-07,
-      "loss": 2.1311,
+      "epoch": 0.49,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00037516787744394615,
+      "loss": 2.1632,
       "step": 2640
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 5.423543862009384e-07,
-      "loss": 2.1087,
+      "epoch": 0.49,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003750117476136895,
+      "loss": 2.2095,
       "step": 2645
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 0.11328125,
-      "learning_rate": 4.5141197508475894e-07,
-      "loss": 2.105,
+      "epoch": 0.49,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003748551611921232,
+      "loss": 2.16,
       "step": 2650
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.115234375,
-      "learning_rate": 3.6880094475664204e-07,
-      "loss": 2.1256,
+      "epoch": 0.49,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003746981185877684,
+      "loss": 2.1668,
       "step": 2655
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.11328125,
-      "learning_rate": 2.945247443212118e-07,
-      "loss": 2.1204,
+      "epoch": 0.49,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003745406202103365,
+      "loss": 2.1289,
       "step": 2660
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 2.2858647489441e-07,
-      "loss": 2.1264,
+      "epoch": 0.49,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003743826664707279,
+      "loss": 2.1539,
       "step": 2665
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 1.7098888947404412e-07,
-      "loss": 2.1144,
+      "epoch": 0.49,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003742242577810307,
+      "loss": 2.1488,
       "step": 2670
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.1142578125,
-      "learning_rate": 1.217343928249237e-07,
-      "loss": 2.0979,
+      "epoch": 0.5,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003740653945545205,
+      "loss": 2.1595,
       "step": 2675
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 8.082504137836288e-08,
-      "loss": 2.136,
+      "epoch": 0.5,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003739060772056583,
+      "loss": 2.2113,
       "step": 2680
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.11279296875,
-      "learning_rate": 4.8262543146382345e-08,
-      "loss": 2.1116,
+      "epoch": 0.5,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00037374630615008987,
+      "loss": 2.1642,
       "step": 2685
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.11474609375,
-      "learning_rate": 2.404825765034424e-08,
-      "loss": 2.1157,
+      "epoch": 0.5,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003735860818046451,
+      "loss": 2.2222,
       "step": 2690
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 0.12255859375,
-      "learning_rate": 8.183195864264192e-09,
-      "loss": 2.1259,
+      "epoch": 0.5,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003734254045873359,
+      "loss": 2.1348,
       "step": 2695
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.1123046875,
-      "learning_rate": 6.680201725117741e-10,
-      "loss": 2.1135,
+      "epoch": 0.5,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00037326427491735613,
+      "loss": 2.1671,
       "step": 2700
     },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00037310269321507984,
+      "loss": 2.1556,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003729406599020605,
+      "loss": 2.1649,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00037277817540102973,
+      "loss": 2.2039,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003726152401358962,
+      "loss": 2.1396,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003724518545317448,
+      "loss": 2.1507,
+      "step": 2725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00037228801901483503,
+      "loss": 2.1799,
+      "step": 2730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00037212373401260046,
+      "loss": 2.187,
+      "step": 2735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003719589999536471,
+      "loss": 2.1976,
+      "step": 2740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0003717938172677527,
+      "loss": 2.1753,
+      "step": 2745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00037162818638586533,
+      "loss": 2.1722,
+      "step": 2750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00037146210774010235,
+      "loss": 2.1647,
+      "step": 2755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00037129558176374937,
+      "loss": 2.1425,
+      "step": 2760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000371128608891259,
+      "loss": 2.1494,
+      "step": 2765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037096118955824985,
+      "loss": 2.1378,
+      "step": 2770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003707933242015051,
+      "loss": 2.2139,
+      "step": 2775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037062501325897185,
+      "loss": 2.1782,
+      "step": 2780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003704562571697595,
+      "loss": 2.1569,
+      "step": 2785
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003702870563741389,
+      "loss": 2.1816,
+      "step": 2790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00037011741131354095,
+      "loss": 2.1699,
+      "step": 2795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00036994732243055586,
+      "loss": 2.1796,
+      "step": 2800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003697767901689315,
+      "loss": 2.1894,
+      "step": 2805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00036960581497357263,
+      "loss": 2.2078,
+      "step": 2810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00036943439729053936,
+      "loss": 2.2227,
+      "step": 2815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003692625375670466,
+      "loss": 2.129,
+      "step": 2820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00036909023625146216,
+      "loss": 2.1497,
+      "step": 2825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003689174937933061,
+      "loss": 2.1787,
+      "step": 2830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003687443106432493,
+      "loss": 2.1767,
+      "step": 2835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00036857068725311247,
+      "loss": 2.2093,
+      "step": 2840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00036839662407586477,
+      "loss": 2.192,
+      "step": 2845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003682221215656227,
+      "loss": 2.1946,
+      "step": 2850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00036804718017764915,
+      "loss": 2.1645,
+      "step": 2855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00036787180036835176,
+      "loss": 2.1705,
+      "step": 2860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003676959825952821,
+      "loss": 2.1524,
+      "step": 2865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00036751972731713443,
+      "loss": 2.1779,
+      "step": 2870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003673430349937442,
+      "loss": 2.1696,
+      "step": 2875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00036716590608608726,
+      "loss": 2.2063,
+      "step": 2880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00036698834105627833,
+      "loss": 2.1803,
+      "step": 2885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00036681034036757025,
+      "loss": 2.1442,
+      "step": 2890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00036663190448435197,
+      "loss": 2.1818,
+      "step": 2895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003664530338721483,
+      "loss": 2.1825,
+      "step": 2900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00036627372899761783,
+      "loss": 2.1626,
+      "step": 2905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00036609399032855246,
+      "loss": 2.1804,
+      "step": 2910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003659138183338755,
+      "loss": 2.2026,
+      "step": 2915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000365733213483641,
+      "loss": 2.1142,
+      "step": 2920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003655521762490322,
+      "loss": 2.1742,
+      "step": 2925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00036537070710236045,
+      "loss": 2.1832,
+      "step": 2930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003651888065170638,
+      "loss": 2.1556,
+      "step": 2935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003650064749677061,
+      "loss": 2.1317,
+      "step": 2940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003648237129299752,
+      "loss": 2.1326,
+      "step": 2945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003646405208806825,
+      "loss": 2.1876,
+      "step": 2950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000364456899297761,
+      "loss": 2.1772,
+      "step": 2955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00036427284866026425,
+      "loss": 2.1754,
+      "step": 2960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00036408836944836544,
+      "loss": 2.1947,
+      "step": 2965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00036390346214335565,
+      "loss": 2.1612,
+      "step": 2970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003637181272276428,
+      "loss": 2.1577,
+      "step": 2975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00036353236518475067,
+      "loss": 2.1848,
+      "step": 2980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00036334617649931713,
+      "loss": 2.1327,
+      "step": 2985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003631595616570931,
+      "loss": 2.165,
+      "step": 2990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003629725211449416,
+      "loss": 2.1891,
+      "step": 2995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003627850554508359,
+      "loss": 2.1425,
+      "step": 3000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003625971650638586,
+      "loss": 2.2033,
+      "step": 3005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00036240885047420044,
+      "loss": 2.1094,
+      "step": 3010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00036222011217315863,
+      "loss": 2.167,
+      "step": 3015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00036203095065313597,
+      "loss": 2.1472,
+      "step": 3020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003618413664076394,
+      "loss": 2.1544,
+      "step": 3025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003616513599312787,
+      "loss": 2.1752,
+      "step": 3030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003614609317197651,
+      "loss": 2.1324,
+      "step": 3035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00036127008226991036,
+      "loss": 2.1255,
+      "step": 3040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00036107881207962493,
+      "loss": 2.1587,
+      "step": 3045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00036088712164791706,
+      "loss": 2.1735,
+      "step": 3050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003606950114748914,
+      "loss": 2.2219,
+      "step": 3055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003605024820617477,
+      "loss": 2.1603,
+      "step": 3060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003603095339107794,
+      "loss": 2.1873,
+      "step": 3065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003601161675253724,
+      "loss": 2.1496,
+      "step": 3070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00035992238341000377,
+      "loss": 2.174,
+      "step": 3075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003597281820702403,
+      "loss": 2.1466,
+      "step": 3080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00035953356401273754,
+      "loss": 2.2021,
+      "step": 3085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00035933852974523797,
+      "loss": 2.1912,
+      "step": 3090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00035914307977657,
+      "loss": 2.1559,
+      "step": 3095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00035894721461664653,
+      "loss": 2.1599,
+      "step": 3100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003587509347764639,
+      "loss": 2.1783,
+      "step": 3105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003585542407681,
+      "loss": 2.1911,
+      "step": 3110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003583571331047133,
+      "loss": 2.1656,
+      "step": 3115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003581596123005417,
+      "loss": 2.2032,
+      "step": 3120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00035796167887090064,
+      "loss": 2.1761,
+      "step": 3125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003577633333321824,
+      "loss": 2.1613,
+      "step": 3130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00035756457620185413,
+      "loss": 2.1708,
+      "step": 3135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003573654079984569,
+      "loss": 2.1343,
+      "step": 3140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003571658292416043,
+      "loss": 2.1691,
+      "step": 3145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00035696584045198085,
+      "loss": 2.1903,
+      "step": 3150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000356765442151341,
+      "loss": 2.1999,
+      "step": 3155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003565646348625074,
+      "loss": 2.178,
+      "step": 3160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00035636341910937,
+      "loss": 2.1452,
+      "step": 3165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00035616179541688396,
+      "loss": 2.1834,
+      "step": 3170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00035595976431106917,
+      "loss": 2.1612,
+      "step": 3175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00035575732631900817,
+      "loss": 2.1553,
+      "step": 3180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003555544819688451,
+      "loss": 2.1546,
+      "step": 3185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00035535123178978415,
+      "loss": 2.1749,
+      "step": 3190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003551475763120886,
+      "loss": 2.1814,
+      "step": 3195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003549435160670787,
+      "loss": 2.1969,
+      "step": 3200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00035473905158713096,
+      "loss": 2.1151,
+      "step": 3205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00035453418340567654,
+      "loss": 2.1841,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003543289120571996,
+      "loss": 2.1678,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00035412323807723644,
+      "loss": 2.1795,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035391716200237346,
+      "loss": 2.1592,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003537106843702464,
+      "loss": 2.1743,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00035350380571953836,
+      "loss": 2.1537,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00035329652658997883,
+      "loss": 2.1708,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00035308884752234207,
+      "loss": 2.1661,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003528807690584458,
+      "loss": 2.1801,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003526722917411497,
+      "loss": 2.1754,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003524634161143539,
+      "loss": 2.1586,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035225414272299784,
+      "loss": 2.1633,
+      "step": 3265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003520444721130586,
+      "loss": 2.1521,
+      "step": 3270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00035183440483154977,
+      "loss": 2.1735,
+      "step": 3275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035162394142651955,
+      "loss": 2.1732,
+      "step": 3280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00035141308244704975,
+      "loss": 2.1746,
+      "step": 3285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003512018284432542,
+      "loss": 2.1395,
+      "step": 3290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00035099017996627726,
+      "loss": 2.1977,
+      "step": 3295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035077813756829247,
+      "loss": 2.175,
+      "step": 3300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035056570180250104,
+      "loss": 2.1792,
+      "step": 3305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00035035287322313053,
+      "loss": 2.1743,
+      "step": 3310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00035013965238543324,
+      "loss": 2.1874,
+      "step": 3315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003499260398456849,
+      "loss": 2.1612,
+      "step": 3320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000349712036161183,
+      "loss": 2.1541,
+      "step": 3325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003494976418902458,
+      "loss": 2.186,
+      "step": 3330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00034928285759221027,
+      "loss": 2.1962,
+      "step": 3335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003490676838274311,
+      "loss": 2.1682,
+      "step": 3340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003488521211572789,
+      "loss": 2.1448,
+      "step": 3345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00034863617014413916,
+      "loss": 2.1853,
+      "step": 3350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003484198313514103,
+      "loss": 2.1469,
+      "step": 3355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003482031053435025,
+      "loss": 2.1847,
+      "step": 3360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003479859926858362,
+      "loss": 2.1409,
+      "step": 3365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00034776849394484047,
+      "loss": 2.1633,
+      "step": 3370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003475506096879518,
+      "loss": 2.1971,
+      "step": 3375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003473323404836122,
+      "loss": 2.2042,
+      "step": 3380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00034711368690126836,
+      "loss": 2.186,
+      "step": 3385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00034689464951136943,
+      "loss": 2.1366,
+      "step": 3390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00034667522888536597,
+      "loss": 2.1616,
+      "step": 3395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003464554255957085,
+      "loss": 2.135,
+      "step": 3400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00034623524021584577,
+      "loss": 2.212,
+      "step": 3405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003460146733202234,
+      "loss": 2.1422,
+      "step": 3410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00034579372548428235,
+      "loss": 2.1923,
+      "step": 3415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003455723972844574,
+      "loss": 2.1822,
+      "step": 3420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003453506892981757,
+      "loss": 2.1625,
+      "step": 3425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00034512860210385516,
+      "loss": 2.1835,
+      "step": 3430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00034490613628090325,
+      "loss": 2.1461,
+      "step": 3435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00034468329240971486,
+      "loss": 2.1501,
+      "step": 3440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003444600710716715,
+      "loss": 2.2022,
+      "step": 3445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00034423647284913934,
+      "loss": 2.1706,
+      "step": 3450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003440124983254678,
+      "loss": 2.2002,
+      "step": 3455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003437881480849881,
+      "loss": 2.1764,
+      "step": 3460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003435634227130116,
+      "loss": 2.1617,
+      "step": 3465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003433383227958284,
+      "loss": 2.1917,
+      "step": 3470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003431128489207058,
+      "loss": 2.161,
+      "step": 3475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00034288700167588653,
+      "loss": 2.1907,
+      "step": 3480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003426607816505877,
+      "loss": 2.1946,
+      "step": 3485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00034243418943499863,
+      "loss": 2.1641,
+      "step": 3490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00034220722562028,
+      "loss": 2.1497,
+      "step": 3495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00034197989079856177,
+      "loss": 2.1762,
+      "step": 3500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0003417521855629418,
+      "loss": 2.1725,
+      "step": 3505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003415241105074845,
+      "loss": 2.1933,
+      "step": 3510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00034129566622721893,
+      "loss": 2.1642,
+      "step": 3515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003410668533181375,
+      "loss": 2.1441,
+      "step": 3520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00034083767237719437,
+      "loss": 2.1493,
+      "step": 3525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003406081240023038,
+      "loss": 2.1511,
+      "step": 3530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00034037820879233874,
+      "loss": 2.124,
+      "step": 3535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003401479273471291,
+      "loss": 2.1946,
+      "step": 3540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00033991728026746035,
+      "loss": 2.1686,
+      "step": 3545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003396862681550717,
+      "loss": 2.1675,
+      "step": 3550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00033945489161265496,
+      "loss": 2.1678,
+      "step": 3555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003392231512438525,
+      "loss": 2.1688,
+      "step": 3560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000338991047653256,
+      "loss": 2.1348,
+      "step": 3565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003387585814464046,
+      "loss": 2.1784,
+      "step": 3570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003385257532297837,
+      "loss": 2.1485,
+      "step": 3575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.000338292563610823,
+      "loss": 2.1796,
+      "step": 3580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00033805901319789523,
+      "loss": 2.1385,
+      "step": 3585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00033782510260031397,
+      "loss": 2.1663,
+      "step": 3590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00033759083242833316,
+      "loss": 2.1829,
+      "step": 3595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00033735620329314423,
+      "loss": 2.1654,
+      "step": 3600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00033712121580687553,
+      "loss": 2.1817,
+      "step": 3605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00033688587058259005,
+      "loss": 2.1509,
+      "step": 3610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003366501682342842,
+      "loss": 2.1625,
+      "step": 3615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00033641410937688623,
+      "loss": 2.1596,
+      "step": 3620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00033617769462625414,
+      "loss": 2.1586,
+      "step": 3625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00033594092459917485,
+      "loss": 2.1758,
+      "step": 3630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003357037999133618,
+      "loss": 2.1764,
+      "step": 3635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000335466321187454,
+      "loss": 2.1582,
+      "step": 3640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00033522848904101387,
+      "loss": 2.2074,
+      "step": 3645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003349903040945261,
+      "loss": 2.2073,
+      "step": 3650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00033475176696939565,
+      "loss": 2.1773,
+      "step": 3655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00033451287828794625,
+      "loss": 2.1756,
+      "step": 3660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000334273638673419,
+      "loss": 2.1762,
+      "step": 3665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003340340487499705,
+      "loss": 2.1871,
+      "step": 3670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00033379410914267113,
+      "loss": 2.123,
+      "step": 3675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003335538204775037,
+      "loss": 2.1596,
+      "step": 3680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00033331318338136154,
+      "loss": 2.1601,
+      "step": 3685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00033307219848204727,
+      "loss": 2.1494,
+      "step": 3690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00033283086640827075,
+      "loss": 2.1872,
+      "step": 3695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003325891877896475,
+      "loss": 2.1453,
+      "step": 3700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00033234716325669726,
+      "loss": 2.1355,
+      "step": 3705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003321047934408422,
+      "loss": 2.1521,
+      "step": 3710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00033186207897440535,
+      "loss": 2.1659,
+      "step": 3715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00033161902049060884,
+      "loss": 2.1289,
+      "step": 3720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00033137561862357225,
+      "loss": 2.1724,
+      "step": 3725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00033113187400831133,
+      "loss": 2.1556,
+      "step": 3730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003308877872807356,
+      "loss": 2.1442,
+      "step": 3735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00033064335907764743,
+      "loss": 2.1844,
+      "step": 3740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003303985900367399,
+      "loss": 2.1469,
+      "step": 3745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00033015348079659543,
+      "loss": 2.1677,
+      "step": 3750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00032990803199668396,
+      "loss": 2.1505,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00032966224427736135,
+      "loss": 2.1784,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003294161182798675,
+      "loss": 2.1666,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00032916965464632503,
+      "loss": 2.184,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003289228540197373,
+      "loss": 2.1779,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00032867571704398707,
+      "loss": 2.1613,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003284282443638343,
+      "loss": 2.1666,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003281804366249151,
+      "loss": 2.1374,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003279322944737394,
+      "loss": 2.1905,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00032768381855768984,
+      "loss": 2.1411,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00032743500952501966,
+      "loss": 2.1392,
+      "step": 3805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003271858680248513,
+      "loss": 2.1566,
+      "step": 3810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003269363947071744,
+      "loss": 2.1638,
+      "step": 3815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003266865902228446,
+      "loss": 2.1398,
+      "step": 3820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00032643645522358115,
+      "loss": 2.1758,
+      "step": 3825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00032618599036196593,
+      "loss": 2.1802,
+      "step": 3830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003259351962914411,
+      "loss": 2.1981,
+      "step": 3835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003256840736663079,
+      "loss": 2.1491,
+      "step": 3840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00032543262314172465,
+      "loss": 2.1215,
+      "step": 3845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00032518084537370523,
+      "loss": 2.1359,
+      "step": 3850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00032492874101911715,
+      "loss": 2.1479,
+      "step": 3855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00032467631073568006,
+      "loss": 2.1548,
+      "step": 3860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003244235551819639,
+      "loss": 2.1486,
+      "step": 3865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003241704750173871,
+      "loss": 2.141,
+      "step": 3870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003239170709022151,
+      "loss": 2.183,
+      "step": 3875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003236633434975586,
+      "loss": 2.1688,
+      "step": 3880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003234092934653714,
+      "loss": 2.1295,
+      "step": 3885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003231549214684495,
+      "loss": 2.1869,
+      "step": 3890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003229002281704283,
+      "loss": 2.2002,
+      "step": 3895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00032264521423578193,
+      "loss": 2.1324,
+      "step": 3900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003223898803298207,
+      "loss": 2.2215,
+      "step": 3905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003221342271186901,
+      "loss": 2.1823,
+      "step": 3910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00032187825526936814,
+      "loss": 2.1501,
+      "step": 3915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00032162196544966457,
+      "loss": 2.1406,
+      "step": 3920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00032136535832821855,
+      "loss": 2.1025,
+      "step": 3925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003211084345744968,
+      "loss": 2.196,
+      "step": 3930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003208511948587926,
+      "loss": 2.163,
+      "step": 3935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00032059363985222316,
+      "loss": 2.1585,
+      "step": 3940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00032033577022672847,
+      "loss": 2.1706,
+      "step": 3945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003200775866550692,
+      "loss": 2.1882,
+      "step": 3950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003198190898108251,
+      "loss": 2.1918,
+      "step": 3955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00031956028036839336,
+      "loss": 2.1635,
+      "step": 3960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00031930115900298653,
+      "loss": 2.1744,
+      "step": 3965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000319041726390631,
+      "loss": 2.1833,
+      "step": 3970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003187819832081653,
+      "loss": 2.184,
+      "step": 3975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031852193013323795,
+      "loss": 2.1278,
+      "step": 3980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003182615678443063,
+      "loss": 2.1921,
+      "step": 3985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00031800089702063413,
+      "loss": 2.1368,
+      "step": 3990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003177399183422903,
+      "loss": 2.1714,
+      "step": 3995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003174786324901467,
+      "loss": 2.1578,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003172170401458768,
+      "loss": 2.1785,
+      "step": 4005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00031695514199195357,
+      "loss": 2.1574,
+      "step": 4010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00031669293871164784,
+      "loss": 2.1741,
+      "step": 4015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031643043098902645,
+      "loss": 2.1515,
+      "step": 4020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031616761950895066,
+      "loss": 2.1695,
+      "step": 4025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031590450495707387,
+      "loss": 2.1727,
+      "step": 4030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003156410880198406,
+      "loss": 2.1336,
+      "step": 4035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031537736938448387,
+      "loss": 2.1436,
+      "step": 4040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003151133497390242,
+      "loss": 2.1573,
+      "step": 4045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00031484902977226703,
+      "loss": 2.0997,
+      "step": 4050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003145844101738017,
+      "loss": 2.1623,
+      "step": 4055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003143194916339987,
+      "loss": 2.1201,
+      "step": 4060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00031405427484400923,
+      "loss": 2.142,
+      "step": 4065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00031378876049576186,
+      "loss": 2.1642,
+      "step": 4070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003135229492819619,
+      "loss": 2.1755,
+      "step": 4075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003132568418960889,
+      "loss": 2.161,
+      "step": 4080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003129904390323954,
+      "loss": 2.1652,
+      "step": 4085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003127237413859046,
+      "loss": 2.1814,
+      "step": 4090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003124567496524087,
+      "loss": 2.1351,
+      "step": 4095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00031218946452846733,
+      "loss": 2.167,
+      "step": 4100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00031192188671140547,
+      "loss": 2.1597,
+      "step": 4105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003116540168993117,
+      "loss": 2.1689,
+      "step": 4110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00031138585579103647,
+      "loss": 2.1414,
+      "step": 4115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003111174040861901,
+      "loss": 2.1896,
+      "step": 4120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003108486624851411,
+      "loss": 2.1622,
+      "step": 4125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00031057963168901426,
+      "loss": 2.1455,
+      "step": 4130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003103103123996889,
+      "loss": 2.1765,
+      "step": 4135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000310040705319797,
+      "loss": 2.1614,
+      "step": 4140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003097708111527214,
+      "loss": 2.1687,
+      "step": 4145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003095006306025937,
+      "loss": 2.1996,
+      "step": 4150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00030923016437429295,
+      "loss": 2.1228,
+      "step": 4155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00030895941317344325,
+      "loss": 2.1984,
+      "step": 4160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003086883777064125,
+      "loss": 2.1719,
+      "step": 4165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003084170586803098,
+      "loss": 2.1309,
+      "step": 4170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00030814545680298445,
+      "loss": 2.1502,
+      "step": 4175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003078735727830234,
+      "loss": 2.1748,
+      "step": 4180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003076014073297498,
+      "loss": 2.1534,
+      "step": 4185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000307328961153221,
+      "loss": 2.169,
+      "step": 4190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00030705623496422673,
+      "loss": 2.2074,
+      "step": 4195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00030678322947428737,
+      "loss": 2.1453,
+      "step": 4200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00030650994539565174,
+      "loss": 2.1617,
+      "step": 4205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003062363834412957,
+      "loss": 2.1778,
+      "step": 4210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00030596254432492,
+      "loss": 2.129,
+      "step": 4215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003056884287609483,
+      "loss": 2.1792,
+      "step": 4220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00030541403746452584,
+      "loss": 2.192,
+      "step": 4225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00030513937115151697,
+      "loss": 2.1739,
+      "step": 4230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003048644305385035,
+      "loss": 2.1651,
+      "step": 4235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00030458921634278306,
+      "loss": 2.162,
+      "step": 4240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00030431372928236693,
+      "loss": 2.1479,
+      "step": 4245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0003040379700759782,
+      "loss": 2.1697,
+      "step": 4250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003037619394430501,
+      "loss": 2.1917,
+      "step": 4255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00030348563810372403,
+      "loss": 2.1322,
+      "step": 4260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00030320906677884735,
+      "loss": 2.2019,
+      "step": 4265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00030293222618997214,
+      "loss": 2.1271,
+      "step": 4270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00030265511705935264,
+      "loss": 2.1734,
+      "step": 4275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0003023777401099442,
+      "loss": 2.1741,
+      "step": 4280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003021000960654001,
+      "loss": 2.1295,
+      "step": 4285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00030182218565007126,
+      "loss": 2.1086,
+      "step": 4290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00030154400958900304,
+      "loss": 2.1965,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00030126556860793406,
+      "loss": 2.1739,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00030098686343329393,
+      "loss": 2.1785,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00030070789479220164,
+      "loss": 2.1833,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00030042866341246357,
+      "loss": 2.1742,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0003001491700225715,
+      "loss": 2.1997,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002998694153517007,
+      "loss": 2.1871,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00029958940012970835,
+      "loss": 2.1697,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000299309125087131,
+      "loss": 2.1582,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00029902859095518345,
+      "loss": 2.1734,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002987477984657562,
+      "loss": 2.1583,
+      "step": 4345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002984667483514139,
+      "loss": 2.1455,
+      "step": 4350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002981854413453931,
+      "loss": 2.1834,
+      "step": 4355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0002979038781816009,
+      "loss": 2.1747,
+      "step": 4360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002976220595946125,
+      "loss": 2.1205,
+      "step": 4365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00029733998631966947,
+      "loss": 2.1561,
+      "step": 4370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002970576590926778,
+      "loss": 2.1635,
+      "step": 4375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002967750786502063,
+      "loss": 2.137,
+      "step": 4380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00029649224572948394,
+      "loss": 2.1344,
+      "step": 4385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002962091610683988,
+      "loss": 2.1707,
+      "step": 4390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002959258254054956,
+      "loss": 2.1537,
+      "step": 4395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002956422394799737,
+      "loss": 2.1807,
+      "step": 4400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002953584040316856,
+      "loss": 2.1838,
+      "step": 4405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00029507431980113483,
+      "loss": 2.1748,
+      "step": 4410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00029478998752947383,
+      "loss": 2.1416,
+      "step": 4415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002945054079585021,
+      "loss": 2.1579,
+      "step": 4420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002942205818306645,
+      "loss": 2.1964,
+      "step": 4425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002939355098890491,
+      "loss": 2.1878,
+      "step": 4430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00029365019287738513,
+      "loss": 2.1768,
+      "step": 4435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002933646315400414,
+      "loss": 2.1224,
+      "step": 4440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000293078826622024,
+      "loss": 2.1614,
+      "step": 4445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002927927788689746,
+      "loss": 2.1222,
+      "step": 4450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00029250648902716815,
+      "loss": 2.1392,
+      "step": 4455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002922199578435116,
+      "loss": 2.1278,
+      "step": 4460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00029193318606554126,
+      "loss": 2.1057,
+      "step": 4465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002916461744414213,
+      "loss": 2.1295,
+      "step": 4470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00029135892371994133,
+      "loss": 2.1465,
+      "step": 4475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002910714346505151,
+      "loss": 2.1864,
+      "step": 4480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00029078370798317786,
+      "loss": 2.1191,
+      "step": 4485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00029049574446858513,
+      "loss": 2.1343,
+      "step": 4490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00029020754485801,
+      "loss": 2.1347,
+      "step": 4495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002899191099033417,
+      "loss": 2.1863,
+      "step": 4500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002896304403570834,
+      "loss": 2.1284,
+      "step": 4505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002893415369723502,
+      "loss": 2.1697,
+      "step": 4510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002890524005028674,
+      "loss": 2.1384,
+      "step": 4515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002887630317029684,
+      "loss": 2.1453,
+      "step": 4520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002884734313275927,
+      "loss": 2.1537,
+      "step": 4525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00028818360013228384,
+      "loss": 2.1534,
+      "step": 4530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002878935388731878,
+      "loss": 2.2001,
+      "step": 4535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002876032483070505,
+      "loss": 2.1625,
+      "step": 4540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00028731272919121643,
+      "loss": 2.1471,
+      "step": 4545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00028702198228362595,
+      "loss": 2.1882,
+      "step": 4550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.000286731008342814,
+      "loss": 2.1713,
+      "step": 4555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002864398081279078,
+      "loss": 2.2102,
+      "step": 4560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002861483823986248,
+      "loss": 2.1601,
+      "step": 4565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00028585673191527087,
+      "loss": 2.1745,
+      "step": 4570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00028556485743873815,
+      "loss": 2.1265,
+      "step": 4575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002852727597305034,
+      "loss": 2.1508,
+      "step": 4580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002849804395526254,
+      "loss": 2.1352,
+      "step": 4585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00028468789766774373,
+      "loss": 2.2117,
+      "step": 4590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002843951348390761,
+      "loss": 2.1643,
+      "step": 4595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002841021518304167,
+      "loss": 2.1697,
+      "step": 4600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00028380894940613436,
+      "loss": 2.1483,
+      "step": 4605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00028351552833116996,
+      "loss": 2.1538,
+      "step": 4610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002832218893710353,
+      "loss": 2.1349,
+      "step": 4615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00028292803329181006,
+      "loss": 2.1824,
+      "step": 4620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002826339608601409,
+      "loss": 2.1932,
+      "step": 4625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00028233967284323864,
+      "loss": 2.1542,
+      "step": 4630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00028204517000887663,
+      "loss": 2.1764,
+      "step": 4635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002817504531253885,
+      "loss": 2.1645,
+      "step": 4640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00028145552296166666,
+      "loss": 2.1308,
+      "step": 4645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002811603802871596,
+      "loss": 2.1788,
+      "step": 4650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002808650258718704,
+      "loss": 2.1833,
+      "step": 4655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00028056946048635455,
+      "loss": 2.1497,
+      "step": 4660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002802736849017179,
+      "loss": 2.148,
+      "step": 4665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002799776998896147,
+      "loss": 2.1204,
+      "step": 4670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00027968150622224565,
+      "loss": 2.1634,
+      "step": 4675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002793851046723557,
+      "loss": 2.1469,
+      "step": 4680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00027908849601323226,
+      "loss": 2.1578,
+      "step": 4685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000278791681018703,
+      "loss": 2.165,
+      "step": 4690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00027849466046313384,
+      "loss": 2.1491,
+      "step": 4695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002781974351214272,
+      "loss": 2.1494,
+      "step": 4700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002779000057690196,
+      "loss": 2.1238,
+      "step": 4705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002776023731818798,
+      "loss": 2.187,
+      "step": 4710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000277304538136507,
+      "loss": 2.1481,
+      "step": 4715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002770065014099282,
+      "loss": 2.1226,
+      "step": 4720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002767082637796969,
+      "loss": 2.1702,
+      "step": 4725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00027640982602389065,
+      "loss": 2.1044,
+      "step": 4730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002761111889211091,
+      "loss": 2.1993,
+      "step": 4735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000275812353250472,
+      "loss": 2.1658,
+      "step": 4740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00027551331979161714,
+      "loss": 2.1546,
+      "step": 4745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00027521408932469816,
+      "loss": 2.1709,
+      "step": 4750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00027491466263038304,
+      "loss": 2.1624,
+      "step": 4755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00027461504048985135,
+      "loss": 2.1396,
+      "step": 4760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00027431522368479277,
+      "loss": 2.1456,
+      "step": 4765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002740152129974048,
+      "loss": 2.1567,
+      "step": 4770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00027371500921039065,
+      "loss": 2.1809,
+      "step": 4775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00027341461310695744,
+      "loss": 2.1142,
+      "step": 4780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00027311402547081407,
+      "loss": 2.1756,
+      "step": 4785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00027281324708616897,
+      "loss": 2.1675,
+      "step": 4790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00027251227873772825,
+      "loss": 2.1433,
+      "step": 4795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00027221112121069384,
+      "loss": 2.168,
+      "step": 4800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00027190977529076087,
+      "loss": 2.2056,
+      "step": 4805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00027160824176411633,
+      "loss": 2.1628,
+      "step": 4810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002713065214174363,
+      "loss": 2.1301,
+      "step": 4815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002710046150378846,
+      "loss": 2.1805,
+      "step": 4820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002707025234131101,
+      "loss": 2.1282,
+      "step": 4825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00027040024733124523,
+      "loss": 2.159,
+      "step": 4830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00027009778758090345,
+      "loss": 2.1703,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00026979514495117745,
+      "loss": 2.1592,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002694923202316371,
+      "loss": 2.1478,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00026918931421232725,
+      "loss": 2.1303,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00026888612768376577,
+      "loss": 2.1344,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00026858276143694145,
+      "loss": 2.1288,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00026827921626331207,
+      "loss": 2.1749,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000267975492954802,
+      "loss": 2.1649,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00026767159230380055,
+      "loss": 2.1581,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002673675151031596,
+      "loss": 2.1654,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002670632621461916,
+      "loss": 2.1446,
+      "step": 4885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002667588342266677,
+      "loss": 2.1597,
+      "step": 4890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00026645423213881523,
+      "loss": 2.1272,
+      "step": 4895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002661494566773162,
+      "loss": 2.1691,
+      "step": 4900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00026584450863730474,
+      "loss": 2.1536,
+      "step": 4905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002655393888143654,
+      "loss": 2.1609,
+      "step": 4910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00026523409800453075,
+      "loss": 2.1631,
+      "step": 4915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00026492863700427945,
+      "loss": 2.1416,
+      "step": 4920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002646230066105343,
+      "loss": 2.1406,
+      "step": 4925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00026431720762065995,
+      "loss": 2.1789,
+      "step": 4930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002640112408324608,
+      "loss": 2.1909,
+      "step": 4935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002637051070441793,
+      "loss": 2.138,
+      "step": 4940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00026339880705449325,
+      "loss": 2.1424,
+      "step": 4945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002630923416625143,
+      "loss": 2.1703,
+      "step": 4950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002627857116677855,
+      "loss": 2.1225,
+      "step": 4955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002624789178702796,
+      "loss": 2.1497,
+      "step": 4960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002621719610703961,
+      "loss": 2.1998,
+      "step": 4965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00026186484206896046,
+      "loss": 2.1573,
+      "step": 4970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00026155756166722086,
+      "loss": 2.169,
+      "step": 4975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002612501206668468,
+      "loss": 2.1583,
+      "step": 4980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00026094251986992647,
+      "loss": 2.1491,
+      "step": 4985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00026063476007896525,
+      "loss": 2.1644,
+      "step": 4990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00026032684209688323,
+      "loss": 2.1645,
+      "step": 4995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002600187667270132,
+      "loss": 2.1521,
+      "step": 5000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00025971053477309854,
+      "loss": 2.2114,
+      "step": 5005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00025940214703929116,
+      "loss": 2.1751,
+      "step": 5010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00025909360433014935,
+      "loss": 2.1664,
+      "step": 5015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002587849074506358,
+      "loss": 2.2008,
+      "step": 5020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00025847605720611533,
+      "loss": 2.1275,
+      "step": 5025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002581670544023531,
+      "loss": 2.16,
+      "step": 5030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00025785789984551186,
+      "loss": 2.1662,
+      "step": 5035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002575485943421506,
+      "loss": 2.1949,
+      "step": 5040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002572391386992221,
+      "loss": 2.1761,
+      "step": 5045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002569295337240708,
+      "loss": 2.1522,
+      "step": 5050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00025661978022443056,
+      "loss": 2.153,
+      "step": 5055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000256309879008423,
+      "loss": 2.1931,
+      "step": 5060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002559998308845549,
+      "loss": 2.1806,
+      "step": 5065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002556896366617166,
+      "loss": 2.1721,
+      "step": 5070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00025537929714917925,
+      "loss": 2.1222,
+      "step": 5075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002550688131565932,
+      "loss": 2.1832,
+      "step": 5080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00025475818549398584,
+      "loss": 2.1428,
+      "step": 5085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00025444741497175925,
+      "loss": 2.1774,
+      "step": 5090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00025413650240068824,
+      "loss": 2.1533,
+      "step": 5095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0002538254485919183,
+      "loss": 2.1717,
+      "step": 5100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00025351425435696316,
+      "loss": 2.1586,
+      "step": 5105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00025320292050770327,
+      "loss": 2.1615,
+      "step": 5110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002528914478563831,
+      "loss": 2.1589,
+      "step": 5115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002525798372156092,
+      "loss": 2.1327,
+      "step": 5120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002522680893983482,
+      "loss": 2.1334,
+      "step": 5125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002519562052179247,
+      "loss": 2.1986,
+      "step": 5130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002516441854880189,
+      "loss": 2.1506,
+      "step": 5135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00025133203102266474,
+      "loss": 2.1611,
+      "step": 5140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00025101974263624764,
+      "loss": 2.1453,
+      "step": 5145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00025070732114350233,
+      "loss": 2.1207,
+      "step": 5150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002503947673595111,
+      "loss": 2.14,
+      "step": 5155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00025008208209970095,
+      "loss": 2.1846,
+      "step": 5160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002497692661798422,
+      "loss": 2.1757,
+      "step": 5165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00024945632041604586,
+      "loss": 2.1336,
+      "step": 5170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00024914324562476197,
+      "loss": 2.1451,
+      "step": 5175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002488300426227768,
+      "loss": 2.1725,
+      "step": 5180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002485167122272115,
+      "loss": 2.1376,
+      "step": 5185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002482032552555192,
+      "loss": 2.179,
+      "step": 5190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00024788967252548365,
+      "loss": 2.1259,
+      "step": 5195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002475759648552164,
+      "loss": 2.2117,
+      "step": 5200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002472621330631551,
+      "loss": 2.1336,
+      "step": 5205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002469481779680611,
+      "loss": 2.1223,
+      "step": 5210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002466341003890176,
+      "loss": 2.143,
+      "step": 5215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00024631990114542716,
+      "loss": 2.1461,
+      "step": 5220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002460055810570101,
+      "loss": 2.1894,
+      "step": 5225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00024569114094380153,
+      "loss": 2.1477,
+      "step": 5230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00024537658162615,
+      "loss": 2.1513,
+      "step": 5235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002450619039247151,
+      "loss": 2.1068,
+      "step": 5240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002447471086604651,
+      "loss": 2.1434,
+      "step": 5245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00024443219665467496,
+      "loss": 2.1596,
+      "step": 5250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002441171687289245,
+      "loss": 2.1451,
+      "step": 5255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00024380202570509557,
+      "loss": 2.1864,
+      "step": 5260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00024348676840537064,
+      "loss": 2.15,
+      "step": 5265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00024317139765223008,
+      "loss": 2.1246,
+      "step": 5270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00024285591426845044,
+      "loss": 2.1653,
+      "step": 5275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002425403190771018,
+      "loss": 2.1562,
+      "step": 5280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00024222461290154647,
+      "loss": 2.1499,
+      "step": 5285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00024190879656543578,
+      "loss": 2.1019,
+      "step": 5290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00024159287089270875,
+      "loss": 2.1572,
+      "step": 5295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00024127683670758954,
+      "loss": 2.1302,
+      "step": 5300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00024096069483458545,
+      "loss": 2.1094,
+      "step": 5305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00024064444609848474,
+      "loss": 2.1348,
+      "step": 5310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00024032809132435446,
+      "loss": 2.1543,
+      "step": 5315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00024001163133753823,
+      "loss": 2.1001,
+      "step": 5320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00023969506696365432,
+      "loss": 2.1323,
+      "step": 5325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00023937839902859315,
+      "loss": 2.1543,
+      "step": 5330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023906162835851555,
+      "loss": 2.1806,
+      "step": 5335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023874475577985007,
+      "loss": 2.1394,
+      "step": 5340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002384277821192914,
+      "loss": 2.1701,
+      "step": 5345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00023811070820379783,
+      "loss": 2.1586,
+      "step": 5350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002377935348605893,
+      "loss": 2.14,
+      "step": 5355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00023747626291714498,
+      "loss": 2.1441,
+      "step": 5360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023715889320120135,
+      "loss": 2.1567,
+      "step": 5365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00023684142654075004,
+      "loss": 2.1384,
+      "step": 5370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023652386376403562,
+      "loss": 2.1509,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00023620620569955332,
+      "loss": 2.1294,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002358884531760469,
+      "loss": 2.1489,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00023557060702250677,
+      "loss": 2.1334,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00023525266806816744,
+      "loss": 2.1313,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00023493463714250568,
+      "loss": 2.1266,
+      "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,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00023429830269631886,
+      "loss": 2.1299,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00023398000083593829,
+      "loss": 2.106,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023366161032451982,
+      "loss": 2.1316,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00023334313199271808,
+      "loss": 2.1283,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00023302456667141704,
+      "loss": 2.1331,
+      "step": 5430
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00023270591519172743,
+      "loss": 2.1087,
+      "step": 5435
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00023238717838498495,
+      "loss": 2.1158,
+      "step": 5440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00023206835708274777,
+      "loss": 2.1277,
+      "step": 5445
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023174945211679458,
+      "loss": 2.1279,
+      "step": 5450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002314304643191222,
+      "loss": 2.1279,
+      "step": 5455
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00023111139452194378,
+      "loss": 2.1358,
+      "step": 5460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00023079224355768618,
+      "loss": 2.0907,
+      "step": 5465
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00023047301225898826,
+      "loss": 2.0924,
+      "step": 5470
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023015370145869812,
+      "loss": 2.1152,
+      "step": 5475
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002298343119898716,
+      "loss": 2.1402,
+      "step": 5480
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00022951484468576965,
+      "loss": 2.1549,
+      "step": 5485
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002291953003798563,
+      "loss": 2.0924,
+      "step": 5490
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00022887567990579637,
+      "loss": 2.1107,
+      "step": 5495
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00022855598409745356,
+      "loss": 2.117,
+      "step": 5500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00022823621378888803,
+      "loss": 2.1362,
+      "step": 5505
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002279163698143544,
+      "loss": 2.0987,
+      "step": 5510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022759645300829924,
+      "loss": 2.122,
+      "step": 5515
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022727646420535934,
+      "loss": 2.0972,
+      "step": 5520
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00022695640424035926,
+      "loss": 2.0972,
+      "step": 5525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00022663627394830935,
+      "loss": 2.1113,
+      "step": 5530
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022631607416440311,
+      "loss": 2.0969,
+      "step": 5535
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00022599580572401568,
+      "loss": 2.1307,
+      "step": 5540
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00022567546946270105,
+      "loss": 2.088,
+      "step": 5545
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00022535506621619046,
+      "loss": 2.1045,
+      "step": 5550
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00022503459682038962,
+      "loss": 2.0976,
+      "step": 5555
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00022471406211137693,
+      "loss": 2.125,
+      "step": 5560
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00022439346292540114,
+      "loss": 2.0827,
+      "step": 5565
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00022407280009887934,
+      "loss": 2.1119,
+      "step": 5570
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00022375207446839453,
+      "loss": 2.1057,
+      "step": 5575
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00022343128687069363,
+      "loss": 2.1151,
+      "step": 5580
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00022311043814268513,
+      "loss": 2.1059,
+      "step": 5585
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002227895291214371,
+      "loss": 2.0808,
+      "step": 5590
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00022246856064417494,
+      "loss": 2.0955,
+      "step": 5595
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00022214753354827908,
+      "loss": 2.1399,
+      "step": 5600
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00022182644867128283,
+      "loss": 2.1297,
+      "step": 5605
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00022150530685087044,
+      "loss": 2.1046,
+      "step": 5610
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00022118410892487455,
+      "loss": 2.1181,
+      "step": 5615
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00022086285573127423,
+      "loss": 2.1405,
+      "step": 5620
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00022054154810819277,
+      "loss": 2.1203,
+      "step": 5625
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00022022018689389535,
+      "loss": 2.1112,
+      "step": 5630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00021989877292678714,
+      "loss": 2.1051,
+      "step": 5635
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00021957730704541084,
+      "loss": 2.0685,
+      "step": 5640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00021925579008844453,
+      "loss": 2.1139,
+      "step": 5645
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002189342228946997,
+      "loss": 2.096,
+      "step": 5650
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00021861260630311863,
+      "loss": 2.1158,
+      "step": 5655
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00021829094115277287,
+      "loss": 2.1403,
+      "step": 5660
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002179692282828603,
+      "loss": 2.1077,
+      "step": 5665
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00021764746853270356,
+      "loss": 2.0722,
+      "step": 5670
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00021732566274174726,
+      "loss": 2.146,
+      "step": 5675
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00021700381174955658,
+      "loss": 2.1231,
+      "step": 5680
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00021668191639581425,
+      "loss": 2.1488,
+      "step": 5685
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00021635997752031893,
+      "loss": 2.0689,
+      "step": 5690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002160379959629828,
+      "loss": 2.1011,
+      "step": 5695
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00021571597256382917,
+      "loss": 2.0967,
+      "step": 5700
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.000215393908162991,
+      "loss": 2.0587,
+      "step": 5705
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00021507180360070765,
+      "loss": 2.091,
+      "step": 5710
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002147496597173237,
+      "loss": 2.0682,
+      "step": 5715
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00021442747735328608,
+      "loss": 2.0907,
+      "step": 5720
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00021410525734914218,
+      "loss": 2.0557,
+      "step": 5725
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00021378300054553754,
+      "loss": 2.1218,
+      "step": 5730
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00021346070778321393,
+      "loss": 2.1368,
+      "step": 5735
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00021313837990300657,
+      "loss": 2.1157,
+      "step": 5740
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00021281601774584256,
+      "loss": 2.1341,
+      "step": 5745
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00021249362215273836,
+      "loss": 2.1038,
+      "step": 5750
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0002121711939647977,
+      "loss": 2.1066,
+      "step": 5755
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002118487340232092,
+      "loss": 2.0586,
+      "step": 5760
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002115262431692445,
+      "loss": 2.0977,
+      "step": 5765
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0002112037222442558,
+      "loss": 2.1219,
+      "step": 5770
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002108811720896738,
+      "loss": 2.1481,
+      "step": 5775
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00021055859354700537,
+      "loss": 2.097,
+      "step": 5780
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00021023598745783145,
+      "loss": 2.1014,
+      "step": 5785
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00020991335466380502,
+      "loss": 2.1153,
+      "step": 5790
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002095906960066486,
+      "loss": 2.0794,
+      "step": 5795
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00020926801232815204,
+      "loss": 2.1081,
+      "step": 5800
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00020894530447017077,
+      "loss": 2.1151,
+      "step": 5805
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00020862257327462297,
+      "loss": 2.1014,
+      "step": 5810
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00020829981958348813,
+      "loss": 2.1248,
+      "step": 5815
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000207977044238804,
+      "loss": 2.1307,
+      "step": 5820
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002076542480826651,
+      "loss": 2.114,
+      "step": 5825
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00020733143195722002,
+      "loss": 2.1148,
+      "step": 5830
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00020700859670466981,
+      "loss": 2.1523,
+      "step": 5835
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000206685743167265,
+      "loss": 2.1059,
+      "step": 5840
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002063628721873041,
+      "loss": 2.1124,
+      "step": 5845
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.000206039984607131,
+      "loss": 2.1202,
+      "step": 5850
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00020571708126913295,
+      "loss": 2.1126,
+      "step": 5855
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00020539416301573832,
+      "loss": 2.11,
+      "step": 5860
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00020507123068941438,
+      "loss": 2.1197,
+      "step": 5865
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00020474828513266505,
+      "loss": 2.1129,
+      "step": 5870
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0002044253271880289,
+      "loss": 2.1116,
+      "step": 5875
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00020410235769807668,
+      "loss": 2.0783,
+      "step": 5880
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00020377937750540946,
+      "loss": 2.1268,
+      "step": 5885
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00020345638745265592,
+      "loss": 2.1024,
+      "step": 5890
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0002031333883824707,
+      "loss": 2.0943,
+      "step": 5895
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00020281038113753198,
+      "loss": 2.1171,
+      "step": 5900
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00020248736656053914,
+      "loss": 2.1008,
+      "step": 5905
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00020216434549421073,
+      "loss": 2.0961,
+      "step": 5910
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002018413187812823,
+      "loss": 2.09,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00020151828726450396,
+      "loss": 2.1428,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00020119525178663862,
+      "loss": 2.1225,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0002008722131904593,
+      "loss": 2.1276,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00020054917231874722,
+      "loss": 2.0879,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00020022613001428953,
+      "loss": 2.1028,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001999030871198772,
+      "loss": 2.1489,
+      "step": 5945
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001995800444783027,
+      "loss": 2.1126,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019925700293235764,
+      "loss": 2.1148,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019893396332483108,
+      "loss": 2.1339,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019861092649850682,
+      "loss": 2.106,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019828789329616163,
+      "loss": 2.1064,
+      "step": 5970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019796486456056245,
+      "loss": 2.1179,
+      "step": 5975
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019764184113446487,
+      "loss": 2.1449,
+      "step": 5980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019731882386061055,
+      "loss": 2.0949,
+      "step": 5985
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019699581358172507,
+      "loss": 2.0989,
+      "step": 5990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019667281114051592,
+      "loss": 2.1064,
+      "step": 5995
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001963498173796698,
+      "loss": 2.144,
+      "step": 6000
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019602683314185105,
+      "loss": 2.1148,
+      "step": 6005
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019570385926969902,
+      "loss": 2.1193,
+      "step": 6010
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001953808966058263,
+      "loss": 2.1129,
+      "step": 6015
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019505794599281577,
+      "loss": 2.128,
+      "step": 6020
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001947350082732192,
+      "loss": 2.116,
+      "step": 6025
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019441208428955467,
+      "loss": 2.0824,
+      "step": 6030
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019408917488430438,
+      "loss": 2.0985,
+      "step": 6035
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019376628089991246,
+      "loss": 2.1691,
+      "step": 6040
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019344340317878293,
+      "loss": 2.1106,
+      "step": 6045
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019312054256327729,
+      "loss": 2.115,
+      "step": 6050
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001927976998957124,
+      "loss": 2.1005,
+      "step": 6055
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019247487601835835,
+      "loss": 2.1078,
+      "step": 6060
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019215207177343617,
+      "loss": 2.1068,
+      "step": 6065
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001918292880031157,
+      "loss": 2.0899,
+      "step": 6070
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001915065255495134,
+      "loss": 2.1286,
+      "step": 6075
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019118378525468996,
+      "loss": 2.1497,
+      "step": 6080
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001908610679606484,
+      "loss": 2.0953,
+      "step": 6085
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019053837450933175,
+      "loss": 2.1246,
+      "step": 6090
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019021570574262076,
+      "loss": 2.124,
+      "step": 6095
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018989306250233177,
+      "loss": 2.1067,
+      "step": 6100
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018957044563021455,
+      "loss": 2.1097,
+      "step": 6105
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018924785596795014,
+      "loss": 2.1008,
+      "step": 6110
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018892529435714854,
+      "loss": 2.0854,
+      "step": 6115
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001886027616393465,
+      "loss": 2.1099,
+      "step": 6120
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001882802586560055,
+      "loss": 2.0994,
+      "step": 6125
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018795778624850935,
+      "loss": 2.1248,
+      "step": 6130
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018763534525816224,
+      "loss": 2.125,
+      "step": 6135
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001873129365261862,
+      "loss": 2.0986,
+      "step": 6140
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001869905608937192,
+      "loss": 2.1644,
+      "step": 6145
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001866682192018129,
+      "loss": 2.0913,
+      "step": 6150
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001863459122914304,
+      "loss": 2.1306,
+      "step": 6155
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00018602364100344382,
+      "loss": 2.1249,
+      "step": 6160
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00018570140617863275,
+      "loss": 2.1154,
+      "step": 6165
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018537920865768136,
+      "loss": 2.1396,
+      "step": 6170
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001850570492811767,
+      "loss": 2.1388,
+      "step": 6175
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000184734928889606,
+      "loss": 2.1193,
+      "step": 6180
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018441284832335496,
+      "loss": 2.1099,
+      "step": 6185
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00018409080842270558,
+      "loss": 2.1267,
+      "step": 6190
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018376881002783355,
+      "loss": 2.0948,
+      "step": 6195
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018344685397880613,
+      "loss": 2.0771,
+      "step": 6200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00018312494111558033,
+      "loss": 2.0696,
+      "step": 6205
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001828030722780005,
+      "loss": 2.0965,
+      "step": 6210
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018248124830579608,
+      "loss": 2.1058,
+      "step": 6215
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001821594700385794,
+      "loss": 2.1195,
+      "step": 6220
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001818377383158435,
+      "loss": 2.1058,
+      "step": 6225
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001815160539769602,
+      "loss": 2.1028,
+      "step": 6230
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00018119441786117763,
+      "loss": 2.1162,
+      "step": 6235
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018087283080761793,
+      "loss": 2.1048,
+      "step": 6240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00018055129365527543,
+      "loss": 2.1006,
+      "step": 6245
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00018022980724301423,
+      "loss": 2.0785,
+      "step": 6250
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017990837240956602,
+      "loss": 2.0936,
+      "step": 6255
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017958698999352794,
+      "loss": 2.1514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00017926566083336035,
+      "loss": 2.0913,
+      "step": 6265
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001789443857673848,
+      "loss": 2.104,
+      "step": 6270
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00017862316563378144,
+      "loss": 2.1296,
+      "step": 6275
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00017830200127058756,
+      "loss": 2.11,
+      "step": 6280
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001779808935156944,
+      "loss": 2.0885,
+      "step": 6285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001776598432068459,
+      "loss": 2.0768,
+      "step": 6290
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017733885118163595,
+      "loss": 2.114,
+      "step": 6295
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017701791827750654,
+      "loss": 2.0768,
+      "step": 6300
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001766970453317452,
+      "loss": 2.1185,
+      "step": 6305
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00017637623318148324,
+      "loss": 2.0814,
+      "step": 6310
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017605548266369321,
+      "loss": 2.1291,
+      "step": 6315
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00017573479461518708,
+      "loss": 2.1129,
+      "step": 6320
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001754141698726135,
+      "loss": 2.099,
+      "step": 6325
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00017509360927245622,
+      "loss": 2.1072,
+      "step": 6330
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00017477311365103162,
+      "loss": 2.1402,
+      "step": 6335
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001744526838444865,
+      "loss": 2.0942,
+      "step": 6340
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00017413232068879596,
+      "loss": 2.1288,
+      "step": 6345
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001738120250197612,
+      "loss": 2.086,
+      "step": 6350
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017349179767300738,
+      "loss": 2.0644,
+      "step": 6355
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001731716394839815,
+      "loss": 2.0885,
+      "step": 6360
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00017285155128794993,
+      "loss": 2.1015,
+      "step": 6365
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00017253153391999656,
+      "loss": 2.0885,
+      "step": 6370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00017221158821502055,
+      "loss": 2.133,
+      "step": 6375
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.000171891715007734,
+      "loss": 2.0947,
+      "step": 6380
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00017157191513265984,
+      "loss": 2.1282,
+      "step": 6385
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001712521894241298,
+      "loss": 2.0747,
+      "step": 6390
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00017093253871628206,
+      "loss": 2.108,
+      "step": 6395
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00017061296384305913,
+      "loss": 2.1235,
+      "step": 6400
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001702934656382056,
+      "loss": 2.0921,
+      "step": 6405
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00016997404493526618,
+      "loss": 2.147,
+      "step": 6410
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00016965470256758328,
+      "loss": 2.1055,
+      "step": 6415
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00016933543936829506,
+      "loss": 2.1194,
+      "step": 6420
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00016901625617033283,
+      "loss": 2.1074,
+      "step": 6425
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001686971538064196,
+      "loss": 2.0911,
+      "step": 6430
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001683781331090672,
+      "loss": 2.1184,
+      "step": 6435
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00016805919491057463,
+      "loss": 2.131,
+      "step": 6440
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00016774034004302525,
+      "loss": 2.1019,
+      "step": 6445
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001674215693382855,
+      "loss": 2.0983,
+      "step": 6450
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00016710288362800196,
+      "loss": 2.1232,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00016678428374359964,
+      "loss": 2.1064,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00016646577051627936,
+      "loss": 2.089,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00016614734477701605,
+      "loss": 2.1403,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00016582900735655649,
+      "loss": 2.1017,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00016551075908541693,
+      "loss": 2.1058,
+      "step": 6480
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001651926007938809,
+      "loss": 2.1283,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00016487453331199737,
+      "loss": 2.0858,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00016455655746957834,
+      "loss": 2.0983,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00016423867409619687,
+      "loss": 2.1359,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001639208840211844,
+      "loss": 2.1151,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001636031880736293,
+      "loss": 2.108,
+      "step": 6510
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001632855870823743,
+      "loss": 2.0985,
+      "step": 6515
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00016296808187601437,
+      "loss": 2.0595,
+      "step": 6520
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00016265067328289447,
+      "loss": 2.0978,
+      "step": 6525
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00016233336213110773,
+      "loss": 2.1,
+      "step": 6530
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00016201614924849287,
+      "loss": 2.1164,
+      "step": 6535
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00016169903546263238,
+      "loss": 2.0785,
+      "step": 6540
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00016138202160085005,
+      "loss": 2.161,
+      "step": 6545
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00016106510849020914,
+      "loss": 2.1207,
+      "step": 6550
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00016074829695750996,
+      "loss": 2.1187,
+      "step": 6555
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001604315878292879,
+      "loss": 2.1297,
+      "step": 6560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00016011498193181108,
+      "loss": 2.1197,
+      "step": 6565
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00015979848009107828,
+      "loss": 2.1178,
+      "step": 6570
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00015948208313281694,
+      "loss": 2.1231,
+      "step": 6575
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001591657918824808,
+      "loss": 2.0752,
+      "step": 6580
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00015884960716524787,
+      "loss": 2.1095,
+      "step": 6585
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00015853352980601802,
+      "loss": 2.1059,
+      "step": 6590
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00015821756062941132,
+      "loss": 2.113,
+      "step": 6595
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001579017004597654,
+      "loss": 2.1061,
+      "step": 6600
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001575859501211336,
+      "loss": 2.1151,
+      "step": 6605
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001572703104372826,
+      "loss": 2.0957,
+      "step": 6610
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00015695478223169053,
+      "loss": 2.1191,
+      "step": 6615
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00015663936632754464,
+      "loss": 2.0812,
+      "step": 6620
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015632406354773918,
+      "loss": 2.1666,
+      "step": 6625
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00015600887471487322,
+      "loss": 2.0995,
+      "step": 6630
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015569380065124863,
+      "loss": 2.1118,
+      "step": 6635
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001553788421788678,
+      "loss": 2.1483,
+      "step": 6640
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001550640001194317,
+      "loss": 2.076,
+      "step": 6645
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001547492752943373,
+      "loss": 2.1103,
+      "step": 6650
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00015443466852467594,
+      "loss": 2.1059,
+      "step": 6655
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00015412018063123091,
+      "loss": 2.1148,
+      "step": 6660
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001538058124344754,
+      "loss": 2.1381,
+      "step": 6665
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00015349156475457021,
+      "loss": 2.0915,
+      "step": 6670
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001531774384113618,
+      "loss": 2.0751,
+      "step": 6675
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001528634342243801,
+      "loss": 2.0934,
+      "step": 6680
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015254955301283635,
+      "loss": 2.127,
+      "step": 6685
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001522357955956207,
+      "loss": 2.1216,
+      "step": 6690
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00015192216279130076,
+      "loss": 2.112,
+      "step": 6695
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015160865541811872,
+      "loss": 2.0921,
+      "step": 6700
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015129527429398975,
+      "loss": 2.146,
+      "step": 6705
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001509820202364993,
+      "loss": 2.095,
+      "step": 6710
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001506688940629017,
+      "loss": 2.0786,
+      "step": 6715
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00015035589659011746,
+      "loss": 2.0724,
+      "step": 6720
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001500430286347314,
+      "loss": 2.1372,
+      "step": 6725
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014973029101299023,
+      "loss": 2.0978,
+      "step": 6730
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001494176845408008,
+      "loss": 2.1331,
+      "step": 6735
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014910521003372792,
+      "loss": 2.1272,
+      "step": 6740
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014879286830699195,
+      "loss": 2.1223,
+      "step": 6745
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00014848066017546674,
+      "loss": 2.093,
+      "step": 6750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00014816858645367778,
+      "loss": 2.1193,
+      "step": 6755
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00014785664795579983,
+      "loss": 2.081,
+      "step": 6760
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.000147544845495655,
+      "loss": 2.1271,
+      "step": 6765
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001472331798867102,
+      "loss": 2.0703,
+      "step": 6770
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001469216519420755,
+      "loss": 2.1123,
+      "step": 6775
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001466102624745018,
+      "loss": 2.1405,
+      "step": 6780
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00014629901229637878,
+      "loss": 2.1119,
+      "step": 6785
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00014598790221973247,
+      "loss": 2.1046,
+      "step": 6790
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00014567693305622377,
+      "loss": 2.1295,
+      "step": 6795
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00014536610561714558,
+      "loss": 2.0998,
+      "step": 6800
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001450554207134214,
+      "loss": 2.1193,
+      "step": 6805
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00014474487915560252,
+      "loss": 2.101,
+      "step": 6810
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001444344817538665,
+      "loss": 2.0741,
+      "step": 6815
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00014412422931801467,
+      "loss": 2.0999,
+      "step": 6820
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00014381412265747034,
+      "loss": 2.1131,
+      "step": 6825
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014350416258127624,
+      "loss": 2.0859,
+      "step": 6830
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014319434989809287,
+      "loss": 2.0994,
+      "step": 6835
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00014288468541619613,
+      "loss": 2.1219,
+      "step": 6840
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00014257516994347532,
+      "loss": 2.1255,
+      "step": 6845
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00014226580428743088,
+      "loss": 2.1276,
+      "step": 6850
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014195658925517246,
+      "loss": 2.0905,
+      "step": 6855
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001416475256534168,
+      "loss": 2.1211,
+      "step": 6860
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00014133861428848547,
+      "loss": 2.1013,
+      "step": 6865
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00014102985596630297,
+      "loss": 2.0874,
+      "step": 6870
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00014072125149239438,
+      "loss": 2.1041,
+      "step": 6875
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00014041280167188352,
+      "loss": 2.1036,
+      "step": 6880
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001401045073094907,
+      "loss": 2.1149,
+      "step": 6885
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001397963692095307,
+      "loss": 2.0599,
+      "step": 6890
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001394883881759105,
+      "loss": 2.1005,
+      "step": 6895
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001391805650121274,
+      "loss": 2.1037,
+      "step": 6900
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00013887290052126683,
+      "loss": 2.0939,
+      "step": 6905
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00013856539550600026,
+      "loss": 2.0898,
+      "step": 6910
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00013825805076858299,
+      "loss": 2.0876,
+      "step": 6915
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00013795086711085235,
+      "loss": 2.0711,
+      "step": 6920
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00013764384533422526,
+      "loss": 2.1145,
+      "step": 6925
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001373369862396965,
+      "loss": 2.1386,
+      "step": 6930
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00013703029062783615,
+      "loss": 2.1496,
+      "step": 6935
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00013672375929878801,
+      "loss": 2.1117,
+      "step": 6940
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00013641739305226723,
+      "loss": 2.1243,
+      "step": 6945
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013611119268755825,
+      "loss": 2.0966,
+      "step": 6950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00013580515900351258,
+      "loss": 2.0824,
+      "step": 6955
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00013549929279854718,
+      "loss": 2.1118,
+      "step": 6960
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001351935948706419,
+      "loss": 2.1052,
+      "step": 6965
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001348880660173376,
+      "loss": 2.1367,
+      "step": 6970
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013458270703573387,
+      "loss": 2.0904,
+      "step": 6975
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013427751872248744,
+      "loss": 2.0932,
+      "step": 6980
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013397250187380957,
+      "loss": 2.1145,
+      "step": 6985
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001336676572854643,
+      "loss": 2.1086,
+      "step": 6990
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000133362985752766,
+      "loss": 2.1105,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013305848807057782,
+      "loss": 2.1029,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001327541650333093,
+      "loss": 2.0997,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00013245001743491437,
+      "loss": 2.1089,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013214604606888902,
+      "loss": 2.1435,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00013184225172826973,
+      "loss": 2.1185,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00013153863520563101,
+      "loss": 2.1492,
+      "step": 7025
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013123519729308364,
+      "loss": 2.1143,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001309319387822721,
+      "loss": 2.0802,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013062886046437302,
+      "loss": 2.1069,
+      "step": 7040
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00013032596313009286,
+      "loss": 2.1407,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00013002324756966618,
+      "loss": 2.0927,
+      "step": 7050
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001297207145728528,
+      "loss": 2.1295,
+      "step": 7055
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001294183649289366,
+      "loss": 2.1743,
+      "step": 7060
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001291161994267231,
+      "loss": 2.075,
+      "step": 7065
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00012881421885453732,
+      "loss": 2.1188,
+      "step": 7070
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00012851242400022177,
+      "loss": 2.1046,
+      "step": 7075
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00012821081565113452,
+      "loss": 2.1022,
+      "step": 7080
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00012790939459414704,
+      "loss": 2.0762,
+      "step": 7085
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00012760816161564223,
+      "loss": 2.1178,
+      "step": 7090
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012730711750151213,
+      "loss": 2.1288,
+      "step": 7095
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012700626303715625,
+      "loss": 2.0667,
+      "step": 7100
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001267055990074792,
+      "loss": 2.0965,
+      "step": 7105
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00012640512619688888,
+      "loss": 2.1014,
+      "step": 7110
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012610484538929412,
+      "loss": 2.1169,
+      "step": 7115
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012580475736810302,
+      "loss": 2.1372,
+      "step": 7120
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001255048629162206,
+      "loss": 2.0957,
+      "step": 7125
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00012520516281604692,
+      "loss": 2.1229,
+      "step": 7130
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000124905657849475,
+      "loss": 2.1027,
+      "step": 7135
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012460634879788874,
+      "loss": 2.0916,
+      "step": 7140
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012430723644216092,
+      "loss": 2.1181,
+      "step": 7145
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012400832156265126,
+      "loss": 2.1138,
+      "step": 7150
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012370960493920404,
+      "loss": 2.1201,
+      "step": 7155
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00012341108735114653,
+      "loss": 2.0924,
+      "step": 7160
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00012311276957728665,
+      "loss": 2.1257,
+      "step": 7165
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00012281465239591106,
+      "loss": 2.1157,
+      "step": 7170
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00012251673658478305,
+      "loss": 2.1112,
+      "step": 7175
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00012221902292114053,
+      "loss": 2.097,
+      "step": 7180
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00012192151218169407,
+      "loss": 2.0966,
+      "step": 7185
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001216242051426248,
+      "loss": 2.0823,
+      "step": 7190
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00012132710257958248,
+      "loss": 2.1166,
+      "step": 7195
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00012103020526768327,
+      "loss": 2.1522,
+      "step": 7200
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00012073351398150796,
+      "loss": 2.1473,
+      "step": 7205
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00012043702949509983,
+      "loss": 2.1095,
+      "step": 7210
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00012014075258196266,
+      "loss": 2.107,
+      "step": 7215
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00011984468401505849,
+      "loss": 2.1213,
+      "step": 7220
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00011954882456680603,
+      "loss": 2.1018,
+      "step": 7225
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00011925317500907837,
+      "loss": 2.1361,
+      "step": 7230
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00011895773611320104,
+      "loss": 2.1327,
+      "step": 7235
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00011866250864994973,
+      "loss": 2.0931,
+      "step": 7240
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00011836749338954885,
+      "loss": 2.1342,
+      "step": 7245
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011807269110166897,
+      "loss": 2.0836,
+      "step": 7250
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011777810255542521,
+      "loss": 2.0761,
+      "step": 7255
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00011748372851937479,
+      "loss": 2.1275,
+      "step": 7260
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00011718956976151552,
+      "loss": 2.1099,
+      "step": 7265
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001168956270492835,
+      "loss": 2.0851,
+      "step": 7270
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00011660190114955129,
+      "loss": 2.0925,
+      "step": 7275
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00011630839282862546,
+      "loss": 2.0861,
+      "step": 7280
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00011601510285224539,
+      "loss": 2.1035,
+      "step": 7285
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00011572203198558043,
+      "loss": 2.0903,
+      "step": 7290
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00011542918099322862,
+      "loss": 2.0787,
+      "step": 7295
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00011513655063921412,
+      "loss": 2.1227,
+      "step": 7300
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00011484414168698547,
+      "loss": 2.0865,
+      "step": 7305
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001145519548994138,
+      "loss": 2.072,
+      "step": 7310
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011425999103879048,
+      "loss": 2.1382,
+      "step": 7315
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011396825086682516,
+      "loss": 2.094,
+      "step": 7320
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001136767351446442,
+      "loss": 2.1178,
+      "step": 7325
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00011338544463278815,
+      "loss": 2.092,
+      "step": 7330
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00011309438009121025,
+      "loss": 2.0974,
+      "step": 7335
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00011280354227927374,
+      "loss": 2.1064,
+      "step": 7340
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00011251293195575084,
+      "loss": 2.0531,
+      "step": 7345
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001122225498788201,
+      "loss": 2.1161,
+      "step": 7350
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011193239680606451,
+      "loss": 2.1124,
+      "step": 7355
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00011164247349446957,
+      "loss": 2.0824,
+      "step": 7360
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00011135278070042163,
+      "loss": 2.0939,
+      "step": 7365
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011106331917970527,
+      "loss": 2.0857,
+      "step": 7370
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00011077408968750214,
+      "loss": 2.1444,
+      "step": 7375
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00011048509297838791,
+      "loss": 2.1137,
+      "step": 7380
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00011019632980633156,
+      "loss": 2.0799,
+      "step": 7385
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010990780092469259,
+      "loss": 2.0915,
+      "step": 7390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001096195070862191,
+      "loss": 2.1333,
+      "step": 7395
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001093314490430461,
+      "loss": 2.1179,
+      "step": 7400
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010904362754669342,
+      "loss": 2.0775,
+      "step": 7405
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001087560433480637,
+      "loss": 2.1088,
+      "step": 7410
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010846869719744086,
+      "loss": 2.1282,
+      "step": 7415
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010818158984448703,
+      "loss": 2.0788,
+      "step": 7420
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010789472203824207,
+      "loss": 2.145,
+      "step": 7425
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010760809452712041,
+      "loss": 2.1578,
+      "step": 7430
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010732170805890987,
+      "loss": 2.0883,
+      "step": 7435
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00010703556338076924,
+      "loss": 2.1002,
+      "step": 7440
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010674966123922645,
+      "loss": 2.1353,
+      "step": 7445
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00010646400238017691,
+      "loss": 2.1553,
+      "step": 7450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00010617858754888103,
+      "loss": 2.1264,
+      "step": 7455
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010589341748996285,
+      "loss": 2.0861,
+      "step": 7460
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00010560849294740765,
+      "loss": 2.1787,
+      "step": 7465
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010532381466456009,
+      "loss": 2.0844,
+      "step": 7470
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010503938338412269,
+      "loss": 2.1122,
+      "step": 7475
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010475519984815325,
+      "loss": 2.1068,
+      "step": 7480
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00010447126479806326,
+      "loss": 2.124,
+      "step": 7485
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010418757897461614,
+      "loss": 2.1239,
+      "step": 7490
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010390414311792485,
+      "loss": 2.1004,
+      "step": 7495
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010362095796745039,
+      "loss": 2.0709,
+      "step": 7500
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00010333802426199963,
+      "loss": 2.1201,
+      "step": 7505
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00010305534273972326,
+      "loss": 2.1216,
+      "step": 7510
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00010277291413811445,
+      "loss": 2.1217,
+      "step": 7515
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010249073919400611,
+      "loss": 2.0801,
+      "step": 7520
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010220881864356955,
+      "loss": 2.0881,
+      "step": 7525
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00010192715322231253,
+      "loss": 2.1399,
+      "step": 7530
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00010164574366507685,
+      "loss": 2.1091,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00010136459070603719,
+      "loss": 2.1038,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010108369507869848,
+      "loss": 2.0849,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010080305751589434,
+      "loss": 2.1102,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010052267874978531,
+      "loss": 2.1513,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010024255951185657,
+      "loss": 2.0989,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.996270053291614e-05,
+      "loss": 2.1153,
+      "step": 7565
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.968310254309335e-05,
+      "loss": 2.1246,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.940376627183628e-05,
+      "loss": 2.1371,
+      "step": 7575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.912469244791054e-05,
+      "loss": 2.1117,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.884588179939677e-05,
+      "loss": 2.1116,
+      "step": 7585
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.856733505368905e-05,
+      "loss": 2.0977,
+      "step": 7590
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.828905293749318e-05,
+      "loss": 2.0909,
+      "step": 7595
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.80110361768243e-05,
+      "loss": 2.1109,
+      "step": 7600
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.224609375,
+      "learning_rate": 9.773328549700526e-05,
+      "loss": 2.0799,
+      "step": 7605
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.745580162266505e-05,
+      "loss": 2.1182,
+      "step": 7610
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.717858527773616e-05,
+      "loss": 2.1301,
+      "step": 7615
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.690163718545356e-05,
+      "loss": 2.1144,
+      "step": 7620
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.662495806835179e-05,
+      "loss": 2.0999,
+      "step": 7625
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.634854864826421e-05,
+      "loss": 2.1331,
+      "step": 7630
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.607240964632039e-05,
+      "loss": 2.1587,
+      "step": 7635
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.579654178294429e-05,
+      "loss": 2.1276,
+      "step": 7640
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.552094577785255e-05,
+      "loss": 2.1238,
+      "step": 7645
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.524562235005252e-05,
+      "loss": 2.1393,
+      "step": 7650
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.497057221784051e-05,
+      "loss": 2.0863,
+      "step": 7655
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.469579609879993e-05,
+      "loss": 2.1032,
+      "step": 7660
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.442129470979888e-05,
+      "loss": 2.09,
+      "step": 7665
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.414706876698922e-05,
+      "loss": 2.0932,
+      "step": 7670
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.387311898580378e-05,
+      "loss": 2.098,
+      "step": 7675
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.359944608095528e-05,
+      "loss": 2.0938,
+      "step": 7680
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.234375,
+      "learning_rate": 9.332605076643379e-05,
+      "loss": 2.0881,
+      "step": 7685
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.305293375550525e-05,
+      "loss": 2.1616,
+      "step": 7690
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.278009576070963e-05,
+      "loss": 2.0904,
+      "step": 7695
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.250753749385903e-05,
+      "loss": 2.0911,
+      "step": 7700
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.223525966603536e-05,
+      "loss": 2.1187,
+      "step": 7705
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.196326298758937e-05,
+      "loss": 2.1144,
+      "step": 7710
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.169154816813798e-05,
+      "loss": 2.0919,
+      "step": 7715
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.142011591656301e-05,
+      "loss": 2.0698,
+      "step": 7720
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.114896694100891e-05,
+      "loss": 2.0918,
+      "step": 7725
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.087810194888107e-05,
+      "loss": 2.1143,
+      "step": 7730
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.06075216468442e-05,
+      "loss": 2.1348,
+      "step": 7735
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.033722674082006e-05,
+      "loss": 2.119,
+      "step": 7740
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.006721793598584e-05,
+      "loss": 2.1144,
+      "step": 7745
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.979749593677253e-05,
+      "loss": 2.0731,
+      "step": 7750
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.952806144686261e-05,
+      "loss": 2.1358,
+      "step": 7755
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.92589151691887e-05,
+      "loss": 2.1075,
+      "step": 7760
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.89900578059313e-05,
+      "loss": 2.0871,
+      "step": 7765
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.872149005851722e-05,
+      "loss": 2.107,
+      "step": 7770
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.845321262761781e-05,
+      "loss": 2.11,
+      "step": 7775
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.818522621314678e-05,
+      "loss": 2.1028,
+      "step": 7780
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.791753151425888e-05,
+      "loss": 2.1378,
+      "step": 7785
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.765012922934755e-05,
+      "loss": 2.1208,
+      "step": 7790
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.738302005604335e-05,
+      "loss": 2.1132,
+      "step": 7795
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.711620469121245e-05,
+      "loss": 2.1078,
+      "step": 7800
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.684968383095411e-05,
+      "loss": 2.0819,
+      "step": 7805
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.658345817059941e-05,
+      "loss": 2.1066,
+      "step": 7810
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.631752840470941e-05,
+      "loss": 2.1254,
+      "step": 7815
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.605189522707291e-05,
+      "loss": 2.0947,
+      "step": 7820
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.57865593307053e-05,
+      "loss": 2.1196,
+      "step": 7825
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.552152140784607e-05,
+      "loss": 2.1187,
+      "step": 7830
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.525678214995738e-05,
+      "loss": 2.0997,
+      "step": 7835
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.499234224772246e-05,
+      "loss": 2.0727,
+      "step": 7840
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.47282023910432e-05,
+      "loss": 2.1121,
+      "step": 7845
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.446436326903882e-05,
+      "loss": 2.133,
+      "step": 7850
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.420082557004409e-05,
+      "loss": 2.0483,
+      "step": 7855
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.393758998160717e-05,
+      "loss": 2.1356,
+      "step": 7860
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.367465719048831e-05,
+      "loss": 2.1049,
+      "step": 7865
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.341202788265734e-05,
+      "loss": 2.0773,
+      "step": 7870
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.314970274329277e-05,
+      "loss": 2.117,
+      "step": 7875
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.288768245677945e-05,
+      "loss": 2.131,
+      "step": 7880
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.262596770670676e-05,
+      "loss": 2.0954,
+      "step": 7885
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.236455917586699e-05,
+      "loss": 2.0664,
+      "step": 7890
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.21034575462537e-05,
+      "loss": 2.1112,
+      "step": 7895
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.184266349905951e-05,
+      "loss": 2.1066,
+      "step": 7900
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.158217771467496e-05,
+      "loss": 2.1083,
+      "step": 7905
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.132200087268577e-05,
+      "loss": 2.1021,
+      "step": 7910
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.106213365187223e-05,
+      "loss": 2.1463,
+      "step": 7915
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.080257673020663e-05,
+      "loss": 2.1233,
+      "step": 7920
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.054333078485168e-05,
+      "loss": 2.1229,
+      "step": 7925
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.028439649215877e-05,
+      "loss": 2.104,
+      "step": 7930
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.002577452766623e-05,
+      "loss": 2.0927,
+      "step": 7935
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.976746556609759e-05,
+      "loss": 2.0963,
+      "step": 7940
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 7.950947028135997e-05,
+      "loss": 2.1209,
+      "step": 7945
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.925178934654161e-05,
+      "loss": 2.0654,
+      "step": 7950
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.899442343391114e-05,
+      "loss": 2.0789,
+      "step": 7955
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.873737321491506e-05,
+      "loss": 2.1175,
+      "step": 7960
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.848063936017642e-05,
+      "loss": 2.086,
+      "step": 7965
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 7.822422253949278e-05,
+      "loss": 2.1389,
+      "step": 7970
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.79681234218345e-05,
+      "loss": 2.0908,
+      "step": 7975
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.771234267534337e-05,
+      "loss": 2.0564,
+      "step": 7980
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.745688096733032e-05,
+      "loss": 2.1368,
+      "step": 7985
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.720173896427394e-05,
+      "loss": 2.0946,
+      "step": 7990
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.694691733181896e-05,
+      "loss": 2.1287,
+      "step": 7995
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.669241673477399e-05,
+      "loss": 2.1275,
+      "step": 8000
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.643823783711037e-05,
+      "loss": 2.0847,
+      "step": 8005
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.618438130195998e-05,
+      "loss": 2.065,
+      "step": 8010
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.593084779161359e-05,
+      "loss": 2.1041,
+      "step": 8015
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.567763796751954e-05,
+      "loss": 2.1025,
+      "step": 8020
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.542475249028145e-05,
+      "loss": 2.069,
+      "step": 8025
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.51721920196567e-05,
+      "loss": 2.074,
+      "step": 8030
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.491995721455505e-05,
+      "loss": 2.1351,
+      "step": 8035
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2421875,
+      "learning_rate": 7.466804873303625e-05,
+      "loss": 2.1296,
+      "step": 8040
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.441646723230906e-05,
+      "loss": 2.1207,
+      "step": 8045
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.416521336872894e-05,
+      "loss": 2.0797,
+      "step": 8050
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.391428779779655e-05,
+      "loss": 2.1238,
+      "step": 8055
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.366369117415626e-05,
+      "loss": 2.0916,
+      "step": 8060
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.341342415159403e-05,
+      "loss": 2.0776,
+      "step": 8065
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.316348738303611e-05,
+      "loss": 2.14,
+      "step": 8070
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.224609375,
+      "learning_rate": 7.291388152054704e-05,
+      "loss": 2.0778,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.266460721532793e-05,
+      "loss": 2.0561,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.241566511771521e-05,
+      "loss": 2.1152,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 7.216705587717831e-05,
+      "loss": 2.12,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.191878014231833e-05,
+      "loss": 2.0983,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.16708385608664e-05,
+      "loss": 2.0884,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.142323177968173e-05,
+      "loss": 2.1197,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.117596044475022e-05,
+      "loss": 2.1033,
+      "step": 8110
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.092902520118243e-05,
+      "loss": 2.1059,
+      "step": 8115
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.068242669321215e-05,
+      "loss": 2.1148,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.043616556419479e-05,
+      "loss": 2.1426,
+      "step": 8125
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.019024245660542e-05,
+      "loss": 2.0757,
+      "step": 8130
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.99446580120372e-05,
+      "loss": 2.1089,
+      "step": 8135
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.969941287119994e-05,
+      "loss": 2.0967,
+      "step": 8140
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.945450767391799e-05,
+      "loss": 2.0861,
+      "step": 8145
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.920994305912915e-05,
+      "loss": 2.1158,
+      "step": 8150
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.896571966488219e-05,
+      "loss": 2.0883,
+      "step": 8155
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.872183812833608e-05,
+      "loss": 2.1185,
+      "step": 8160
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.847829908575782e-05,
+      "loss": 2.1195,
+      "step": 8165
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.823510317252075e-05,
+      "loss": 2.1461,
+      "step": 8170
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.79922510231031e-05,
+      "loss": 2.0658,
+      "step": 8175
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.774974327108614e-05,
+      "loss": 2.0876,
+      "step": 8180
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.750758054915278e-05,
+      "loss": 2.0941,
+      "step": 8185
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.726576348908589e-05,
+      "loss": 2.1018,
+      "step": 8190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.702429272176607e-05,
+      "loss": 2.117,
+      "step": 8195
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.6783168877171e-05,
+      "loss": 2.1286,
+      "step": 8200
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.654239258437285e-05,
+      "loss": 2.0733,
+      "step": 8205
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.63019644715374e-05,
+      "loss": 2.0924,
+      "step": 8210
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.606188516592185e-05,
+      "loss": 2.1021,
+      "step": 8215
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.582215529387335e-05,
+      "loss": 2.1414,
+      "step": 8220
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.558277548082754e-05,
+      "loss": 2.1095,
+      "step": 8225
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.534374635130692e-05,
+      "loss": 2.0879,
+      "step": 8230
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.510506852891858e-05,
+      "loss": 2.086,
+      "step": 8235
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.486674263635364e-05,
+      "loss": 2.0797,
+      "step": 8240
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.462876929538468e-05,
+      "loss": 2.1481,
+      "step": 8245
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.439114912686479e-05,
+      "loss": 2.1469,
+      "step": 8250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.415388275072544e-05,
+      "loss": 2.1036,
+      "step": 8255
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.391697078597512e-05,
+      "loss": 2.0785,
+      "step": 8260
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.368041385069785e-05,
+      "loss": 2.0955,
+      "step": 8265
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.344421256205122e-05,
+      "loss": 2.1202,
+      "step": 8270
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.3208367536265e-05,
+      "loss": 2.1314,
+      "step": 8275
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.297287938863963e-05,
+      "loss": 2.1448,
+      "step": 8280
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.273774873354427e-05,
+      "loss": 2.079,
+      "step": 8285
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.25029761844157e-05,
+      "loss": 2.0911,
+      "step": 8290
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.226856235375615e-05,
+      "loss": 2.0911,
+      "step": 8295
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.203450785313208e-05,
+      "loss": 2.1259,
+      "step": 8300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.180081329317259e-05,
+      "loss": 2.1198,
+      "step": 8305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.15674792835676e-05,
+      "loss": 2.1614,
+      "step": 8310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.133450643306637e-05,
+      "loss": 2.0845,
+      "step": 8315
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.110189534947606e-05,
+      "loss": 2.1205,
+      "step": 8320
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.086964663965984e-05,
+      "loss": 2.096,
+      "step": 8325
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.063776090953568e-05,
+      "loss": 2.1206,
+      "step": 8330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.0406238764074386e-05,
+      "loss": 2.0934,
+      "step": 8335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.017508080729819e-05,
+      "loss": 2.087,
+      "step": 8340
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.9944287642279415e-05,
+      "loss": 2.1131,
+      "step": 8345
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.971385987113836e-05,
+      "loss": 2.128,
+      "step": 8350
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.9483798095042364e-05,
+      "loss": 2.1452,
+      "step": 8355
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.9254102914203634e-05,
+      "loss": 2.1297,
+      "step": 8360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.9024774927878054e-05,
+      "loss": 2.0584,
+      "step": 8365
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.879581473436368e-05,
+      "loss": 2.1078,
+      "step": 8370
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.856722293099881e-05,
+      "loss": 2.1047,
+      "step": 8375
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.8339000114160666e-05,
+      "loss": 2.0854,
+      "step": 8380
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.811114687926402e-05,
+      "loss": 2.1011,
+      "step": 8385
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.788366382075916e-05,
+      "loss": 2.1448,
+      "step": 8390
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.7656551532130945e-05,
+      "loss": 2.1345,
+      "step": 8395
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.742981060589643e-05,
+      "loss": 2.0944,
+      "step": 8400
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.7203441633604294e-05,
+      "loss": 2.0984,
+      "step": 8405
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.697744520583268e-05,
+      "loss": 2.086,
+      "step": 8410
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.675182191218769e-05,
+      "loss": 2.1013,
+      "step": 8415
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.652657234130194e-05,
+      "loss": 2.0664,
+      "step": 8420
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.630169708083324e-05,
+      "loss": 2.1041,
+      "step": 8425
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.607719671746259e-05,
+      "loss": 2.0922,
+      "step": 8430
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.5853071836893214e-05,
+      "loss": 2.1008,
+      "step": 8435
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.562932302384834e-05,
+      "loss": 2.0781,
+      "step": 8440
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.5405950862070385e-05,
+      "loss": 2.1269,
+      "step": 8445
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.51829559343191e-05,
+      "loss": 2.117,
+      "step": 8450
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.49603388223699e-05,
+      "loss": 2.146,
+      "step": 8455
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.4738100107012615e-05,
+      "loss": 2.0968,
+      "step": 8460
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.451624036804976e-05,
+      "loss": 2.1071,
+      "step": 8465
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.429476018429529e-05,
+      "loss": 2.151,
+      "step": 8470
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.407366013357298e-05,
+      "loss": 2.1123,
+      "step": 8475
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.3852940792714476e-05,
+      "loss": 2.1213,
+      "step": 8480
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.363260273755866e-05,
+      "loss": 2.1184,
+      "step": 8485
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.341264654294935e-05,
+      "loss": 2.1055,
+      "step": 8490
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.319307278273433e-05,
+      "loss": 2.1069,
+      "step": 8495
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.297388202976352e-05,
+      "loss": 2.1162,
+      "step": 8500
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.275507485588753e-05,
+      "loss": 2.104,
+      "step": 8505
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.253665183195653e-05,
+      "loss": 2.1353,
+      "step": 8510
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.231861352781817e-05,
+      "loss": 2.0981,
+      "step": 8515
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.21009605123165e-05,
+      "loss": 2.106,
+      "step": 8520
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.188369335329051e-05,
+      "loss": 2.1331,
+      "step": 8525
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.1666812617572316e-05,
+      "loss": 2.0954,
+      "step": 8530
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.145031887098606e-05,
+      "loss": 2.1032,
+      "step": 8535
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.1234212678346184e-05,
+      "loss": 2.0879,
+      "step": 8540
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.1018494603455934e-05,
+      "loss": 2.0836,
+      "step": 8545
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.080316520910622e-05,
+      "loss": 2.1151,
+      "step": 8550
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.0588225057073744e-05,
+      "loss": 2.1116,
+      "step": 8555
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.037367470811969e-05,
+      "loss": 2.1097,
+      "step": 8560
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.015951472198845e-05,
+      "loss": 2.1463,
+      "step": 8565
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.994574565740577e-05,
+      "loss": 2.1102,
+      "step": 8570
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.973236807207773e-05,
+      "loss": 2.1069,
+      "step": 8575
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.9519382522688903e-05,
+      "loss": 2.1303,
+      "step": 8580
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.930678956490109e-05,
+      "loss": 2.1173,
+      "step": 8585
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.909458975335202e-05,
+      "loss": 2.1001,
+      "step": 8590
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.888278364165355e-05,
+      "loss": 2.0999,
+      "step": 8595
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.8671371782390405e-05,
+      "loss": 2.1242,
+      "step": 8600
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.846035472711894e-05,
+      "loss": 2.1004,
+      "step": 8605
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.824973302636522e-05,
+      "loss": 2.1379,
+      "step": 8610
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.803950722962414e-05,
+      "loss": 2.1155,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.7829677885357526e-05,
+      "loss": 2.0763,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.762024554099285e-05,
+      "loss": 2.1104,
+      "step": 8625
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.741121074292209e-05,
+      "loss": 2.1256,
+      "step": 8630
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.720257403649977e-05,
+      "loss": 2.085,
+      "step": 8635
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.69943359660421e-05,
+      "loss": 2.1261,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.6786497074825056e-05,
+      "loss": 2.1117,
+      "step": 8645
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.657905790508324e-05,
+      "loss": 2.1095,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.637201899800856e-05,
+      "loss": 2.0897,
+      "step": 8655
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.616538089374851e-05,
+      "loss": 2.0951,
+      "step": 8660
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.5959144131404896e-05,
+      "loss": 2.1346,
+      "step": 8665
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.575330924903265e-05,
+      "loss": 2.125,
+      "step": 8670
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.554787678363801e-05,
+      "loss": 2.1235,
+      "step": 8675
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.5342847271177616e-05,
+      "loss": 2.0816,
+      "step": 8680
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.513822124655644e-05,
+      "loss": 2.133,
+      "step": 8685
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.493399924362713e-05,
+      "loss": 2.0935,
+      "step": 8690
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.47301817951882e-05,
+      "loss": 2.1127,
+      "step": 8695
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.452676943298264e-05,
+      "loss": 2.1171,
+      "step": 8700
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.4323762687696604e-05,
+      "loss": 2.1185,
+      "step": 8705
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.4121162088958026e-05,
+      "loss": 2.088,
+      "step": 8710
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.391896816533532e-05,
+      "loss": 2.1312,
+      "step": 8715
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.3717181444335964e-05,
+      "loss": 2.0713,
+      "step": 8720
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.351580245240476e-05,
+      "loss": 2.1042,
+      "step": 8725
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.331483171492319e-05,
+      "loss": 2.0868,
+      "step": 8730
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.3114269756207315e-05,
+      "loss": 2.0766,
+      "step": 8735
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.2914117099506926e-05,
+      "loss": 2.1004,
+      "step": 8740
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.271437426700391e-05,
+      "loss": 2.0826,
+      "step": 8745
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.251504177981085e-05,
+      "loss": 2.1315,
+      "step": 8750
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.231612015796995e-05,
+      "loss": 2.1251,
+      "step": 8755
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.211760992045153e-05,
+      "loss": 2.1379,
+      "step": 8760
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.19195115851523e-05,
+      "loss": 2.0842,
+      "step": 8765
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.1721825668894776e-05,
+      "loss": 2.1471,
+      "step": 8770
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.152455268742517e-05,
+      "loss": 2.0909,
+      "step": 8775
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.132769315541267e-05,
+      "loss": 2.1121,
+      "step": 8780
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.1131247586447597e-05,
+      "loss": 2.1032,
+      "step": 8785
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.0935216493040285e-05,
+      "loss": 2.0749,
+      "step": 8790
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.07396003866199e-05,
+      "loss": 2.107,
+      "step": 8795
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.054439977753282e-05,
+      "loss": 2.1109,
+      "step": 8800
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.034961517504132e-05,
+      "loss": 2.1148,
+      "step": 8805
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.015524708732263e-05,
+      "loss": 2.1391,
+      "step": 8810
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.9961296021467056e-05,
+      "loss": 2.1037,
+      "step": 8815
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.976776248347709e-05,
+      "loss": 2.1215,
+      "step": 8820
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.9574646978265874e-05,
+      "loss": 2.0873,
+      "step": 8825
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.938195000965588e-05,
+      "loss": 2.0711,
+      "step": 8830
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.918967208037778e-05,
+      "loss": 2.0757,
+      "step": 8835
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.8997813692068944e-05,
+      "loss": 2.0977,
+      "step": 8840
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.880637534527207e-05,
+      "loss": 2.1179,
+      "step": 8845
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.861535753943426e-05,
+      "loss": 2.0717,
+      "step": 8850
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.842476077290522e-05,
+      "loss": 2.0851,
+      "step": 8855
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.8234585542936376e-05,
+      "loss": 2.1166,
+      "step": 8860
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.804483234567928e-05,
+      "loss": 2.0894,
+      "step": 8865
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.7855501676184416e-05,
+      "loss": 2.0819,
+      "step": 8870
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.766659402840011e-05,
+      "loss": 2.1043,
+      "step": 8875
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.747810989517087e-05,
+      "loss": 2.1277,
+      "step": 8880
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.7290049768236316e-05,
+      "loss": 2.1355,
+      "step": 8885
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.710241413822999e-05,
+      "loss": 2.0989,
+      "step": 8890
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.691520349467781e-05,
+      "loss": 2.1278,
+      "step": 8895
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.672841832599712e-05,
+      "loss": 2.112,
+      "step": 8900
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.6542059119494954e-05,
+      "loss": 2.1045,
+      "step": 8905
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.635612636136727e-05,
+      "loss": 2.0608,
+      "step": 8910
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.617062053669749e-05,
+      "loss": 2.1293,
+      "step": 8915
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.5985542129455e-05,
+      "loss": 2.1204,
+      "step": 8920
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.580089162249429e-05,
+      "loss": 2.1028,
+      "step": 8925
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.5616669497553336e-05,
+      "loss": 2.1302,
+      "step": 8930
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.543287623525256e-05,
+      "loss": 2.0837,
+      "step": 8935
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.524951231509359e-05,
+      "loss": 2.1037,
+      "step": 8940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.506657821545785e-05,
+      "loss": 2.1115,
+      "step": 8945
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.488407441360535e-05,
+      "loss": 2.1232,
+      "step": 8950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.4702001385673634e-05,
+      "loss": 2.1757,
+      "step": 8955
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.452035960667623e-05,
+      "loss": 2.1113,
+      "step": 8960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.4339149550501836e-05,
+      "loss": 2.1785,
+      "step": 8965
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.415837168991238e-05,
+      "loss": 2.0961,
+      "step": 8970
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.397802649654265e-05,
+      "loss": 2.0961,
+      "step": 8975
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.379811444089849e-05,
+      "loss": 2.1169,
+      "step": 8980
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.3618635992355686e-05,
+      "loss": 2.1381,
+      "step": 8985
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.343959161915877e-05,
+      "loss": 2.1424,
+      "step": 8990
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.326098178841983e-05,
+      "loss": 2.1002,
+      "step": 8995
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.30828069661173e-05,
+      "loss": 2.1073,
+      "step": 9000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.2905067617094796e-05,
+      "loss": 2.1393,
+      "step": 9005
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.272776420505954e-05,
+      "loss": 2.1097,
+      "step": 9010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.255089719258173e-05,
+      "loss": 2.0935,
+      "step": 9015
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.237446704109284e-05,
+      "loss": 2.1066,
+      "step": 9020
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.219847421088473e-05,
+      "loss": 2.0793,
+      "step": 9025
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.202291916110827e-05,
+      "loss": 2.1244,
+      "step": 9030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.184780234977214e-05,
+      "loss": 2.0899,
+      "step": 9035
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.1673124233741824e-05,
+      "loss": 2.0989,
+      "step": 9040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.149888526873821e-05,
+      "loss": 2.1033,
+      "step": 9045
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.132508590933638e-05,
+      "loss": 2.1003,
+      "step": 9050
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.115172660896477e-05,
+      "loss": 2.1204,
+      "step": 9055
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.097880781990348e-05,
+      "loss": 2.1247,
+      "step": 9060
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.080632999328352e-05,
+      "loss": 2.0663,
+      "step": 9065
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.0634293579085404e-05,
+      "loss": 2.1224,
+      "step": 9070
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.0462699026137966e-05,
+      "loss": 2.1124,
+      "step": 9075
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.0291546782117452e-05,
+      "loss": 2.0772,
+      "step": 9080
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.012083729354598e-05,
+      "loss": 2.0876,
+      "step": 9085
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.9950571005790573e-05,
+      "loss": 2.0978,
+      "step": 9090
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.978074836306213e-05,
+      "loss": 2.1272,
+      "step": 9095
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.9611369808413904e-05,
+      "loss": 2.1163,
+      "step": 9100
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.9442435783740775e-05,
+      "loss": 2.1526,
+      "step": 9105
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.9273946729777702e-05,
+      "loss": 2.1216,
+      "step": 9110
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.9105903086098775e-05,
+      "loss": 2.1364,
+      "step": 9115
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.893830529111621e-05,
+      "loss": 2.0959,
+      "step": 9120
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.8771153782078864e-05,
+      "loss": 2.0946,
+      "step": 9125
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.8604448995071285e-05,
+      "loss": 2.1365,
+      "step": 9130
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.8438191365012667e-05,
+      "loss": 2.14,
+      "step": 9135
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.8272381325655483e-05,
+      "loss": 2.0476,
+      "step": 9140
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.8107019309584636e-05,
+      "loss": 2.0812,
+      "step": 9145
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.794210574821601e-05,
+      "loss": 2.0666,
+      "step": 9150
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.7777641071795545e-05,
+      "loss": 2.1169,
+      "step": 9155
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.761362570939816e-05,
+      "loss": 2.0937,
+      "step": 9160
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.74500600889265e-05,
+      "loss": 2.0642,
+      "step": 9165
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.7286944637109812e-05,
+      "loss": 2.1106,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.7124279779503005e-05,
+      "loss": 2.1315,
+      "step": 9175
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.6962065940485294e-05,
+      "loss": 2.0988,
+      "step": 9180
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.680030354325944e-05,
+      "loss": 2.0955,
+      "step": 9185
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.663899300985009e-05,
+      "loss": 2.1067,
+      "step": 9190
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.647813476110328e-05,
+      "loss": 2.1418,
+      "step": 9195
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.6317729216685095e-05,
+      "loss": 2.0693,
+      "step": 9200
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.6157776795080357e-05,
+      "loss": 2.1104,
+      "step": 9205
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.5998277913591863e-05,
+      "loss": 2.0949,
+      "step": 9210
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.583923298833906e-05,
+      "loss": 2.1104,
+      "step": 9215
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.5680642434257208e-05,
+      "loss": 2.113,
+      "step": 9220
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.552250666509606e-05,
+      "loss": 2.1012,
+      "step": 9225
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.536482609341888e-05,
+      "loss": 2.0663,
+      "step": 9230
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.5207601130601322e-05,
+      "loss": 2.1224,
+      "step": 9235
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.5050832186830442e-05,
+      "loss": 2.1241,
+      "step": 9240
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4894519671103546e-05,
+      "loss": 2.1329,
+      "step": 9245
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.473866399122733e-05,
+      "loss": 2.1264,
+      "step": 9250
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.458326555381627e-05,
+      "loss": 2.091,
+      "step": 9255
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.4428324764292265e-05,
+      "loss": 2.1031,
+      "step": 9260
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.4273842026883165e-05,
+      "loss": 2.1216,
+      "step": 9265
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.4119817744621777e-05,
+      "loss": 2.0704,
+      "step": 9270
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.3966252319344774e-05,
+      "loss": 2.0667,
+      "step": 9275
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.3813146151691766e-05,
+      "loss": 2.1076,
+      "step": 9280
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.3660499641104217e-05,
+      "loss": 2.1104,
+      "step": 9285
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.3508313185824472e-05,
+      "loss": 2.131,
+      "step": 9290
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.3356587182894353e-05,
+      "loss": 2.0922,
+      "step": 9295
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.3205322028154687e-05,
+      "loss": 2.0962,
+      "step": 9300
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.3054518116243773e-05,
+      "loss": 2.0909,
+      "step": 9305
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.2904175840596742e-05,
+      "loss": 2.0702,
+      "step": 9310
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.2754295593444262e-05,
+      "loss": 2.1061,
+      "step": 9315
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.2604877765811573e-05,
+      "loss": 2.1015,
+      "step": 9320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.24559227475176e-05,
+      "loss": 2.1243,
+      "step": 9325
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.2307430927173755e-05,
+      "loss": 2.1422,
+      "step": 9330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.2159402692182974e-05,
+      "loss": 2.0982,
+      "step": 9335
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.2011838428738908e-05,
+      "loss": 2.1523,
+      "step": 9340
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.186473852182451e-05,
+      "loss": 2.0777,
+      "step": 9345
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.1718103355211496e-05,
+      "loss": 2.068,
+      "step": 9350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.1571933311458924e-05,
+      "loss": 2.1138,
+      "step": 9355
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.142622877191247e-05,
+      "loss": 2.1396,
+      "step": 9360
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.128099011670339e-05,
+      "loss": 2.1507,
+      "step": 9365
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.1136217724747386e-05,
+      "loss": 2.0835,
+      "step": 9370
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.0991911973743748e-05,
+      "loss": 2.0951,
+      "step": 9375
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.0848073240174438e-05,
+      "loss": 2.0804,
+      "step": 9380
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.0704701899302827e-05,
+      "loss": 2.111,
+      "step": 9385
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.0561798325173067e-05,
+      "loss": 2.1172,
+      "step": 9390
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.041936289060886e-05,
+      "loss": 2.1233,
+      "step": 9395
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.0277395967212543e-05,
+      "loss": 2.155,
+      "step": 9400
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.0135897925364277e-05,
+      "loss": 2.0959,
+      "step": 9405
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.9994869134220772e-05,
+      "loss": 2.1398,
+      "step": 9410
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.98543099617146e-05,
+      "loss": 2.0972,
+      "step": 9415
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.971422077455316e-05,
+      "loss": 2.1109,
+      "step": 9420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.957460193821763e-05,
+      "loss": 2.0976,
+      "step": 9425
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.943545381696219e-05,
+      "loss": 2.1277,
+      "step": 9430
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.9296776773812762e-05,
+      "loss": 2.0844,
+      "step": 9435
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.9158571170566453e-05,
+      "loss": 2.0883,
+      "step": 9440
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.902083736779039e-05,
+      "loss": 2.149,
+      "step": 9445
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.888357572482076e-05,
+      "loss": 2.1262,
+      "step": 9450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.8746786599761878e-05,
+      "loss": 2.1061,
+      "step": 9455
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.8610470349485444e-05,
+      "loss": 2.0833,
+      "step": 9460
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.8474627329629348e-05,
+      "loss": 2.0643,
+      "step": 9465
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.8339257894596962e-05,
+      "loss": 2.1025,
+      "step": 9470
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.8204362397555917e-05,
+      "loss": 2.0692,
+      "step": 9475
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.806994119043761e-05,
+      "loss": 2.14,
+      "step": 9480
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.7935994623935957e-05,
+      "loss": 2.1017,
+      "step": 9485
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.7802523047506582e-05,
+      "loss": 2.0842,
+      "step": 9490
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.7669526809365845e-05,
+      "loss": 2.0907,
+      "step": 9495
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.7537006256489997e-05,
+      "loss": 2.096,
+      "step": 9500
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.740496173461439e-05,
+      "loss": 2.1358,
+      "step": 9505
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.7273393588232344e-05,
+      "loss": 2.0871,
+      "step": 9510
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.7142302160594383e-05,
+      "loss": 2.1203,
+      "step": 9515
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.7011687793707252e-05,
+      "loss": 2.0944,
+      "step": 9520
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.6881550828333137e-05,
+      "loss": 2.1036,
+      "step": 9525
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.6751891603988735e-05,
+      "loss": 2.1204,
+      "step": 9530
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.6622710458944413e-05,
+      "loss": 2.16,
+      "step": 9535
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.649400773022305e-05,
+      "loss": 2.1025,
+      "step": 9540
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.6365783753599608e-05,
+      "loss": 2.129,
+      "step": 9545
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.6238038863599912e-05,
+      "loss": 2.0574,
+      "step": 9550
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.6110773393499934e-05,
+      "loss": 2.0829,
+      "step": 9555
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5983987675324853e-05,
+      "loss": 2.0842,
+      "step": 9560
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5857682039848187e-05,
+      "loss": 2.094,
+      "step": 9565
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.5731856816591017e-05,
+      "loss": 2.1373,
+      "step": 9570
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.5606512333821065e-05,
+      "loss": 2.0888,
+      "step": 9575
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.548164891855175e-05,
+      "loss": 2.1476,
+      "step": 9580
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.5357266896541555e-05,
+      "loss": 2.0712,
+      "step": 9585
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.5233366592292909e-05,
+      "loss": 2.1163,
+      "step": 9590
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.510994832905166e-05,
+      "loss": 2.1408,
+      "step": 9595
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.4987012428805891e-05,
+      "loss": 2.1092,
+      "step": 9600
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.4864559212285245e-05,
+      "loss": 2.0868,
+      "step": 9605
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.4742588998960194e-05,
+      "loss": 2.1332,
+      "step": 9610
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4621102107041041e-05,
+      "loss": 2.0777,
+      "step": 9615
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.4500098853477052e-05,
+      "loss": 2.1007,
+      "step": 9620
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.437957955395588e-05,
+      "loss": 2.1006,
+      "step": 9625
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4259544522902458e-05,
+      "loss": 2.0954,
+      "step": 9630
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.4139994073478369e-05,
+      "loss": 2.0669,
+      "step": 9635
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4020928517580944e-05,
+      "loss": 2.1218,
+      "step": 9640
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.3902348165842438e-05,
+      "loss": 2.1473,
+      "step": 9645
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.3784253327629316e-05,
+      "loss": 2.1085,
+      "step": 9650
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.3666644311041299e-05,
+      "loss": 2.0782,
+      "step": 9655
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.3549521422910639e-05,
+      "loss": 2.1336,
+      "step": 9660
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.343288496880144e-05,
+      "loss": 2.1178,
+      "step": 9665
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.3316735253008562e-05,
+      "loss": 2.0642,
+      "step": 9670
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.3201072578557206e-05,
+      "loss": 2.0848,
+      "step": 9675
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.3085897247201729e-05,
+      "loss": 2.0419,
+      "step": 9680
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.2971209559425124e-05,
+      "loss": 2.1245,
+      "step": 9685
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.2857009814438247e-05,
+      "loss": 2.0752,
+      "step": 9690
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.2743298310178842e-05,
+      "loss": 2.1045,
+      "step": 9695
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.2630075343310844e-05,
+      "loss": 2.1198,
+      "step": 9700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.2517341209223787e-05,
+      "loss": 2.1022,
+      "step": 9705
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.2405096202031718e-05,
+      "loss": 2.1245,
+      "step": 9710
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.2293340614572769e-05,
+      "loss": 2.0759,
+      "step": 9715
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.2182074738407955e-05,
+      "loss": 2.1398,
+      "step": 9720
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.2071298863820946e-05,
+      "loss": 2.1084,
+      "step": 9725
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.196101327981689e-05,
+      "loss": 2.0783,
+      "step": 9730
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.1851218274121856e-05,
+      "loss": 2.1567,
+      "step": 9735
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.1741914133181997e-05,
+      "loss": 2.1273,
+      "step": 9740
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.1633101142162828e-05,
+      "loss": 2.1442,
+      "step": 9745
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.1524779584948576e-05,
+      "loss": 2.1221,
+      "step": 9750
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.1416949744141358e-05,
+      "loss": 2.0778,
+      "step": 9755
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.1309611901060257e-05,
+      "loss": 2.1166,
+      "step": 9760
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.1202766335741022e-05,
+      "loss": 2.0887,
+      "step": 9765
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.1096413326934917e-05,
+      "loss": 2.0831,
+      "step": 9770
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.0990553152108263e-05,
+      "loss": 2.0461,
+      "step": 9775
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.0885186087441602e-05,
+      "loss": 2.1048,
+      "step": 9780
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.0780312407828885e-05,
+      "loss": 2.1238,
+      "step": 9785
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.0675932386877008e-05,
+      "loss": 2.0902,
+      "step": 9790
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.0572046296904959e-05,
+      "loss": 2.0718,
+      "step": 9795
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.0468654408942912e-05,
+      "loss": 2.1241,
+      "step": 9800
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.03657569927319e-05,
+      "loss": 2.1015,
+      "step": 9805
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.0263354316722807e-05,
+      "loss": 2.1432,
+      "step": 9810
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0161446648075857e-05,
+      "loss": 2.1003,
+      "step": 9815
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.0060034252659867e-05,
+      "loss": 2.0906,
+      "step": 9820
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.959117395051355e-06,
+      "loss": 2.128,
+      "step": 9825
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.85869633853418e-06,
+      "loss": 2.0762,
+      "step": 9830
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.758771345098661e-06,
+      "loss": 2.0651,
+      "step": 9835
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.659342675440909e-06,
+      "loss": 2.148,
+      "step": 9840
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.560410588962198e-06,
+      "loss": 2.1074,
+      "step": 9845
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.461975343768181e-06,
+      "loss": 2.1155,
+      "step": 9850
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.364037196668386e-06,
+      "loss": 2.0541,
+      "step": 9855
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.266596403175443e-06,
+      "loss": 2.0918,
+      "step": 9860
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.169653217504315e-06,
+      "loss": 2.0792,
+      "step": 9865
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.07320789257191e-06,
+      "loss": 2.1151,
+      "step": 9870
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.977260679996136e-06,
+      "loss": 2.1125,
+      "step": 9875
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.881811830095421e-06,
+      "loss": 2.0979,
+      "step": 9880
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.786861591888018e-06,
+      "loss": 2.1435,
+      "step": 9885
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.692410213091262e-06,
+      "loss": 2.1159,
+      "step": 9890
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.598457940121107e-06,
+      "loss": 2.0526,
+      "step": 9895
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.505005018091327e-06,
+      "loss": 2.1262,
+      "step": 9900
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.412051690812917e-06,
+      "loss": 2.1046,
+      "step": 9905
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.31959820079351e-06,
+      "loss": 2.0946,
+      "step": 9910
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.227644789236677e-06,
+      "loss": 2.0598,
+      "step": 9915
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.136191696041363e-06,
+      "loss": 2.1327,
+      "step": 9920
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21875,
+      "learning_rate": 8.045239159801177e-06,
+      "loss": 2.104,
+      "step": 9925
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.954787417803799e-06,
+      "loss": 2.0897,
+      "step": 9930
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.864836706030466e-06,
+      "loss": 2.0884,
+      "step": 9935
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.775387259155186e-06,
+      "loss": 2.1002,
+      "step": 9940
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.686439310544203e-06,
+      "loss": 2.0566,
+      "step": 9945
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.5979930922555285e-06,
+      "loss": 2.0904,
+      "step": 9950
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.510048835037986e-06,
+      "loss": 2.1242,
+      "step": 9955
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.4226067683310195e-06,
+      "loss": 2.0756,
+      "step": 9960
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.335667120263767e-06,
+      "loss": 2.1141,
+      "step": 9965
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.24923011765466e-06,
+      "loss": 2.1056,
+      "step": 9970
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.163295986010776e-06,
+      "loss": 2.0881,
+      "step": 9975
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.077864949527202e-06,
+      "loss": 2.0661,
+      "step": 9980
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.9929372310864936e-06,
+      "loss": 2.0908,
+      "step": 9985
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.90851305225817e-06,
+      "loss": 2.0908,
+      "step": 9990
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.8245926332979105e-06,
+      "loss": 2.1092,
+      "step": 9995
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.741176193147336e-06,
+      "loss": 2.108,
+      "step": 10000
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.658263949432963e-06,
+      "loss": 2.0959,
+      "step": 10005
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.5758561184661174e-06,
+      "loss": 2.0973,
+      "step": 10010
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.493952915242085e-06,
+      "loss": 2.0725,
+      "step": 10015
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.412554553439587e-06,
+      "loss": 2.1116,
+      "step": 10020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.331661245420306e-06,
+      "loss": 2.1142,
+      "step": 10025
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.251273202228203e-06,
+      "loss": 2.1287,
+      "step": 10030
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.171390633589113e-06,
+      "loss": 2.0923,
+      "step": 10035
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.0920137479101965e-06,
+      "loss": 2.0858,
+      "step": 10040
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.013142752279133e-06,
+      "loss": 2.0913,
+      "step": 10045
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.934777852463946e-06,
+      "loss": 2.1084,
+      "step": 10050
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.856919252912186e-06,
+      "loss": 2.1087,
+      "step": 10055
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.779567156750632e-06,
+      "loss": 2.1063,
+      "step": 10060
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.702721765784524e-06,
+      "loss": 2.1162,
+      "step": 10065
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.6263832804971825e-06,
+      "loss": 2.1194,
+      "step": 10070
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.550551900049494e-06,
+      "loss": 2.122,
+      "step": 10075
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.475227822279294e-06,
+      "loss": 2.1018,
+      "step": 10080
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.40041124370092e-06,
+      "loss": 2.1249,
+      "step": 10085
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.326102359504725e-06,
+      "loss": 2.1241,
+      "step": 10090
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.252301363556455e-06,
+      "loss": 2.1329,
+      "step": 10095
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.179008448396916e-06,
+      "loss": 2.0565,
+      "step": 10100
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.1062238052412835e-06,
+      "loss": 2.1075,
+      "step": 10105
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.0339476239787074e-06,
+      "loss": 2.1011,
+      "step": 10110
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.962180093171886e-06,
+      "loss": 2.1014,
+      "step": 10115
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.890921400056359e-06,
+      "loss": 2.0966,
+      "step": 10120
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.82017173054028e-06,
+      "loss": 2.1117,
+      "step": 10125
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.749931269203711e-06,
+      "loss": 2.072,
+      "step": 10130
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.6802001992982234e-06,
+      "loss": 2.0992,
+      "step": 10135
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.610978702746538e-06,
+      "loss": 2.0872,
+      "step": 10140
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.542266960141816e-06,
+      "loss": 2.122,
+      "step": 10145
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.474065150747375e-06,
+      "loss": 2.1117,
+      "step": 10150
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.406373452496148e-06,
+      "loss": 2.1459,
+      "step": 10155
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.339192041990226e-06,
+      "loss": 2.0726,
+      "step": 10160
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.272521094500426e-06,
+      "loss": 2.1388,
+      "step": 10165
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.206360783965768e-06,
+      "loss": 2.1223,
+      "step": 10170
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.140711282993071e-06,
+      "loss": 2.1004,
+      "step": 10175
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.075572762856528e-06,
+      "loss": 2.1162,
+      "step": 10180
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.0109453934972005e-06,
+      "loss": 2.0616,
+      "step": 10185
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.946829343522551e-06,
+      "loss": 2.0862,
+      "step": 10190
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.883224780206174e-06,
+      "loss": 2.0968,
+      "step": 10195
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.820131869487109e-06,
+      "loss": 2.1148,
+      "step": 10200
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.7575507759696427e-06,
+      "loss": 2.133,
+      "step": 10205
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.6954816629226618e-06,
+      "loss": 2.0871,
+      "step": 10210
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.6339246922794114e-06,
+      "loss": 2.16,
+      "step": 10215
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.5728800246370265e-06,
+      "loss": 2.1124,
+      "step": 10220
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.512347819256001e-06,
+      "loss": 2.0771,
+      "step": 10225
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.452328234059876e-06,
+      "loss": 2.1171,
+      "step": 10230
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.3928214256349067e-06,
+      "loss": 2.1057,
+      "step": 10235
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.3338275492293956e-06,
+      "loss": 2.1071,
+      "step": 10240
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.275346758753628e-06,
+      "loss": 2.0921,
+      "step": 10245
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.217379206779114e-06,
+      "loss": 2.1025,
+      "step": 10250
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.1599250445384587e-06,
+      "loss": 2.1286,
+      "step": 10255
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.102984421924915e-06,
+      "loss": 2.0973,
+      "step": 10260
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.0465574874918524e-06,
+      "loss": 2.1132,
+      "step": 10265
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.990644388452535e-06,
+      "loss": 2.1013,
+      "step": 10270
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.935245270679654e-06,
+      "loss": 2.0976,
+      "step": 10275
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.880360278704952e-06,
+      "loss": 2.1008,
+      "step": 10280
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.8259895557189332e-06,
+      "loss": 2.1045,
+      "step": 10285
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.772133243570285e-06,
+      "loss": 2.0994,
+      "step": 10290
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.7187914827657702e-06,
+      "loss": 2.0822,
+      "step": 10295
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.6659644124696235e-06,
+      "loss": 2.1504,
+      "step": 10300
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.6136521705034e-06,
+      "loss": 2.0762,
+      "step": 10305
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.5618548933453944e-06,
+      "loss": 2.1154,
+      "step": 10310
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.5105727161304883e-06,
+      "loss": 2.1195,
+      "step": 10315
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.4598057726496814e-06,
+      "loss": 2.0759,
+      "step": 10320
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.4095541953498057e-06,
+      "loss": 2.1001,
+      "step": 10325
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.3598181153330123e-06,
+      "loss": 2.1505,
+      "step": 10330
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.3105976623567505e-06,
+      "loss": 2.1319,
+      "step": 10335
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.261892964833101e-06,
+      "loss": 2.1112,
+      "step": 10340
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.2137041498286436e-06,
+      "loss": 2.1086,
+      "step": 10345
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.166031343064057e-06,
+      "loss": 2.0987,
+      "step": 10350
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.1188746689137395e-06,
+      "loss": 2.0885,
+      "step": 10355
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.072234250405658e-06,
+      "loss": 2.0984,
+      "step": 10360
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.026110209220833e-06,
+      "loss": 2.1191,
+      "step": 10365
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.9805026656930514e-06,
+      "loss": 2.1586,
+      "step": 10370
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.9354117388087125e-06,
+      "loss": 2.0988,
+      "step": 10375
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.8908375462063144e-06,
+      "loss": 2.0778,
+      "step": 10380
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.846780204176324e-06,
+      "loss": 2.1076,
+      "step": 10385
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.8032398276606854e-06,
+      "loss": 2.1024,
+      "step": 10390
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.7602165302526673e-06,
+      "loss": 2.1098,
+      "step": 10395
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.7177104241965946e-06,
+      "loss": 2.0772,
+      "step": 10400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.6757216203873382e-06,
+      "loss": 2.0987,
+      "step": 10405
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.6342502283702932e-06,
+      "loss": 2.0965,
+      "step": 10410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.5932963563409341e-06,
+      "loss": 2.0916,
+      "step": 10415
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.5528601111445051e-06,
+      "loss": 2.1307,
+      "step": 10420
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.5129415982758855e-06,
+      "loss": 2.1692,
+      "step": 10425
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.473540921879235e-06,
+      "loss": 2.1008,
+      "step": 10430
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.434658184747617e-06,
+      "loss": 2.1081,
+      "step": 10435
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.3962934883229751e-06,
+      "loss": 2.1117,
+      "step": 10440
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.3584469326956228e-06,
+      "loss": 2.1159,
+      "step": 10445
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.3211186166041557e-06,
+      "loss": 2.0878,
+      "step": 10450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.28430863743505e-06,
+      "loss": 2.1206,
+      "step": 10455
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.248017091222553e-06,
+      "loss": 2.1235,
+      "step": 10460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.2122440726483275e-06,
+      "loss": 2.128,
+      "step": 10465
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.1769896750412512e-06,
+      "loss": 2.1555,
+      "step": 10470
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.142253990377129e-06,
+      "loss": 2.1062,
+      "step": 10475
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.1080371092785591e-06,
+      "loss": 2.0768,
+      "step": 10480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.074339121014556e-06,
+      "loss": 2.1237,
+      "step": 10485
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.041160113500439e-06,
+      "loss": 2.1123,
+      "step": 10490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.0085001732974774e-06,
+      "loss": 2.115,
+      "step": 10495
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.763593856128017e-07,
+      "loss": 2.1598,
+      "step": 10500
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.447378342991142e-07,
+      "loss": 2.1414,
+      "step": 10505
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.136356018544012e-07,
+      "loss": 2.1251,
+      "step": 10510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.830527694218882e-07,
+      "loss": 2.0766,
+      "step": 10515
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.529894167896402e-07,
+      "loss": 2.0971,
+      "step": 10520
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.234456223904951e-07,
+      "loss": 2.0995,
+      "step": 10525
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.944214633018199e-07,
+      "loss": 2.1257,
+      "step": 10530
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.659170152452211e-07,
+      "loss": 2.1156,
+      "step": 10535
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.379323525865012e-07,
+      "loss": 2.1039,
+      "step": 10540
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.104675483353474e-07,
+      "loss": 2.1651,
+      "step": 10545
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.835226741451539e-07,
+      "loss": 2.0815,
+      "step": 10550
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.570978003129336e-07,
+      "loss": 2.0955,
+      "step": 10555
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.311929957789842e-07,
+      "loss": 2.1086,
+      "step": 10560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.058083281268223e-07,
+      "loss": 2.1192,
+      "step": 10565
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.809438635829833e-07,
+      "loss": 2.0822,
+      "step": 10570
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.565996670168216e-07,
+      "loss": 2.1074,
+      "step": 10575
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.327758019403106e-07,
+      "loss": 2.0805,
+      "step": 10580
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.094723305079984e-07,
+      "loss": 2.0862,
+      "step": 10585
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.866893135167416e-07,
+      "loss": 2.1187,
+      "step": 10590
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.6442681040552714e-07,
+      "loss": 2.0898,
+      "step": 10595
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.4268487925542835e-07,
+      "loss": 2.0712,
+      "step": 10600
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.2146357678936045e-07,
+      "loss": 2.0944,
+      "step": 10605
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.0076295837194745e-07,
+      "loss": 2.0915,
+      "step": 10610
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.805830780094333e-07,
+      "loss": 2.102,
+      "step": 10615
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.609239883494375e-07,
+      "loss": 2.163,
+      "step": 10620
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.417857406809333e-07,
+      "loss": 2.1239,
+      "step": 10625
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.2316838493404723e-07,
+      "loss": 2.1066,
+      "step": 10630
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.0507196967992644e-07,
+      "loss": 2.1214,
+      "step": 10635
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.8749654213062747e-07,
+      "loss": 2.0529,
+      "step": 10640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.7044214813902737e-07,
+      "loss": 2.1039,
+      "step": 10645
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.539088321986016e-07,
+      "loss": 2.1145,
+      "step": 10650
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.3789663744344659e-07,
+      "loss": 2.1126,
+      "step": 10655
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.2240560564807945e-07,
+      "loss": 2.1045,
+      "step": 10660
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.0743577722730501e-07,
+      "loss": 2.0901,
+      "step": 10665
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.9298719123617138e-07,
+      "loss": 2.1305,
+      "step": 10670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.7905988536988105e-07,
+      "loss": 2.0512,
+      "step": 10675
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.6565389596363555e-07,
+      "loss": 2.1194,
+      "step": 10680
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.527692579925466e-07,
+      "loss": 2.0744,
+      "step": 10685
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.4040600507156943e-07,
+      "loss": 2.052,
+      "step": 10690
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.2856416945545847e-07,
+      "loss": 2.1111,
+      "step": 10695
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.1724378203854524e-07,
+      "loss": 2.1183,
+      "step": 10700
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.0644487235482725e-07,
+      "loss": 2.1012,
+      "step": 10705
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.616746857774583e-08,
+      "loss": 2.1225,
+      "step": 10710
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.641159752023064e-08,
+      "loss": 2.1564,
+      "step": 10715
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.717728463454421e-08,
+      "loss": 2.0847,
+      "step": 10720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.846455401225971e-08,
+      "loss": 2.0842,
+      "step": 10725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.027342838414996e-08,
+      "loss": 2.1097,
+      "step": 10730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.260392912020962e-08,
+      "loss": 2.0956,
+      "step": 10735
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.545607622954417e-08,
+      "loss": 2.0895,
+      "step": 10740
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.882988836028112e-08,
+      "loss": 2.0985,
+      "step": 10745
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.272538279963655e-08,
+      "loss": 2.0959,
+      "step": 10750
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.714257547375976e-08,
+      "loss": 2.1242,
+      "step": 10755
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.2081480947711007e-08,
+      "loss": 2.0885,
+      "step": 10760
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.7542112425505964e-08,
+      "loss": 2.0554,
+      "step": 10765
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.3524481749982442e-08,
+      "loss": 2.1695,
+      "step": 10770
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.0028599402778226e-08,
+      "loss": 2.1069,
+      "step": 10775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.054474504375463e-09,
+      "loss": 2.1023,
+      "step": 10780
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.602114814056258e-09,
+      "loss": 2.0682,
+      "step": 10785
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.671526729769447e-09,
+      "loss": 2.0856,
+      "step": 10790
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.262715288308236e-09,
+      "loss": 2.1324,
+      "step": 10795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.7568416513256157e-10,
+      "loss": 2.0969,
+      "step": 10800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.0435674435704812e-11,
+      "loss": 2.1084,
+      "step": 10805
+    },
     {
       "epoch": 2.0,
-      "eval_loss": 2.14662766456604,
-      "eval_runtime": 188.6485,
-      "eval_samples_per_second": 25.778,
-      "eval_steps_per_second": 3.223,
-      "step": 2702
+      "eval_loss": 2.1439597606658936,
+      "eval_runtime": 190.2812,
+      "eval_samples_per_second": 25.557,
+      "eval_steps_per_second": 3.195,
+      "step": 10806
     },
     {
       "epoch": 2.0,
-      "step": 2702,
-      "total_flos": 6.137045069867254e+17,
-      "train_loss": 2.1601854102158,
-      "train_runtime": 21984.2259,
-      "train_samples_per_second": 7.864,
-      "train_steps_per_second": 0.123
+      "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
     }
   ],
   "logging_steps": 5,
-  "max_steps": 2702,
+  "max_steps": 10806,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 100,
-  "total_flos": 6.137045069867254e+17,
-  "train_batch_size": 32,
+  "total_flos": 6.137045068818678e+17,
+  "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null
 }