diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,3352 +1,13015 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 6.0,
-  "global_step": 2586,
+  "epoch": 2.0,
+  "global_step": 10752,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 2.2895645784996133e-05,
-      "loss": 3.3205,
+      "epoch": 0.0,
+      "learning_rate": 1.1472385386519655e-07,
+      "loss": 4.2326,
       "step": 5
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 2.2851291569992266e-05,
-      "loss": 2.9999,
+      "epoch": 0.0,
+      "learning_rate": 2.294477077303931e-07,
+      "loss": 4.5232,
       "step": 10
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 2.28069373549884e-05,
-      "loss": 2.9241,
+      "epoch": 0.0,
+      "learning_rate": 3.441715615955896e-07,
+      "loss": 4.2695,
       "step": 15
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 2.2762583139984532e-05,
-      "loss": 2.7706,
+      "epoch": 0.0,
+      "learning_rate": 4.588954154607862e-07,
+      "loss": 4.1255,
       "step": 20
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 2.2718228924980665e-05,
-      "loss": 2.8128,
+      "epoch": 0.0,
+      "learning_rate": 5.736192693259828e-07,
+      "loss": 4.4481,
       "step": 25
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 2.2673874709976798e-05,
-      "loss": 2.7359,
+      "epoch": 0.01,
+      "learning_rate": 6.883431231911792e-07,
+      "loss": 4.0765,
       "step": 30
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 2.262952049497293e-05,
-      "loss": 2.7496,
+      "epoch": 0.01,
+      "learning_rate": 8.030669770563759e-07,
+      "loss": 4.0945,
       "step": 35
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 2.2585166279969064e-05,
-      "loss": 2.6543,
+      "epoch": 0.01,
+      "learning_rate": 9.177908309215724e-07,
+      "loss": 3.9355,
       "step": 40
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 2.2540812064965197e-05,
-      "loss": 2.6493,
+      "epoch": 0.01,
+      "learning_rate": 1.0325146847867689e-06,
+      "loss": 3.7374,
       "step": 45
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 2.2496457849961333e-05,
-      "loss": 2.6312,
+      "epoch": 0.01,
+      "learning_rate": 1.1472385386519655e-06,
+      "loss": 3.6504,
       "step": 50
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 2.2452103634957466e-05,
-      "loss": 2.6222,
+      "epoch": 0.01,
+      "learning_rate": 1.261962392517162e-06,
+      "loss": 3.9264,
       "step": 55
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 2.2407749419953595e-05,
-      "loss": 2.6328,
+      "epoch": 0.01,
+      "learning_rate": 1.3766862463823585e-06,
+      "loss": 3.9033,
       "step": 60
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 2.236339520494973e-05,
-      "loss": 2.6085,
+      "epoch": 0.01,
+      "learning_rate": 1.4914101002475552e-06,
+      "loss": 3.5127,
       "step": 65
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 2.2319040989945864e-05,
-      "loss": 2.5869,
+      "epoch": 0.01,
+      "learning_rate": 1.6061339541127518e-06,
+      "loss": 3.2964,
       "step": 70
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 2.2274686774941994e-05,
-      "loss": 2.6353,
+      "epoch": 0.01,
+      "learning_rate": 1.720857807977948e-06,
+      "loss": 3.4927,
       "step": 75
     },
     {
-      "epoch": 0.19,
-      "learning_rate": 2.223033255993813e-05,
-      "loss": 2.6266,
+      "epoch": 0.01,
+      "learning_rate": 1.8355816618431448e-06,
+      "loss": 3.1635,
       "step": 80
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 2.2185978344934263e-05,
-      "loss": 2.6341,
+      "epoch": 0.02,
+      "learning_rate": 1.9503055157083414e-06,
+      "loss": 3.4046,
       "step": 85
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 2.2141624129930393e-05,
-      "loss": 2.601,
+      "epoch": 0.02,
+      "learning_rate": 2.0650293695735377e-06,
+      "loss": 3.2964,
       "step": 90
     },
     {
-      "epoch": 0.22,
-      "learning_rate": 2.209726991492653e-05,
-      "loss": 2.5009,
+      "epoch": 0.02,
+      "learning_rate": 2.1797532234387344e-06,
+      "loss": 3.265,
       "step": 95
     },
     {
-      "epoch": 0.23,
-      "learning_rate": 2.205291569992266e-05,
-      "loss": 2.5507,
-      "step": 100
-    },
-    {
-      "epoch": 0.23,
-      "eval_accuracy": 0.5567914705949215,
-      "eval_loss": 2.5061137676239014,
-      "eval_runtime": 58.2489,
-      "eval_samples_per_second": 7.382,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.02,
+      "learning_rate": 2.294477077303931e-06,
+      "loss": 2.852,
       "step": 100
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 2.2008561484918795e-05,
-      "loss": 2.5848,
+      "epoch": 0.02,
+      "learning_rate": 2.4092009311691273e-06,
+      "loss": 3.2298,
       "step": 105
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 2.1964207269914927e-05,
-      "loss": 2.5978,
+      "epoch": 0.02,
+      "learning_rate": 2.523924785034324e-06,
+      "loss": 3.3424,
       "step": 110
     },
     {
-      "epoch": 0.27,
-      "learning_rate": 2.191985305491106e-05,
-      "loss": 2.4714,
+      "epoch": 0.02,
+      "learning_rate": 2.6386486388995207e-06,
+      "loss": 3.2604,
       "step": 115
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 2.1875498839907193e-05,
-      "loss": 2.5151,
+      "epoch": 0.02,
+      "learning_rate": 2.753372492764717e-06,
+      "loss": 3.3028,
       "step": 120
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 2.1831144624903326e-05,
-      "loss": 2.5547,
+      "epoch": 0.02,
+      "learning_rate": 2.8680963466299136e-06,
+      "loss": 2.93,
       "step": 125
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 2.178679040989946e-05,
-      "loss": 2.5302,
+      "epoch": 0.02,
+      "learning_rate": 2.9828202004951103e-06,
+      "loss": 3.0727,
       "step": 130
     },
     {
-      "epoch": 0.31,
-      "learning_rate": 2.1742436194895592e-05,
-      "loss": 2.5061,
+      "epoch": 0.03,
+      "learning_rate": 3.097544054360307e-06,
+      "loss": 3.0509,
       "step": 135
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 2.1698081979891725e-05,
-      "loss": 2.4813,
+      "epoch": 0.03,
+      "learning_rate": 3.2122679082255037e-06,
+      "loss": 2.8428,
       "step": 140
     },
     {
-      "epoch": 0.34,
-      "learning_rate": 2.1653727764887858e-05,
-      "loss": 2.44,
+      "epoch": 0.03,
+      "learning_rate": 3.3269917620906995e-06,
+      "loss": 2.9271,
       "step": 145
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 2.1609373549883994e-05,
-      "loss": 2.4794,
+      "epoch": 0.03,
+      "learning_rate": 3.441715615955896e-06,
+      "loss": 2.9629,
       "step": 150
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 2.1565019334880123e-05,
-      "loss": 2.5501,
+      "epoch": 0.03,
+      "learning_rate": 3.556439469821093e-06,
+      "loss": 2.7586,
       "step": 155
     },
     {
-      "epoch": 0.37,
-      "learning_rate": 2.1520665119876256e-05,
-      "loss": 2.4405,
+      "epoch": 0.03,
+      "learning_rate": 3.6711633236862895e-06,
+      "loss": 2.8137,
       "step": 160
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 2.1476310904872393e-05,
-      "loss": 2.4833,
+      "epoch": 0.03,
+      "learning_rate": 3.7858871775514862e-06,
+      "loss": 3.1196,
       "step": 165
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 2.1431956689868522e-05,
-      "loss": 2.5244,
+      "epoch": 0.03,
+      "learning_rate": 3.900611031416683e-06,
+      "loss": 2.6954,
       "step": 170
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 2.1387602474864655e-05,
-      "loss": 2.5057,
+      "epoch": 0.03,
+      "learning_rate": 4.015334885281879e-06,
+      "loss": 2.913,
       "step": 175
     },
     {
-      "epoch": 0.42,
-      "learning_rate": 2.134324825986079e-05,
-      "loss": 2.4521,
+      "epoch": 0.03,
+      "learning_rate": 4.130058739147075e-06,
+      "loss": 3.045,
       "step": 180
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 2.129889404485692e-05,
-      "loss": 2.4377,
+      "epoch": 0.03,
+      "learning_rate": 4.244782593012272e-06,
+      "loss": 3.1913,
       "step": 185
     },
     {
-      "epoch": 0.44,
-      "learning_rate": 2.1254539829853054e-05,
-      "loss": 2.4831,
+      "epoch": 0.04,
+      "learning_rate": 4.359506446877469e-06,
+      "loss": 2.9136,
       "step": 190
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 2.121018561484919e-05,
-      "loss": 2.4314,
+      "epoch": 0.04,
+      "learning_rate": 4.4742303007426655e-06,
+      "loss": 2.6609,
       "step": 195
     },
     {
-      "epoch": 0.46,
-      "learning_rate": 2.1165831399845323e-05,
-      "loss": 2.465,
-      "step": 200
-    },
-    {
-      "epoch": 0.46,
-      "eval_accuracy": 0.5671849780626975,
-      "eval_loss": 2.42543625831604,
-      "eval_runtime": 58.1969,
-      "eval_samples_per_second": 7.389,
-      "eval_steps_per_second": 1.856,
+      "epoch": 0.04,
+      "learning_rate": 4.588954154607862e-06,
+      "loss": 2.6252,
       "step": 200
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 2.1121477184841452e-05,
-      "loss": 2.475,
+      "epoch": 0.04,
+      "learning_rate": 4.703678008473058e-06,
+      "loss": 3.0444,
       "step": 205
     },
     {
-      "epoch": 0.49,
-      "learning_rate": 2.107712296983759e-05,
-      "loss": 2.4854,
+      "epoch": 0.04,
+      "learning_rate": 4.818401862338255e-06,
+      "loss": 2.7076,
       "step": 210
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 2.103276875483372e-05,
-      "loss": 2.4716,
+      "epoch": 0.04,
+      "learning_rate": 4.933125716203451e-06,
+      "loss": 2.7443,
       "step": 215
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 2.0988414539829854e-05,
-      "loss": 2.4928,
+      "epoch": 0.04,
+      "learning_rate": 5.047849570068648e-06,
+      "loss": 2.6593,
       "step": 220
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 2.0944060324825987e-05,
-      "loss": 2.4601,
+      "epoch": 0.04,
+      "learning_rate": 5.162573423933845e-06,
+      "loss": 3.2937,
       "step": 225
     },
     {
-      "epoch": 0.53,
-      "learning_rate": 2.089970610982212e-05,
-      "loss": 2.4256,
+      "epoch": 0.04,
+      "learning_rate": 5.277297277799041e-06,
+      "loss": 2.8728,
       "step": 230
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 2.0855351894818253e-05,
-      "loss": 2.4221,
+      "epoch": 0.04,
+      "learning_rate": 5.392021131664237e-06,
+      "loss": 2.8222,
       "step": 235
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 2.0810997679814386e-05,
-      "loss": 2.43,
+      "epoch": 0.04,
+      "learning_rate": 5.506744985529434e-06,
+      "loss": 2.781,
       "step": 240
     },
     {
-      "epoch": 0.57,
-      "learning_rate": 2.076664346481052e-05,
-      "loss": 2.4551,
+      "epoch": 0.05,
+      "learning_rate": 5.6214688393946306e-06,
+      "loss": 2.5678,
       "step": 245
     },
     {
-      "epoch": 0.58,
-      "learning_rate": 2.072228924980665e-05,
-      "loss": 2.4228,
+      "epoch": 0.05,
+      "learning_rate": 5.736192693259827e-06,
+      "loss": 2.5339,
       "step": 250
     },
     {
-      "epoch": 0.59,
-      "learning_rate": 2.0677935034802784e-05,
-      "loss": 2.4224,
+      "epoch": 0.05,
+      "learning_rate": 5.850916547125024e-06,
+      "loss": 2.6287,
       "step": 255
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 2.0633580819798917e-05,
-      "loss": 2.5017,
+      "epoch": 0.05,
+      "learning_rate": 5.965640400990221e-06,
+      "loss": 2.9172,
       "step": 260
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 2.058922660479505e-05,
-      "loss": 2.4203,
+      "epoch": 0.05,
+      "learning_rate": 6.080364254855417e-06,
+      "loss": 2.6293,
       "step": 265
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 2.0544872389791183e-05,
-      "loss": 2.4648,
+      "epoch": 0.05,
+      "learning_rate": 6.195088108720614e-06,
+      "loss": 2.8912,
       "step": 270
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 2.0500518174787316e-05,
-      "loss": 2.4495,
+      "epoch": 0.05,
+      "learning_rate": 6.309811962585811e-06,
+      "loss": 3.2172,
       "step": 275
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 2.045616395978345e-05,
-      "loss": 2.4038,
+      "epoch": 0.05,
+      "learning_rate": 6.424535816451007e-06,
+      "loss": 2.5888,
       "step": 280
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 2.0411809744779582e-05,
-      "loss": 2.4055,
+      "epoch": 0.05,
+      "learning_rate": 6.539259670316202e-06,
+      "loss": 2.724,
       "step": 285
     },
     {
-      "epoch": 0.67,
-      "learning_rate": 2.0367455529775715e-05,
-      "loss": 2.4111,
+      "epoch": 0.05,
+      "learning_rate": 6.653983524181399e-06,
+      "loss": 2.5867,
       "step": 290
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 2.032310131477185e-05,
-      "loss": 2.4462,
+      "epoch": 0.05,
+      "learning_rate": 6.768707378046596e-06,
+      "loss": 2.6797,
       "step": 295
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 2.027874709976798e-05,
-      "loss": 2.3919,
-      "step": 300
-    },
-    {
-      "epoch": 0.7,
-      "eval_accuracy": 0.5726067880606516,
-      "eval_loss": 2.382695436477661,
-      "eval_runtime": 58.2041,
-      "eval_samples_per_second": 7.388,
-      "eval_steps_per_second": 1.856,
+      "epoch": 0.06,
+      "learning_rate": 6.883431231911792e-06,
+      "loss": 3.0558,
       "step": 300
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 2.0234392884764117e-05,
-      "loss": 2.3573,
+      "epoch": 0.06,
+      "learning_rate": 6.998155085776989e-06,
+      "loss": 2.9952,
       "step": 305
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 2.019003866976025e-05,
-      "loss": 2.4338,
+      "epoch": 0.06,
+      "learning_rate": 7.112878939642186e-06,
+      "loss": 2.7067,
       "step": 310
     },
     {
-      "epoch": 0.73,
-      "learning_rate": 2.014568445475638e-05,
-      "loss": 2.398,
+      "epoch": 0.06,
+      "learning_rate": 7.227602793507382e-06,
+      "loss": 2.7179,
       "step": 315
     },
     {
-      "epoch": 0.74,
-      "learning_rate": 2.0101330239752515e-05,
-      "loss": 2.3701,
+      "epoch": 0.06,
+      "learning_rate": 7.342326647372579e-06,
+      "loss": 2.7779,
       "step": 320
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 2.0056976024748648e-05,
-      "loss": 2.4837,
+      "epoch": 0.06,
+      "learning_rate": 7.457050501237776e-06,
+      "loss": 2.761,
       "step": 325
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 2.0012621809744778e-05,
-      "loss": 2.3821,
+      "epoch": 0.06,
+      "learning_rate": 7.5717743551029724e-06,
+      "loss": 2.7821,
       "step": 330
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 1.9968267594740914e-05,
-      "loss": 2.408,
+      "epoch": 0.06,
+      "learning_rate": 7.686498208968169e-06,
+      "loss": 2.5932,
       "step": 335
     },
     {
-      "epoch": 0.79,
-      "learning_rate": 1.9923913379737047e-05,
-      "loss": 2.386,
+      "epoch": 0.06,
+      "learning_rate": 7.801222062833366e-06,
+      "loss": 2.6604,
       "step": 340
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 1.987955916473318e-05,
-      "loss": 2.4007,
+      "epoch": 0.06,
+      "learning_rate": 7.91594591669856e-06,
+      "loss": 2.6451,
       "step": 345
     },
     {
-      "epoch": 0.81,
-      "learning_rate": 1.9835204949729313e-05,
-      "loss": 2.44,
+      "epoch": 0.07,
+      "learning_rate": 8.030669770563757e-06,
+      "loss": 2.55,
       "step": 350
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 1.9790850734725446e-05,
-      "loss": 2.3848,
+      "epoch": 0.07,
+      "learning_rate": 8.145393624428954e-06,
+      "loss": 2.754,
       "step": 355
     },
     {
-      "epoch": 0.84,
-      "learning_rate": 1.974649651972158e-05,
-      "loss": 2.3972,
+      "epoch": 0.07,
+      "learning_rate": 8.26011747829415e-06,
+      "loss": 2.7822,
       "step": 360
     },
     {
-      "epoch": 0.85,
-      "learning_rate": 1.970214230471771e-05,
-      "loss": 2.3759,
+      "epoch": 0.07,
+      "learning_rate": 8.374841332159348e-06,
+      "loss": 2.6191,
       "step": 365
     },
     {
-      "epoch": 0.86,
-      "learning_rate": 1.9657788089713844e-05,
-      "loss": 2.3842,
+      "epoch": 0.07,
+      "learning_rate": 8.489565186024544e-06,
+      "loss": 2.7476,
       "step": 370
     },
     {
-      "epoch": 0.87,
-      "learning_rate": 1.9613433874709977e-05,
-      "loss": 2.3788,
+      "epoch": 0.07,
+      "learning_rate": 8.604289039889741e-06,
+      "loss": 3.0726,
       "step": 375
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 1.956907965970611e-05,
-      "loss": 2.4706,
+      "epoch": 0.07,
+      "learning_rate": 8.719012893754938e-06,
+      "loss": 2.621,
       "step": 380
     },
     {
-      "epoch": 0.89,
-      "learning_rate": 1.9524725444702243e-05,
-      "loss": 2.3689,
+      "epoch": 0.07,
+      "learning_rate": 8.833736747620134e-06,
+      "loss": 2.8915,
       "step": 385
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 1.948037122969838e-05,
-      "loss": 2.3853,
+      "epoch": 0.07,
+      "learning_rate": 8.948460601485331e-06,
+      "loss": 2.5251,
       "step": 390
     },
     {
-      "epoch": 0.92,
-      "learning_rate": 1.943601701469451e-05,
-      "loss": 2.4519,
+      "epoch": 0.07,
+      "learning_rate": 9.063184455350528e-06,
+      "loss": 2.7159,
       "step": 395
     },
     {
-      "epoch": 0.93,
-      "learning_rate": 1.939166279969064e-05,
-      "loss": 2.4222,
-      "step": 400
-    },
-    {
-      "epoch": 0.93,
-      "eval_accuracy": 0.5760144581599945,
-      "eval_loss": 2.3488640785217285,
-      "eval_runtime": 58.2409,
-      "eval_samples_per_second": 7.383,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.07,
+      "learning_rate": 9.177908309215724e-06,
+      "loss": 2.7079,
       "step": 400
     },
     {
-      "epoch": 0.94,
-      "learning_rate": 1.9347308584686778e-05,
-      "loss": 2.3769,
+      "epoch": 0.08,
+      "learning_rate": 9.292632163080921e-06,
+      "loss": 2.8255,
       "step": 405
     },
     {
-      "epoch": 0.95,
-      "learning_rate": 1.9302954369682907e-05,
-      "loss": 2.3834,
+      "epoch": 0.08,
+      "learning_rate": 9.407356016946116e-06,
+      "loss": 2.6322,
       "step": 410
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 1.925860015467904e-05,
-      "loss": 2.4063,
+      "epoch": 0.08,
+      "learning_rate": 9.522079870811313e-06,
+      "loss": 2.8833,
       "step": 415
     },
     {
-      "epoch": 0.97,
-      "learning_rate": 1.9214245939675176e-05,
-      "loss": 2.3603,
+      "epoch": 0.08,
+      "learning_rate": 9.63680372467651e-06,
+      "loss": 2.7325,
       "step": 420
     },
     {
-      "epoch": 0.99,
-      "learning_rate": 1.9169891724671306e-05,
-      "loss": 2.3989,
+      "epoch": 0.08,
+      "learning_rate": 9.751527578541706e-06,
+      "loss": 2.9166,
       "step": 425
     },
     {
-      "epoch": 1.0,
-      "learning_rate": 1.912553750966744e-05,
-      "loss": 2.3888,
+      "epoch": 0.08,
+      "learning_rate": 9.866251432406903e-06,
+      "loss": 2.8784,
       "step": 430
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 1.9081183294663575e-05,
-      "loss": 2.2754,
+      "epoch": 0.08,
+      "learning_rate": 9.9809752862721e-06,
+      "loss": 2.7534,
       "step": 435
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 1.9036829079659708e-05,
-      "loss": 2.2926,
+      "epoch": 0.08,
+      "learning_rate": 1.0095699140137296e-05,
+      "loss": 2.6256,
       "step": 440
     },
     {
-      "epoch": 1.03,
-      "learning_rate": 1.8992474864655837e-05,
-      "loss": 2.2617,
+      "epoch": 0.08,
+      "learning_rate": 1.0210422994002493e-05,
+      "loss": 2.5401,
       "step": 445
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 1.8948120649651974e-05,
-      "loss": 2.2712,
+      "epoch": 0.08,
+      "learning_rate": 1.032514684786769e-05,
+      "loss": 2.6513,
       "step": 450
     },
     {
-      "epoch": 1.06,
-      "learning_rate": 1.8903766434648107e-05,
-      "loss": 2.2109,
+      "epoch": 0.08,
+      "learning_rate": 1.0439870701732886e-05,
+      "loss": 2.7761,
       "step": 455
     },
     {
-      "epoch": 1.07,
-      "learning_rate": 1.885941221964424e-05,
-      "loss": 2.258,
+      "epoch": 0.09,
+      "learning_rate": 1.0554594555598083e-05,
+      "loss": 2.5604,
       "step": 460
     },
     {
-      "epoch": 1.08,
-      "learning_rate": 1.8815058004640372e-05,
-      "loss": 2.2625,
+      "epoch": 0.09,
+      "learning_rate": 1.066931840946328e-05,
+      "loss": 2.8281,
       "step": 465
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 1.8770703789636505e-05,
-      "loss": 2.3066,
+      "epoch": 0.09,
+      "learning_rate": 1.0784042263328474e-05,
+      "loss": 2.6211,
       "step": 470
     },
     {
-      "epoch": 1.1,
-      "learning_rate": 1.8726349574632638e-05,
-      "loss": 2.2384,
+      "epoch": 0.09,
+      "learning_rate": 1.0898766117193671e-05,
+      "loss": 2.6652,
       "step": 475
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 1.868199535962877e-05,
-      "loss": 2.2411,
+      "epoch": 0.09,
+      "learning_rate": 1.1013489971058868e-05,
+      "loss": 2.4588,
       "step": 480
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 1.8637641144624904e-05,
-      "loss": 2.2545,
+      "epoch": 0.09,
+      "learning_rate": 1.1128213824924064e-05,
+      "loss": 2.9148,
       "step": 485
     },
     {
-      "epoch": 1.14,
-      "learning_rate": 1.8593286929621037e-05,
-      "loss": 2.2706,
+      "epoch": 0.09,
+      "learning_rate": 1.1242937678789261e-05,
+      "loss": 2.6054,
       "step": 490
     },
     {
-      "epoch": 1.15,
-      "learning_rate": 1.854893271461717e-05,
-      "loss": 2.2407,
+      "epoch": 0.09,
+      "learning_rate": 1.1357661532654458e-05,
+      "loss": 2.7581,
       "step": 495
     },
     {
-      "epoch": 1.16,
-      "learning_rate": 1.8504578499613303e-05,
-      "loss": 2.1958,
-      "step": 500
-    },
-    {
-      "epoch": 1.16,
-      "eval_accuracy": 0.5792607242719771,
-      "eval_loss": 2.3301777839660645,
-      "eval_runtime": 58.2054,
-      "eval_samples_per_second": 7.388,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.09,
+      "learning_rate": 1.1472385386519654e-05,
+      "loss": 2.8565,
       "step": 500
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 1.8460224284609435e-05,
-      "loss": 2.2448,
+      "epoch": 0.09,
+      "learning_rate": 1.1587109240384851e-05,
+      "loss": 2.7272,
       "step": 505
     },
     {
-      "epoch": 1.18,
-      "learning_rate": 1.841587006960557e-05,
-      "loss": 2.3031,
+      "epoch": 0.09,
+      "learning_rate": 1.1701833094250048e-05,
+      "loss": 2.6562,
       "step": 510
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 1.83715158546017e-05,
-      "loss": 2.196,
+      "epoch": 0.1,
+      "learning_rate": 1.1816556948115245e-05,
+      "loss": 2.5933,
       "step": 515
     },
     {
-      "epoch": 1.21,
-      "learning_rate": 1.8327161639597834e-05,
-      "loss": 2.3195,
+      "epoch": 0.1,
+      "learning_rate": 1.1931280801980441e-05,
+      "loss": 2.6228,
       "step": 520
     },
     {
-      "epoch": 1.22,
-      "learning_rate": 1.8282807424593967e-05,
-      "loss": 2.2407,
+      "epoch": 0.1,
+      "learning_rate": 1.2046004655845638e-05,
+      "loss": 2.6905,
       "step": 525
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 1.82384532095901e-05,
-      "loss": 2.2314,
+      "epoch": 0.1,
+      "learning_rate": 1.2160728509710835e-05,
+      "loss": 2.1833,
       "step": 530
     },
     {
-      "epoch": 1.24,
-      "learning_rate": 1.8194098994586236e-05,
-      "loss": 2.2374,
+      "epoch": 0.1,
+      "learning_rate": 1.2275452363576031e-05,
+      "loss": 2.6952,
       "step": 535
     },
     {
-      "epoch": 1.25,
-      "learning_rate": 1.8149744779582366e-05,
-      "loss": 2.2567,
+      "epoch": 0.1,
+      "learning_rate": 1.2390176217441228e-05,
+      "loss": 2.537,
       "step": 540
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 1.8105390564578502e-05,
-      "loss": 2.2234,
+      "epoch": 0.1,
+      "learning_rate": 1.2504900071306425e-05,
+      "loss": 2.7027,
       "step": 545
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 1.8061036349574635e-05,
-      "loss": 2.2688,
+      "epoch": 0.1,
+      "learning_rate": 1.2619623925171621e-05,
+      "loss": 2.6045,
       "step": 550
     },
     {
-      "epoch": 1.29,
-      "learning_rate": 1.8016682134570764e-05,
-      "loss": 2.2644,
+      "epoch": 0.1,
+      "learning_rate": 1.2734347779036818e-05,
+      "loss": 2.6345,
       "step": 555
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 1.79723279195669e-05,
-      "loss": 2.255,
+      "epoch": 0.1,
+      "learning_rate": 1.2849071632902015e-05,
+      "loss": 2.8346,
       "step": 560
     },
     {
-      "epoch": 1.31,
-      "learning_rate": 1.7927973704563033e-05,
-      "loss": 2.2402,
+      "epoch": 0.11,
+      "learning_rate": 1.2963795486767208e-05,
+      "loss": 2.4451,
       "step": 565
     },
     {
-      "epoch": 1.32,
-      "learning_rate": 1.7883619489559163e-05,
-      "loss": 2.1857,
+      "epoch": 0.11,
+      "learning_rate": 1.3078519340632405e-05,
+      "loss": 2.1453,
       "step": 570
     },
     {
-      "epoch": 1.33,
-      "learning_rate": 1.78392652745553e-05,
-      "loss": 2.278,
+      "epoch": 0.11,
+      "learning_rate": 1.3193243194497601e-05,
+      "loss": 2.1882,
       "step": 575
     },
     {
-      "epoch": 1.35,
-      "learning_rate": 1.7794911059551432e-05,
-      "loss": 2.2623,
+      "epoch": 0.11,
+      "learning_rate": 1.3307967048362798e-05,
+      "loss": 2.8247,
       "step": 580
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 1.7750556844547565e-05,
-      "loss": 2.2192,
+      "epoch": 0.11,
+      "learning_rate": 1.3422690902227995e-05,
+      "loss": 2.6848,
       "step": 585
     },
     {
-      "epoch": 1.37,
-      "learning_rate": 1.7706202629543698e-05,
-      "loss": 2.2716,
+      "epoch": 0.11,
+      "learning_rate": 1.3537414756093191e-05,
+      "loss": 2.4843,
       "step": 590
     },
     {
-      "epoch": 1.38,
-      "learning_rate": 1.766184841453983e-05,
-      "loss": 2.2689,
+      "epoch": 0.11,
+      "learning_rate": 1.3652138609958388e-05,
+      "loss": 2.6365,
       "step": 595
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 1.7617494199535964e-05,
-      "loss": 2.2087,
-      "step": 600
-    },
-    {
-      "epoch": 1.39,
-      "eval_accuracy": 0.5817590761326695,
-      "eval_loss": 2.3123257160186768,
-      "eval_runtime": 58.2597,
-      "eval_samples_per_second": 7.381,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.11,
+      "learning_rate": 1.3766862463823585e-05,
+      "loss": 2.4,
       "step": 600
     },
     {
-      "epoch": 1.4,
-      "learning_rate": 1.7573139984532096e-05,
-      "loss": 2.2795,
+      "epoch": 0.11,
+      "learning_rate": 1.3881586317688781e-05,
+      "loss": 2.5909,
       "step": 605
     },
     {
-      "epoch": 1.42,
-      "learning_rate": 1.752878576952823e-05,
-      "loss": 2.2978,
+      "epoch": 0.11,
+      "learning_rate": 1.3996310171553978e-05,
+      "loss": 2.8485,
       "step": 610
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 1.7484431554524362e-05,
-      "loss": 2.202,
+      "epoch": 0.11,
+      "learning_rate": 1.4111034025419175e-05,
+      "loss": 2.5356,
       "step": 615
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 1.7440077339520495e-05,
-      "loss": 2.2099,
+      "epoch": 0.12,
+      "learning_rate": 1.4225757879284371e-05,
+      "loss": 2.5269,
       "step": 620
     },
     {
-      "epoch": 1.45,
-      "learning_rate": 1.7395723124516628e-05,
-      "loss": 2.2414,
+      "epoch": 0.12,
+      "learning_rate": 1.4340481733149568e-05,
+      "loss": 2.7219,
       "step": 625
     },
     {
-      "epoch": 1.46,
-      "learning_rate": 1.7351368909512764e-05,
-      "loss": 2.2039,
+      "epoch": 0.12,
+      "learning_rate": 1.4455205587014765e-05,
+      "loss": 2.73,
       "step": 630
     },
     {
-      "epoch": 1.47,
-      "learning_rate": 1.7307014694508894e-05,
-      "loss": 2.2568,
+      "epoch": 0.12,
+      "learning_rate": 1.4569929440879961e-05,
+      "loss": 2.561,
       "step": 635
     },
     {
-      "epoch": 1.48,
-      "learning_rate": 1.7262660479505027e-05,
-      "loss": 2.2013,
+      "epoch": 0.12,
+      "learning_rate": 1.4684653294745158e-05,
+      "loss": 2.425,
       "step": 640
     },
     {
-      "epoch": 1.5,
-      "learning_rate": 1.7218306264501163e-05,
-      "loss": 2.2123,
+      "epoch": 0.12,
+      "learning_rate": 1.4799377148610355e-05,
+      "loss": 2.3738,
       "step": 645
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 1.7173952049497292e-05,
-      "loss": 2.2459,
+      "epoch": 0.12,
+      "learning_rate": 1.4914101002475552e-05,
+      "loss": 2.7559,
       "step": 650
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 1.7129597834493425e-05,
-      "loss": 2.3154,
+      "epoch": 0.12,
+      "learning_rate": 1.5028824856340748e-05,
+      "loss": 2.7757,
       "step": 655
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 1.708524361948956e-05,
-      "loss": 2.1588,
+      "epoch": 0.12,
+      "learning_rate": 1.5143548710205945e-05,
+      "loss": 2.6654,
       "step": 660
     },
     {
-      "epoch": 1.54,
-      "learning_rate": 1.704088940448569e-05,
-      "loss": 2.2633,
+      "epoch": 0.12,
+      "learning_rate": 1.5258272564071142e-05,
+      "loss": 2.766,
       "step": 665
     },
     {
-      "epoch": 1.55,
-      "learning_rate": 1.6996535189481824e-05,
-      "loss": 2.2424,
+      "epoch": 0.12,
+      "learning_rate": 1.5372996417936338e-05,
+      "loss": 2.2905,
       "step": 670
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 1.695218097447796e-05,
-      "loss": 2.2245,
+      "epoch": 0.13,
+      "learning_rate": 1.5487720271801535e-05,
+      "loss": 2.8645,
       "step": 675
     },
     {
-      "epoch": 1.58,
-      "learning_rate": 1.6907826759474093e-05,
-      "loss": 2.2103,
+      "epoch": 0.13,
+      "learning_rate": 1.560244412566673e-05,
+      "loss": 2.2969,
       "step": 680
     },
     {
-      "epoch": 1.59,
-      "learning_rate": 1.6863472544470223e-05,
-      "loss": 2.1734,
+      "epoch": 0.13,
+      "learning_rate": 1.5717167979531928e-05,
+      "loss": 2.6437,
       "step": 685
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 1.681911832946636e-05,
-      "loss": 2.1947,
+      "epoch": 0.13,
+      "learning_rate": 1.583189183339712e-05,
+      "loss": 2.4631,
       "step": 690
     },
     {
-      "epoch": 1.61,
-      "learning_rate": 1.6774764114462492e-05,
-      "loss": 2.2474,
+      "epoch": 0.13,
+      "learning_rate": 1.5946615687262318e-05,
+      "loss": 2.6186,
       "step": 695
     },
     {
-      "epoch": 1.62,
-      "learning_rate": 1.673040989945862e-05,
-      "loss": 2.2436,
-      "step": 700
-    },
-    {
-      "epoch": 1.62,
-      "eval_accuracy": 0.5841005705971948,
-      "eval_loss": 2.2960166931152344,
-      "eval_runtime": 58.2447,
-      "eval_samples_per_second": 7.383,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.13,
+      "learning_rate": 1.6061339541127515e-05,
+      "loss": 2.5039,
       "step": 700
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 1.6686055684454758e-05,
-      "loss": 2.2242,
+      "epoch": 0.13,
+      "learning_rate": 1.617606339499271e-05,
+      "loss": 2.5645,
       "step": 705
     },
     {
-      "epoch": 1.65,
-      "learning_rate": 1.664170146945089e-05,
-      "loss": 2.1763,
+      "epoch": 0.13,
+      "learning_rate": 1.6290787248857908e-05,
+      "loss": 2.5131,
       "step": 710
     },
     {
-      "epoch": 1.66,
-      "learning_rate": 1.6597347254447023e-05,
-      "loss": 2.2591,
+      "epoch": 0.13,
+      "learning_rate": 1.6405511102723105e-05,
+      "loss": 2.8885,
       "step": 715
     },
     {
-      "epoch": 1.67,
-      "learning_rate": 1.6552993039443156e-05,
-      "loss": 2.2595,
+      "epoch": 0.13,
+      "learning_rate": 1.65202349565883e-05,
+      "loss": 2.569,
       "step": 720
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 1.650863882443929e-05,
-      "loss": 2.26,
+      "epoch": 0.13,
+      "learning_rate": 1.66349588104535e-05,
+      "loss": 2.5193,
       "step": 725
     },
     {
-      "epoch": 1.69,
-      "learning_rate": 1.6464284609435422e-05,
-      "loss": 2.2275,
+      "epoch": 0.14,
+      "learning_rate": 1.6749682664318695e-05,
+      "loss": 2.4725,
       "step": 730
     },
     {
-      "epoch": 1.71,
-      "learning_rate": 1.6419930394431555e-05,
-      "loss": 2.1895,
+      "epoch": 0.14,
+      "learning_rate": 1.686440651818389e-05,
+      "loss": 2.5602,
       "step": 735
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 1.6375576179427688e-05,
-      "loss": 2.1943,
+      "epoch": 0.14,
+      "learning_rate": 1.697913037204909e-05,
+      "loss": 2.7077,
       "step": 740
     },
     {
-      "epoch": 1.73,
-      "learning_rate": 1.633122196442382e-05,
-      "loss": 2.2282,
+      "epoch": 0.14,
+      "learning_rate": 1.7093854225914285e-05,
+      "loss": 2.551,
       "step": 745
     },
     {
-      "epoch": 1.74,
-      "learning_rate": 1.6286867749419954e-05,
-      "loss": 2.1926,
+      "epoch": 0.14,
+      "learning_rate": 1.7208578079779482e-05,
+      "loss": 2.6332,
       "step": 750
     },
     {
-      "epoch": 1.75,
-      "learning_rate": 1.6242513534416086e-05,
-      "loss": 2.2124,
+      "epoch": 0.14,
+      "learning_rate": 1.732330193364468e-05,
+      "loss": 2.6119,
       "step": 755
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 1.619815931941222e-05,
-      "loss": 2.2104,
+      "epoch": 0.14,
+      "learning_rate": 1.7438025787509875e-05,
+      "loss": 2.3745,
       "step": 760
     },
     {
-      "epoch": 1.77,
-      "learning_rate": 1.6153805104408352e-05,
-      "loss": 2.1821,
+      "epoch": 0.14,
+      "learning_rate": 1.7552749641375072e-05,
+      "loss": 2.6072,
       "step": 765
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 1.6109450889404485e-05,
-      "loss": 2.2004,
+      "epoch": 0.14,
+      "learning_rate": 1.766747349524027e-05,
+      "loss": 2.5743,
       "step": 770
     },
     {
-      "epoch": 1.8,
-      "learning_rate": 1.606509667440062e-05,
-      "loss": 2.213,
+      "epoch": 0.14,
+      "learning_rate": 1.7782197349105465e-05,
+      "loss": 2.6184,
       "step": 775
     },
     {
-      "epoch": 1.81,
-      "learning_rate": 1.602074245939675e-05,
-      "loss": 2.2022,
+      "epoch": 0.15,
+      "learning_rate": 1.7896921202970662e-05,
+      "loss": 2.4753,
       "step": 780
     },
     {
-      "epoch": 1.82,
-      "learning_rate": 1.5976388244392884e-05,
-      "loss": 2.2303,
+      "epoch": 0.15,
+      "learning_rate": 1.801164505683586e-05,
+      "loss": 2.5953,
       "step": 785
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 1.593203402938902e-05,
-      "loss": 2.173,
+      "epoch": 0.15,
+      "learning_rate": 1.8126368910701055e-05,
+      "loss": 2.3535,
       "step": 790
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 1.588767981438515e-05,
-      "loss": 2.1799,
+      "epoch": 0.15,
+      "learning_rate": 1.8241092764566252e-05,
+      "loss": 2.5899,
       "step": 795
     },
     {
-      "epoch": 1.86,
-      "learning_rate": 1.5843325599381286e-05,
-      "loss": 2.1737,
-      "step": 800
-    },
-    {
-      "epoch": 1.86,
-      "eval_accuracy": 0.5866239287094501,
-      "eval_loss": 2.2810451984405518,
-      "eval_runtime": 58.2336,
-      "eval_samples_per_second": 7.384,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.15,
+      "learning_rate": 1.835581661843145e-05,
+      "loss": 2.5311,
       "step": 800
     },
     {
-      "epoch": 1.87,
-      "learning_rate": 1.579897138437742e-05,
-      "loss": 2.1741,
+      "epoch": 0.15,
+      "learning_rate": 1.8470540472296645e-05,
+      "loss": 2.7737,
       "step": 805
     },
     {
-      "epoch": 1.88,
-      "learning_rate": 1.5754617169373548e-05,
-      "loss": 2.2198,
+      "epoch": 0.15,
+      "learning_rate": 1.8585264326161842e-05,
+      "loss": 2.5879,
       "step": 810
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 1.5710262954369684e-05,
-      "loss": 2.2216,
+      "epoch": 0.15,
+      "learning_rate": 1.8699988180027035e-05,
+      "loss": 2.5586,
       "step": 815
     },
     {
-      "epoch": 1.9,
-      "learning_rate": 1.5665908739365817e-05,
-      "loss": 2.2283,
+      "epoch": 0.15,
+      "learning_rate": 1.8814712033892232e-05,
+      "loss": 2.6661,
       "step": 820
     },
     {
-      "epoch": 1.91,
-      "learning_rate": 1.562155452436195e-05,
-      "loss": 2.1685,
+      "epoch": 0.15,
+      "learning_rate": 1.892943588775743e-05,
+      "loss": 2.6164,
       "step": 825
     },
     {
-      "epoch": 1.93,
-      "learning_rate": 1.5577200309358083e-05,
-      "loss": 2.2341,
+      "epoch": 0.15,
+      "learning_rate": 1.9044159741622625e-05,
+      "loss": 2.7437,
       "step": 830
     },
     {
-      "epoch": 1.94,
-      "learning_rate": 1.5532846094354216e-05,
-      "loss": 2.2037,
+      "epoch": 0.16,
+      "learning_rate": 1.9158883595487822e-05,
+      "loss": 2.4667,
       "step": 835
     },
     {
-      "epoch": 1.95,
-      "learning_rate": 1.548849187935035e-05,
-      "loss": 2.1839,
+      "epoch": 0.16,
+      "learning_rate": 1.927360744935302e-05,
+      "loss": 2.6063,
       "step": 840
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 1.5444137664346482e-05,
-      "loss": 2.1978,
+      "epoch": 0.16,
+      "learning_rate": 1.9388331303218215e-05,
+      "loss": 2.7781,
       "step": 845
     },
     {
-      "epoch": 1.97,
-      "learning_rate": 1.5399783449342615e-05,
-      "loss": 2.2348,
+      "epoch": 0.16,
+      "learning_rate": 1.9503055157083412e-05,
+      "loss": 2.2327,
       "step": 850
     },
     {
-      "epoch": 1.98,
-      "learning_rate": 1.5355429234338747e-05,
-      "loss": 2.1423,
+      "epoch": 0.16,
+      "learning_rate": 1.961777901094861e-05,
+      "loss": 2.6243,
       "step": 855
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 1.531107501933488e-05,
-      "loss": 2.2643,
+      "epoch": 0.16,
+      "learning_rate": 1.9732502864813805e-05,
+      "loss": 2.5766,
       "step": 860
     },
     {
-      "epoch": 2.01,
-      "learning_rate": 1.5266720804331013e-05,
-      "loss": 2.2016,
+      "epoch": 0.16,
+      "learning_rate": 1.9847226718679002e-05,
+      "loss": 2.3461,
       "step": 865
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 1.5222366589327146e-05,
-      "loss": 2.1568,
+      "epoch": 0.16,
+      "learning_rate": 1.99619505725442e-05,
+      "loss": 2.4645,
       "step": 870
     },
     {
-      "epoch": 2.03,
-      "learning_rate": 1.517801237432328e-05,
-      "loss": 2.1041,
+      "epoch": 0.16,
+      "learning_rate": 2.0076674426409395e-05,
+      "loss": 2.4267,
       "step": 875
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 1.5133658159319412e-05,
-      "loss": 2.0597,
+      "epoch": 0.16,
+      "learning_rate": 2.0191398280274592e-05,
+      "loss": 2.4677,
       "step": 880
     },
     {
-      "epoch": 2.05,
-      "learning_rate": 1.5089303944315546e-05,
-      "loss": 2.0842,
+      "epoch": 0.16,
+      "learning_rate": 2.030612213413979e-05,
+      "loss": 2.389,
       "step": 885
     },
     {
-      "epoch": 2.06,
-      "learning_rate": 1.504494972931168e-05,
-      "loss": 2.09,
+      "epoch": 0.17,
+      "learning_rate": 2.0420845988004985e-05,
+      "loss": 2.4977,
       "step": 890
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 1.500059551430781e-05,
-      "loss": 2.1268,
+      "epoch": 0.17,
+      "learning_rate": 2.0535569841870182e-05,
+      "loss": 2.5927,
       "step": 895
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 1.4956241299303945e-05,
-      "loss": 2.0763,
-      "step": 900
-    },
-    {
-      "epoch": 2.09,
-      "eval_accuracy": 0.5876264520675624,
-      "eval_loss": 2.2779171466827393,
-      "eval_runtime": 58.2423,
-      "eval_samples_per_second": 7.383,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.17,
+      "learning_rate": 2.065029369573538e-05,
+      "loss": 2.2544,
       "step": 900
     },
     {
-      "epoch": 2.1,
-      "learning_rate": 1.4911887084300078e-05,
-      "loss": 2.0782,
+      "epoch": 0.17,
+      "learning_rate": 2.0765017549600575e-05,
+      "loss": 2.615,
       "step": 905
     },
     {
-      "epoch": 2.11,
-      "learning_rate": 1.486753286929621e-05,
-      "loss": 2.0983,
+      "epoch": 0.17,
+      "learning_rate": 2.0879741403465772e-05,
+      "loss": 2.6026,
       "step": 910
     },
     {
-      "epoch": 2.12,
-      "learning_rate": 1.4823178654292344e-05,
-      "loss": 2.0846,
+      "epoch": 0.17,
+      "learning_rate": 2.099446525733097e-05,
+      "loss": 2.6839,
       "step": 915
     },
     {
-      "epoch": 2.13,
-      "learning_rate": 1.4778824439288477e-05,
-      "loss": 2.0959,
+      "epoch": 0.17,
+      "learning_rate": 2.1109189111196165e-05,
+      "loss": 2.3331,
       "step": 920
     },
     {
-      "epoch": 2.15,
-      "learning_rate": 1.473447022428461e-05,
-      "loss": 2.091,
+      "epoch": 0.17,
+      "learning_rate": 2.1223912965061362e-05,
+      "loss": 2.5703,
       "step": 925
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 1.4690116009280744e-05,
-      "loss": 2.1233,
+      "epoch": 0.17,
+      "learning_rate": 2.133863681892656e-05,
+      "loss": 2.8509,
       "step": 930
     },
     {
-      "epoch": 2.17,
-      "learning_rate": 1.4645761794276875e-05,
-      "loss": 2.0943,
+      "epoch": 0.17,
+      "learning_rate": 2.1453360672791756e-05,
+      "loss": 2.5015,
       "step": 935
     },
     {
-      "epoch": 2.18,
-      "learning_rate": 1.4601407579273008e-05,
-      "loss": 2.103,
+      "epoch": 0.17,
+      "learning_rate": 2.156808452665695e-05,
+      "loss": 2.5277,
       "step": 940
     },
     {
-      "epoch": 2.19,
-      "learning_rate": 1.4557053364269143e-05,
-      "loss": 2.1249,
+      "epoch": 0.18,
+      "learning_rate": 2.1682808380522145e-05,
+      "loss": 2.4037,
       "step": 945
     },
     {
-      "epoch": 2.2,
-      "learning_rate": 1.4512699149265274e-05,
-      "loss": 2.0495,
+      "epoch": 0.18,
+      "learning_rate": 2.1797532234387342e-05,
+      "loss": 2.4789,
       "step": 950
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 1.4468344934261409e-05,
-      "loss": 2.0974,
+      "epoch": 0.18,
+      "learning_rate": 2.191225608825254e-05,
+      "loss": 2.8376,
       "step": 955
     },
     {
-      "epoch": 2.23,
-      "learning_rate": 1.4423990719257541e-05,
-      "loss": 2.1103,
+      "epoch": 0.18,
+      "learning_rate": 2.2026979942117736e-05,
+      "loss": 2.6608,
       "step": 960
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 1.4379636504253674e-05,
-      "loss": 2.1507,
+      "epoch": 0.18,
+      "learning_rate": 2.2141703795982932e-05,
+      "loss": 2.5525,
       "step": 965
     },
     {
-      "epoch": 2.25,
-      "learning_rate": 1.4335282289249809e-05,
-      "loss": 2.1357,
+      "epoch": 0.18,
+      "learning_rate": 2.225642764984813e-05,
+      "loss": 2.5682,
       "step": 970
     },
     {
-      "epoch": 2.26,
-      "learning_rate": 1.429092807424594e-05,
-      "loss": 2.0539,
+      "epoch": 0.18,
+      "learning_rate": 2.2371151503713326e-05,
+      "loss": 2.8494,
       "step": 975
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 1.4246573859242073e-05,
-      "loss": 2.0619,
+      "epoch": 0.18,
+      "learning_rate": 2.2485875357578522e-05,
+      "loss": 2.4472,
       "step": 980
     },
     {
-      "epoch": 2.29,
-      "learning_rate": 1.4202219644238208e-05,
-      "loss": 2.091,
+      "epoch": 0.18,
+      "learning_rate": 2.260059921144372e-05,
+      "loss": 2.2839,
       "step": 985
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 1.4157865429234339e-05,
-      "loss": 2.1239,
+      "epoch": 0.18,
+      "learning_rate": 2.2715323065308916e-05,
+      "loss": 2.2856,
       "step": 990
     },
     {
-      "epoch": 2.31,
-      "learning_rate": 1.4113511214230472e-05,
-      "loss": 2.0991,
+      "epoch": 0.19,
+      "learning_rate": 2.2830046919174112e-05,
+      "loss": 2.5793,
       "step": 995
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 1.4069156999226606e-05,
-      "loss": 2.0852,
+      "epoch": 0.19,
+      "learning_rate": 2.294477077303931e-05,
+      "loss": 2.4891,
       "step": 1000
     },
     {
-      "epoch": 2.32,
-      "eval_accuracy": 0.5893882561549478,
-      "eval_loss": 2.2678375244140625,
-      "eval_runtime": 58.2516,
-      "eval_samples_per_second": 7.382,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.19,
+      "eval_accuracy": 0.8446346521695359,
+      "eval_loss": 2.446685791015625,
+      "eval_runtime": 129.7958,
+      "eval_samples_per_second": 82.822,
+      "eval_steps_per_second": 10.355,
       "step": 1000
     },
     {
-      "epoch": 2.33,
-      "learning_rate": 1.4024802784222737e-05,
-      "loss": 2.0668,
+      "epoch": 0.19,
+      "learning_rate": 2.2933006637081024e-05,
+      "loss": 2.1619,
       "step": 1005
     },
     {
-      "epoch": 2.34,
-      "learning_rate": 1.398044856921887e-05,
-      "loss": 2.1403,
+      "epoch": 0.19,
+      "learning_rate": 2.292124250112274e-05,
+      "loss": 2.8486,
       "step": 1010
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 1.3936094354215005e-05,
-      "loss": 2.1447,
+      "epoch": 0.19,
+      "learning_rate": 2.2909478365164453e-05,
+      "loss": 2.6101,
       "step": 1015
     },
     {
-      "epoch": 2.37,
-      "learning_rate": 1.3891740139211138e-05,
-      "loss": 2.0871,
+      "epoch": 0.19,
+      "learning_rate": 2.2897714229206168e-05,
+      "loss": 2.4108,
       "step": 1020
     },
     {
-      "epoch": 2.38,
-      "learning_rate": 1.3847385924207269e-05,
-      "loss": 2.1181,
+      "epoch": 0.19,
+      "learning_rate": 2.2885950093247883e-05,
+      "loss": 2.4996,
       "step": 1025
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 1.3803031709203403e-05,
-      "loss": 2.0435,
+      "epoch": 0.19,
+      "learning_rate": 2.2874185957289597e-05,
+      "loss": 2.4466,
       "step": 1030
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 1.3758677494199536e-05,
-      "loss": 2.0696,
+      "epoch": 0.19,
+      "learning_rate": 2.2862421821331315e-05,
+      "loss": 2.4066,
       "step": 1035
     },
     {
-      "epoch": 2.41,
-      "learning_rate": 1.3714323279195671e-05,
-      "loss": 2.1185,
+      "epoch": 0.19,
+      "learning_rate": 2.2850657685373027e-05,
+      "loss": 2.1801,
       "step": 1040
     },
     {
-      "epoch": 2.42,
-      "learning_rate": 1.3669969064191802e-05,
-      "loss": 2.0651,
+      "epoch": 0.19,
+      "learning_rate": 2.283889354941474e-05,
+      "loss": 2.6545,
       "step": 1045
     },
     {
-      "epoch": 2.44,
-      "learning_rate": 1.3625614849187935e-05,
-      "loss": 2.0914,
+      "epoch": 0.2,
+      "learning_rate": 2.282712941345646e-05,
+      "loss": 2.6547,
       "step": 1050
     },
     {
-      "epoch": 2.45,
-      "learning_rate": 1.358126063418407e-05,
-      "loss": 2.0895,
+      "epoch": 0.2,
+      "learning_rate": 2.281536527749817e-05,
+      "loss": 2.5683,
       "step": 1055
     },
     {
-      "epoch": 2.46,
-      "learning_rate": 1.35369064191802e-05,
-      "loss": 2.1037,
+      "epoch": 0.2,
+      "learning_rate": 2.280360114153989e-05,
+      "loss": 2.3329,
       "step": 1060
     },
     {
-      "epoch": 2.47,
-      "learning_rate": 1.3492552204176334e-05,
-      "loss": 2.1207,
+      "epoch": 0.2,
+      "learning_rate": 2.2791837005581604e-05,
+      "loss": 2.5781,
       "step": 1065
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 1.3448197989172468e-05,
-      "loss": 2.0958,
+      "epoch": 0.2,
+      "learning_rate": 2.2780072869623315e-05,
+      "loss": 2.8662,
       "step": 1070
     },
     {
-      "epoch": 2.49,
-      "learning_rate": 1.3403843774168601e-05,
-      "loss": 2.1155,
+      "epoch": 0.2,
+      "learning_rate": 2.2768308733665033e-05,
+      "loss": 2.576,
       "step": 1075
     },
     {
-      "epoch": 2.51,
-      "learning_rate": 1.3359489559164732e-05,
-      "loss": 2.0974,
+      "epoch": 0.2,
+      "learning_rate": 2.2756544597706748e-05,
+      "loss": 2.4152,
       "step": 1080
     },
     {
-      "epoch": 2.52,
-      "learning_rate": 1.3315135344160867e-05,
-      "loss": 2.0724,
+      "epoch": 0.2,
+      "learning_rate": 2.2744780461748463e-05,
+      "loss": 2.44,
       "step": 1085
     },
     {
-      "epoch": 2.53,
-      "learning_rate": 1.3270781129157e-05,
-      "loss": 2.052,
+      "epoch": 0.2,
+      "learning_rate": 2.2733016325790177e-05,
+      "loss": 2.6596,
       "step": 1090
     },
     {
-      "epoch": 2.54,
-      "learning_rate": 1.3226426914153131e-05,
-      "loss": 2.0653,
+      "epoch": 0.2,
+      "learning_rate": 2.2721252189831892e-05,
+      "loss": 2.1703,
       "step": 1095
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 1.3182072699149266e-05,
-      "loss": 2.0946,
-      "step": 1100
-    },
-    {
-      "epoch": 2.55,
-      "eval_accuracy": 0.5905817363431767,
-      "eval_loss": 2.25939679145813,
-      "eval_runtime": 58.2432,
-      "eval_samples_per_second": 7.383,
-      "eval_steps_per_second": 1.854,
+      "epoch": 0.2,
+      "learning_rate": 2.2709488053873607e-05,
+      "loss": 3.0027,
       "step": 1100
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 1.3137718484145398e-05,
-      "loss": 2.1256,
+      "epoch": 0.21,
+      "learning_rate": 2.269772391791532e-05,
+      "loss": 2.5836,
       "step": 1105
     },
     {
-      "epoch": 2.58,
-      "learning_rate": 1.3093364269141531e-05,
-      "loss": 2.0605,
+      "epoch": 0.21,
+      "learning_rate": 2.2685959781957036e-05,
+      "loss": 2.6088,
       "step": 1110
     },
     {
-      "epoch": 2.59,
-      "learning_rate": 1.3049010054137666e-05,
-      "loss": 2.0764,
+      "epoch": 0.21,
+      "learning_rate": 2.267419564599875e-05,
+      "loss": 2.7861,
       "step": 1115
     },
     {
-      "epoch": 2.6,
-      "learning_rate": 1.3004655839133797e-05,
-      "loss": 2.1162,
+      "epoch": 0.21,
+      "learning_rate": 2.2662431510040466e-05,
+      "loss": 2.498,
       "step": 1120
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 1.2960301624129932e-05,
-      "loss": 2.1343,
+      "epoch": 0.21,
+      "learning_rate": 2.2650667374082184e-05,
+      "loss": 2.4744,
       "step": 1125
     },
     {
-      "epoch": 2.62,
-      "learning_rate": 1.2915947409126065e-05,
-      "loss": 2.1427,
+      "epoch": 0.21,
+      "learning_rate": 2.2638903238123895e-05,
+      "loss": 2.5484,
       "step": 1130
     },
     {
-      "epoch": 2.63,
-      "learning_rate": 1.2871593194122196e-05,
-      "loss": 2.0705,
+      "epoch": 0.21,
+      "learning_rate": 2.262713910216561e-05,
+      "loss": 2.8166,
       "step": 1135
     },
     {
-      "epoch": 2.65,
-      "learning_rate": 1.282723897911833e-05,
-      "loss": 2.0804,
+      "epoch": 0.21,
+      "learning_rate": 2.2615374966207328e-05,
+      "loss": 2.3641,
       "step": 1140
     },
     {
-      "epoch": 2.66,
-      "learning_rate": 1.2782884764114463e-05,
-      "loss": 2.0478,
+      "epoch": 0.21,
+      "learning_rate": 2.260361083024904e-05,
+      "loss": 2.4315,
       "step": 1145
     },
     {
-      "epoch": 2.67,
-      "learning_rate": 1.2738530549110594e-05,
-      "loss": 2.0797,
+      "epoch": 0.21,
+      "learning_rate": 2.2591846694290757e-05,
+      "loss": 2.3159,
       "step": 1150
     },
     {
-      "epoch": 2.68,
-      "learning_rate": 1.2694176334106729e-05,
-      "loss": 2.0719,
+      "epoch": 0.21,
+      "learning_rate": 2.2580082558332472e-05,
+      "loss": 2.6,
       "step": 1155
     },
     {
-      "epoch": 2.69,
-      "learning_rate": 1.2649822119102862e-05,
-      "loss": 2.0958,
+      "epoch": 0.22,
+      "learning_rate": 2.2568318422374183e-05,
+      "loss": 2.2356,
       "step": 1160
     },
     {
-      "epoch": 2.7,
-      "learning_rate": 1.2605467904098995e-05,
-      "loss": 2.0709,
+      "epoch": 0.22,
+      "learning_rate": 2.25565542864159e-05,
+      "loss": 2.3218,
       "step": 1165
     },
     {
-      "epoch": 2.71,
-      "learning_rate": 1.256111368909513e-05,
-      "loss": 2.0724,
+      "epoch": 0.22,
+      "learning_rate": 2.2544790150457616e-05,
+      "loss": 2.5485,
       "step": 1170
     },
     {
-      "epoch": 2.73,
-      "learning_rate": 1.251675947409126e-05,
-      "loss": 2.0813,
+      "epoch": 0.22,
+      "learning_rate": 2.253302601449933e-05,
+      "loss": 2.7599,
       "step": 1175
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 1.2472405259087393e-05,
-      "loss": 2.1098,
+      "epoch": 0.22,
+      "learning_rate": 2.2521261878541046e-05,
+      "loss": 2.5984,
       "step": 1180
     },
     {
-      "epoch": 2.75,
-      "learning_rate": 1.2428051044083528e-05,
-      "loss": 2.0715,
+      "epoch": 0.22,
+      "learning_rate": 2.2509497742582757e-05,
+      "loss": 2.297,
       "step": 1185
     },
     {
-      "epoch": 2.76,
-      "learning_rate": 1.238369682907966e-05,
-      "loss": 2.0848,
+      "epoch": 0.22,
+      "learning_rate": 2.2497733606624475e-05,
+      "loss": 2.4894,
       "step": 1190
     },
     {
-      "epoch": 2.77,
-      "learning_rate": 1.2339342614075792e-05,
-      "loss": 2.1323,
+      "epoch": 0.22,
+      "learning_rate": 2.248596947066619e-05,
+      "loss": 2.4761,
       "step": 1195
     },
     {
-      "epoch": 2.78,
-      "learning_rate": 1.2294988399071927e-05,
-      "loss": 2.0497,
-      "step": 1200
-    },
-    {
-      "epoch": 2.78,
-      "eval_accuracy": 0.5920070926822615,
-      "eval_loss": 2.2515528202056885,
-      "eval_runtime": 58.2138,
-      "eval_samples_per_second": 7.387,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.22,
+      "learning_rate": 2.2474205334707904e-05,
+      "loss": 2.2026,
       "step": 1200
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 1.2250634184068058e-05,
-      "loss": 2.0428,
+      "epoch": 0.22,
+      "learning_rate": 2.246244119874962e-05,
+      "loss": 2.6491,
       "step": 1205
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 1.2206279969064194e-05,
-      "loss": 2.0776,
+      "epoch": 0.23,
+      "learning_rate": 2.2450677062791334e-05,
+      "loss": 2.3317,
       "step": 1210
     },
     {
-      "epoch": 2.82,
-      "learning_rate": 1.2161925754060325e-05,
-      "loss": 2.1259,
+      "epoch": 0.23,
+      "learning_rate": 2.243891292683305e-05,
+      "loss": 2.52,
       "step": 1215
     },
     {
-      "epoch": 2.83,
-      "learning_rate": 1.2117571539056458e-05,
-      "loss": 2.131,
+      "epoch": 0.23,
+      "learning_rate": 2.2427148790874763e-05,
+      "loss": 2.2593,
       "step": 1220
     },
     {
-      "epoch": 2.84,
-      "learning_rate": 1.2073217324052593e-05,
-      "loss": 2.0149,
+      "epoch": 0.23,
+      "learning_rate": 2.2415384654916478e-05,
+      "loss": 2.573,
       "step": 1225
     },
     {
-      "epoch": 2.85,
-      "learning_rate": 1.2028863109048724e-05,
-      "loss": 2.1355,
+      "epoch": 0.23,
+      "learning_rate": 2.2403620518958193e-05,
+      "loss": 2.7403,
       "step": 1230
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 1.1984508894044857e-05,
-      "loss": 2.0664,
+      "epoch": 0.23,
+      "learning_rate": 2.2391856382999907e-05,
+      "loss": 2.3135,
       "step": 1235
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 1.1940154679040991e-05,
-      "loss": 2.0407,
+      "epoch": 0.23,
+      "learning_rate": 2.2380092247041622e-05,
+      "loss": 2.3107,
       "step": 1240
     },
     {
-      "epoch": 2.89,
-      "learning_rate": 1.1895800464037123e-05,
-      "loss": 2.0636,
+      "epoch": 0.23,
+      "learning_rate": 2.2368328111083337e-05,
+      "loss": 2.47,
       "step": 1245
     },
     {
-      "epoch": 2.9,
-      "learning_rate": 1.1851446249033255e-05,
-      "loss": 2.0867,
+      "epoch": 0.23,
+      "learning_rate": 2.235656397512505e-05,
+      "loss": 2.6732,
       "step": 1250
     },
     {
-      "epoch": 2.91,
-      "learning_rate": 1.180709203402939e-05,
-      "loss": 2.0693,
+      "epoch": 0.23,
+      "learning_rate": 2.234479983916677e-05,
+      "loss": 2.7278,
       "step": 1255
     },
     {
-      "epoch": 2.92,
-      "learning_rate": 1.1762737819025523e-05,
-      "loss": 2.0812,
+      "epoch": 0.23,
+      "learning_rate": 2.233303570320848e-05,
+      "loss": 2.6984,
       "step": 1260
     },
     {
-      "epoch": 2.94,
-      "learning_rate": 1.1718383604021654e-05,
-      "loss": 2.1114,
+      "epoch": 0.24,
+      "learning_rate": 2.2321271567250196e-05,
+      "loss": 2.7326,
       "step": 1265
     },
     {
-      "epoch": 2.95,
-      "learning_rate": 1.1674029389017789e-05,
-      "loss": 2.158,
+      "epoch": 0.24,
+      "learning_rate": 2.2309507431291914e-05,
+      "loss": 2.5126,
       "step": 1270
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 1.1629675174013922e-05,
-      "loss": 2.0781,
+      "epoch": 0.24,
+      "learning_rate": 2.2297743295333625e-05,
+      "loss": 2.552,
       "step": 1275
     },
     {
-      "epoch": 2.97,
-      "learning_rate": 1.1585320959010053e-05,
-      "loss": 2.1295,
+      "epoch": 0.24,
+      "learning_rate": 2.2285979159375343e-05,
+      "loss": 2.3433,
       "step": 1280
     },
     {
-      "epoch": 2.98,
-      "learning_rate": 1.1540966744006187e-05,
-      "loss": 2.0517,
+      "epoch": 0.24,
+      "learning_rate": 2.2274215023417058e-05,
+      "loss": 2.3537,
       "step": 1285
     },
     {
-      "epoch": 2.99,
-      "learning_rate": 1.149661252900232e-05,
-      "loss": 2.0954,
+      "epoch": 0.24,
+      "learning_rate": 2.226245088745877e-05,
+      "loss": 2.636,
       "step": 1290
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 1.1452258313998453e-05,
-      "loss": 2.0296,
+      "epoch": 0.24,
+      "learning_rate": 2.2250686751500487e-05,
+      "loss": 2.5576,
       "step": 1295
     },
     {
-      "epoch": 3.02,
-      "learning_rate": 1.1407904098994586e-05,
-      "loss": 2.0141,
-      "step": 1300
-    },
-    {
-      "epoch": 3.02,
-      "eval_accuracy": 0.5928277523926436,
-      "eval_loss": 2.251300573348999,
-      "eval_runtime": 58.199,
-      "eval_samples_per_second": 7.388,
-      "eval_steps_per_second": 1.856,
+      "epoch": 0.24,
+      "learning_rate": 2.2238922615542202e-05,
+      "loss": 2.2962,
       "step": 1300
     },
     {
-      "epoch": 3.03,
-      "learning_rate": 1.1363549883990719e-05,
-      "loss": 1.9846,
+      "epoch": 0.24,
+      "learning_rate": 2.2227158479583917e-05,
+      "loss": 2.457,
       "step": 1305
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 1.1319195668986852e-05,
-      "loss": 2.0017,
+      "epoch": 0.24,
+      "learning_rate": 2.221539434362563e-05,
+      "loss": 2.375,
       "step": 1310
     },
     {
-      "epoch": 3.05,
-      "learning_rate": 1.1274841453982986e-05,
-      "loss": 2.0075,
+      "epoch": 0.24,
+      "learning_rate": 2.2203630207667346e-05,
+      "loss": 2.536,
       "step": 1315
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 1.123048723897912e-05,
-      "loss": 1.9859,
+      "epoch": 0.25,
+      "learning_rate": 2.219186607170906e-05,
+      "loss": 2.5033,
       "step": 1320
     },
     {
-      "epoch": 3.07,
-      "learning_rate": 1.118613302397525e-05,
-      "loss": 1.959,
+      "epoch": 0.25,
+      "learning_rate": 2.2180101935750776e-05,
+      "loss": 2.4336,
       "step": 1325
     },
     {
-      "epoch": 3.09,
-      "learning_rate": 1.1141778808971385e-05,
-      "loss": 2.0016,
+      "epoch": 0.25,
+      "learning_rate": 2.216833779979249e-05,
+      "loss": 2.4067,
       "step": 1330
     },
     {
-      "epoch": 3.1,
-      "learning_rate": 1.1097424593967518e-05,
-      "loss": 2.0378,
+      "epoch": 0.25,
+      "learning_rate": 2.2156573663834205e-05,
+      "loss": 2.583,
       "step": 1335
     },
     {
-      "epoch": 3.11,
-      "learning_rate": 1.105307037896365e-05,
-      "loss": 1.9978,
+      "epoch": 0.25,
+      "learning_rate": 2.214480952787592e-05,
+      "loss": 2.2076,
       "step": 1340
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 1.1008716163959784e-05,
-      "loss": 2.0025,
+      "epoch": 0.25,
+      "learning_rate": 2.2133045391917635e-05,
+      "loss": 2.2939,
       "step": 1345
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 1.0964361948955917e-05,
-      "loss": 1.9221,
+      "epoch": 0.25,
+      "learning_rate": 2.212128125595935e-05,
+      "loss": 2.6029,
       "step": 1350
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.0920007733952051e-05,
-      "loss": 1.9976,
+      "epoch": 0.25,
+      "learning_rate": 2.2109517120001064e-05,
+      "loss": 2.231,
       "step": 1355
     },
     {
-      "epoch": 3.16,
-      "learning_rate": 1.0875653518948182e-05,
-      "loss": 1.9718,
+      "epoch": 0.25,
+      "learning_rate": 2.2097752984042782e-05,
+      "loss": 2.3771,
       "step": 1360
     },
     {
-      "epoch": 3.17,
-      "learning_rate": 1.0831299303944315e-05,
-      "loss": 2.0334,
+      "epoch": 0.25,
+      "learning_rate": 2.2085988848084493e-05,
+      "loss": 2.5636,
       "step": 1365
     },
     {
-      "epoch": 3.18,
-      "learning_rate": 1.078694508894045e-05,
-      "loss": 2.0087,
+      "epoch": 0.25,
+      "learning_rate": 2.2074224712126208e-05,
+      "loss": 2.6581,
       "step": 1370
     },
     {
-      "epoch": 3.19,
-      "learning_rate": 1.0742590873936581e-05,
-      "loss": 2.0111,
+      "epoch": 0.26,
+      "learning_rate": 2.2062460576167926e-05,
+      "loss": 2.6923,
       "step": 1375
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 1.0698236658932716e-05,
-      "loss": 2.0332,
+      "epoch": 0.26,
+      "learning_rate": 2.2050696440209638e-05,
+      "loss": 2.3704,
       "step": 1380
     },
     {
-      "epoch": 3.21,
-      "learning_rate": 1.0653882443928848e-05,
-      "loss": 1.9301,
+      "epoch": 0.26,
+      "learning_rate": 2.2038932304251356e-05,
+      "loss": 2.4697,
       "step": 1385
     },
     {
-      "epoch": 3.23,
-      "learning_rate": 1.0609528228924981e-05,
-      "loss": 1.9898,
+      "epoch": 0.26,
+      "learning_rate": 2.202716816829307e-05,
+      "loss": 2.4409,
       "step": 1390
     },
     {
-      "epoch": 3.24,
-      "learning_rate": 1.0565174013921114e-05,
-      "loss": 1.9965,
+      "epoch": 0.26,
+      "learning_rate": 2.201540403233478e-05,
+      "loss": 2.779,
       "step": 1395
     },
     {
-      "epoch": 3.25,
-      "learning_rate": 1.0520819798917247e-05,
-      "loss": 2.0316,
-      "step": 1400
-    },
-    {
-      "epoch": 3.25,
-      "eval_accuracy": 0.5931710200277343,
-      "eval_loss": 2.2504708766937256,
-      "eval_runtime": 58.2338,
-      "eval_samples_per_second": 7.384,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.26,
+      "learning_rate": 2.20036398963765e-05,
+      "loss": 2.5542,
       "step": 1400
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 1.047646558391338e-05,
-      "loss": 2.0107,
+      "epoch": 0.26,
+      "learning_rate": 2.1991875760418215e-05,
+      "loss": 2.5959,
       "step": 1405
     },
     {
-      "epoch": 3.27,
-      "learning_rate": 1.0432111368909513e-05,
-      "loss": 1.9955,
+      "epoch": 0.26,
+      "learning_rate": 2.198011162445993e-05,
+      "loss": 1.97,
       "step": 1410
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 1.0387757153905646e-05,
-      "loss": 1.9849,
+      "epoch": 0.26,
+      "learning_rate": 2.1968347488501644e-05,
+      "loss": 2.4478,
       "step": 1415
     },
     {
-      "epoch": 3.29,
-      "learning_rate": 1.0343402938901779e-05,
-      "loss": 1.9644,
+      "epoch": 0.26,
+      "learning_rate": 2.195658335254336e-05,
+      "loss": 2.1968,
       "step": 1420
     },
     {
-      "epoch": 3.31,
-      "learning_rate": 1.0299048723897911e-05,
-      "loss": 1.9953,
+      "epoch": 0.27,
+      "learning_rate": 2.1944819216585073e-05,
+      "loss": 2.4421,
       "step": 1425
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 1.0254694508894044e-05,
-      "loss": 1.9783,
+      "epoch": 0.27,
+      "learning_rate": 2.1933055080626788e-05,
+      "loss": 2.0877,
       "step": 1430
     },
     {
-      "epoch": 3.33,
-      "learning_rate": 1.0210340293890179e-05,
-      "loss": 2.0231,
+      "epoch": 0.27,
+      "learning_rate": 2.1921290944668503e-05,
+      "loss": 2.2301,
       "step": 1435
     },
     {
-      "epoch": 3.34,
-      "learning_rate": 1.0165986078886312e-05,
-      "loss": 1.9832,
+      "epoch": 0.27,
+      "learning_rate": 2.1909526808710218e-05,
+      "loss": 2.5126,
       "step": 1440
     },
     {
-      "epoch": 3.35,
-      "learning_rate": 1.0121631863882443e-05,
-      "loss": 2.0026,
+      "epoch": 0.27,
+      "learning_rate": 2.1897762672751932e-05,
+      "loss": 2.5239,
       "step": 1445
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 1.0077277648878578e-05,
-      "loss": 2.0112,
+      "epoch": 0.27,
+      "learning_rate": 2.188599853679365e-05,
+      "loss": 2.5952,
       "step": 1450
     },
     {
-      "epoch": 3.38,
-      "learning_rate": 1.003292343387471e-05,
-      "loss": 2.0493,
+      "epoch": 0.27,
+      "learning_rate": 2.187423440083536e-05,
+      "loss": 2.4527,
       "step": 1455
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 9.988569218870843e-06,
-      "loss": 2.0369,
+      "epoch": 0.27,
+      "learning_rate": 2.1862470264877076e-05,
+      "loss": 2.3795,
       "step": 1460
     },
     {
-      "epoch": 3.4,
-      "learning_rate": 9.944215003866976e-06,
-      "loss": 1.9922,
+      "epoch": 0.27,
+      "learning_rate": 2.1850706128918794e-05,
+      "loss": 2.513,
       "step": 1465
     },
     {
-      "epoch": 3.41,
-      "learning_rate": 9.899860788863109e-06,
-      "loss": 2.0055,
+      "epoch": 0.27,
+      "learning_rate": 2.1838941992960506e-05,
+      "loss": 2.3036,
       "step": 1470
     },
     {
-      "epoch": 3.42,
-      "learning_rate": 9.855506573859244e-06,
-      "loss": 1.9911,
+      "epoch": 0.27,
+      "learning_rate": 2.1827177857002224e-05,
+      "loss": 2.1708,
       "step": 1475
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 9.811152358855375e-06,
-      "loss": 2.0245,
+      "epoch": 0.28,
+      "learning_rate": 2.181541372104394e-05,
+      "loss": 2.3897,
       "step": 1480
     },
     {
-      "epoch": 3.45,
-      "learning_rate": 9.766798143851508e-06,
-      "loss": 1.9636,
+      "epoch": 0.28,
+      "learning_rate": 2.180364958508565e-05,
+      "loss": 2.402,
       "step": 1485
     },
     {
-      "epoch": 3.46,
-      "learning_rate": 9.722443928847642e-06,
-      "loss": 1.9372,
+      "epoch": 0.28,
+      "learning_rate": 2.1791885449127368e-05,
+      "loss": 2.2133,
       "step": 1490
     },
     {
-      "epoch": 3.47,
-      "learning_rate": 9.678089713843774e-06,
-      "loss": 1.9605,
+      "epoch": 0.28,
+      "learning_rate": 2.1780121313169083e-05,
+      "loss": 2.3358,
       "step": 1495
     },
     {
-      "epoch": 3.48,
-      "learning_rate": 9.633735498839908e-06,
-      "loss": 1.9783,
-      "step": 1500
-    },
-    {
-      "epoch": 3.48,
-      "eval_accuracy": 0.593768896769647,
-      "eval_loss": 2.2429721355438232,
-      "eval_runtime": 58.2214,
-      "eval_samples_per_second": 7.386,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.28,
+      "learning_rate": 2.1768357177210798e-05,
+      "loss": 2.4947,
       "step": 1500
     },
     {
-      "epoch": 3.49,
-      "learning_rate": 9.589381283836041e-06,
-      "loss": 1.9702,
+      "epoch": 0.28,
+      "learning_rate": 2.1756593041252512e-05,
+      "loss": 2.2271,
       "step": 1505
     },
     {
-      "epoch": 3.5,
-      "learning_rate": 9.545027068832172e-06,
-      "loss": 1.9922,
+      "epoch": 0.28,
+      "learning_rate": 2.1744828905294227e-05,
+      "loss": 2.4711,
       "step": 1510
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 9.500672853828307e-06,
-      "loss": 1.9904,
+      "epoch": 0.28,
+      "learning_rate": 2.173306476933594e-05,
+      "loss": 2.3014,
       "step": 1515
     },
     {
-      "epoch": 3.53,
-      "learning_rate": 9.45631863882444e-06,
-      "loss": 2.0219,
+      "epoch": 0.28,
+      "learning_rate": 2.1721300633377656e-05,
+      "loss": 2.4893,
       "step": 1520
     },
     {
-      "epoch": 3.54,
-      "learning_rate": 9.411964423820573e-06,
-      "loss": 2.0095,
+      "epoch": 0.28,
+      "learning_rate": 2.170953649741937e-05,
+      "loss": 2.0982,
       "step": 1525
     },
     {
-      "epoch": 3.55,
-      "learning_rate": 9.367610208816705e-06,
-      "loss": 1.9953,
+      "epoch": 0.28,
+      "learning_rate": 2.1697772361461086e-05,
+      "loss": 2.4746,
       "step": 1530
     },
     {
-      "epoch": 3.56,
-      "learning_rate": 9.323255993812838e-06,
-      "loss": 1.9974,
+      "epoch": 0.29,
+      "learning_rate": 2.16860082255028e-05,
+      "loss": 2.1026,
       "step": 1535
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 9.278901778808971e-06,
-      "loss": 1.9874,
+      "epoch": 0.29,
+      "learning_rate": 2.1674244089544515e-05,
+      "loss": 2.1993,
       "step": 1540
     },
     {
-      "epoch": 3.58,
-      "learning_rate": 9.234547563805104e-06,
-      "loss": 1.9871,
+      "epoch": 0.29,
+      "learning_rate": 2.166247995358623e-05,
+      "loss": 2.5912,
       "step": 1545
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 9.190193348801237e-06,
-      "loss": 1.997,
+      "epoch": 0.29,
+      "learning_rate": 2.1650715817627945e-05,
+      "loss": 2.6338,
       "step": 1550
     },
     {
-      "epoch": 3.61,
-      "learning_rate": 9.145839133797372e-06,
-      "loss": 2.0049,
+      "epoch": 0.29,
+      "learning_rate": 2.163895168166966e-05,
+      "loss": 2.4865,
       "step": 1555
     },
     {
-      "epoch": 3.62,
-      "learning_rate": 9.101484918793504e-06,
-      "loss": 1.9514,
+      "epoch": 0.29,
+      "learning_rate": 2.1627187545711374e-05,
+      "loss": 2.2172,
       "step": 1560
     },
     {
-      "epoch": 3.63,
-      "learning_rate": 9.057130703789636e-06,
-      "loss": 1.9647,
+      "epoch": 0.29,
+      "learning_rate": 2.161542340975309e-05,
+      "loss": 2.4172,
       "step": 1565
     },
     {
-      "epoch": 3.64,
-      "learning_rate": 9.01277648878577e-06,
-      "loss": 1.9638,
+      "epoch": 0.29,
+      "learning_rate": 2.1603659273794804e-05,
+      "loss": 2.7129,
       "step": 1570
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 8.968422273781903e-06,
-      "loss": 2.0392,
+      "epoch": 0.29,
+      "learning_rate": 2.1591895137836518e-05,
+      "loss": 2.3931,
       "step": 1575
     },
     {
-      "epoch": 3.67,
-      "learning_rate": 8.924068058778036e-06,
-      "loss": 2.0167,
+      "epoch": 0.29,
+      "learning_rate": 2.1580131001878236e-05,
+      "loss": 2.2633,
       "step": 1580
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 8.879713843774169e-06,
-      "loss": 2.0097,
+      "epoch": 0.29,
+      "learning_rate": 2.1568366865919948e-05,
+      "loss": 2.3415,
       "step": 1585
     },
     {
-      "epoch": 3.69,
-      "learning_rate": 8.835359628770302e-06,
-      "loss": 1.9624,
+      "epoch": 0.3,
+      "learning_rate": 2.1556602729961662e-05,
+      "loss": 2.3735,
       "step": 1590
     },
     {
-      "epoch": 3.7,
-      "learning_rate": 8.791005413766435e-06,
-      "loss": 2.0064,
+      "epoch": 0.3,
+      "learning_rate": 2.154483859400338e-05,
+      "loss": 2.4076,
       "step": 1595
     },
     {
-      "epoch": 3.71,
-      "learning_rate": 8.746651198762568e-06,
-      "loss": 1.9917,
-      "step": 1600
-    },
-    {
-      "epoch": 3.71,
-      "eval_accuracy": 0.5947941530837255,
-      "eval_loss": 2.238585948944092,
-      "eval_runtime": 58.2344,
-      "eval_samples_per_second": 7.384,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.3,
+      "learning_rate": 2.1533074458045092e-05,
+      "loss": 2.4192,
       "step": 1600
     },
     {
-      "epoch": 3.72,
-      "learning_rate": 8.7022969837587e-06,
-      "loss": 2.0178,
+      "epoch": 0.3,
+      "learning_rate": 2.152131032208681e-05,
+      "loss": 2.9549,
       "step": 1605
     },
     {
-      "epoch": 3.74,
-      "learning_rate": 8.657942768754835e-06,
-      "loss": 1.995,
+      "epoch": 0.3,
+      "learning_rate": 2.1509546186128525e-05,
+      "loss": 2.5723,
       "step": 1610
     },
     {
-      "epoch": 3.75,
-      "learning_rate": 8.613588553750966e-06,
-      "loss": 2.0199,
+      "epoch": 0.3,
+      "learning_rate": 2.1497782050170236e-05,
+      "loss": 2.4359,
       "step": 1615
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 8.5692343387471e-06,
-      "loss": 2.0151,
+      "epoch": 0.3,
+      "learning_rate": 2.1486017914211954e-05,
+      "loss": 2.3857,
       "step": 1620
     },
     {
-      "epoch": 3.77,
-      "learning_rate": 8.524880123743234e-06,
-      "loss": 1.9532,
+      "epoch": 0.3,
+      "learning_rate": 2.147425377825367e-05,
+      "loss": 2.2842,
       "step": 1625
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 8.480525908739365e-06,
-      "loss": 1.9242,
+      "epoch": 0.3,
+      "learning_rate": 2.1462489642295384e-05,
+      "loss": 2.4698,
       "step": 1630
     },
     {
-      "epoch": 3.79,
-      "learning_rate": 8.4361716937355e-06,
-      "loss": 1.9664,
+      "epoch": 0.3,
+      "learning_rate": 2.1450725506337098e-05,
+      "loss": 2.2655,
       "step": 1635
     },
     {
-      "epoch": 3.81,
-      "learning_rate": 8.391817478731632e-06,
-      "loss": 1.9545,
+      "epoch": 0.31,
+      "learning_rate": 2.1438961370378813e-05,
+      "loss": 2.6379,
       "step": 1640
     },
     {
-      "epoch": 3.82,
-      "learning_rate": 8.347463263727765e-06,
-      "loss": 1.9732,
+      "epoch": 0.31,
+      "learning_rate": 2.1427197234420528e-05,
+      "loss": 2.3113,
       "step": 1645
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 8.303109048723898e-06,
-      "loss": 1.9884,
+      "epoch": 0.31,
+      "learning_rate": 2.1415433098462242e-05,
+      "loss": 2.3197,
       "step": 1650
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 8.258754833720031e-06,
-      "loss": 2.0224,
+      "epoch": 0.31,
+      "learning_rate": 2.1403668962503957e-05,
+      "loss": 2.2366,
       "step": 1655
     },
     {
-      "epoch": 3.85,
-      "learning_rate": 8.214400618716164e-06,
-      "loss": 1.9914,
+      "epoch": 0.31,
+      "learning_rate": 2.1391904826545672e-05,
+      "loss": 2.6301,
       "step": 1660
     },
     {
-      "epoch": 3.86,
-      "learning_rate": 8.170046403712297e-06,
-      "loss": 1.9627,
+      "epoch": 0.31,
+      "learning_rate": 2.1380140690587387e-05,
+      "loss": 2.4991,
       "step": 1665
     },
     {
-      "epoch": 3.87,
-      "learning_rate": 8.12569218870843e-06,
-      "loss": 1.977,
+      "epoch": 0.31,
+      "learning_rate": 2.13683765546291e-05,
+      "loss": 2.2349,
       "step": 1670
     },
     {
-      "epoch": 3.89,
-      "learning_rate": 8.081337973704564e-06,
-      "loss": 2.0249,
+      "epoch": 0.31,
+      "learning_rate": 2.1356612418670816e-05,
+      "loss": 2.1911,
       "step": 1675
     },
     {
-      "epoch": 3.9,
-      "learning_rate": 8.036983758700697e-06,
-      "loss": 2.067,
+      "epoch": 0.31,
+      "learning_rate": 2.134484828271253e-05,
+      "loss": 2.6309,
       "step": 1680
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 7.992629543696828e-06,
-      "loss": 1.9708,
+      "epoch": 0.31,
+      "learning_rate": 2.133308414675425e-05,
+      "loss": 2.3671,
       "step": 1685
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 7.948275328692963e-06,
-      "loss": 1.9988,
+      "epoch": 0.31,
+      "learning_rate": 2.132132001079596e-05,
+      "loss": 2.2787,
       "step": 1690
     },
     {
-      "epoch": 3.93,
-      "learning_rate": 7.903921113689096e-06,
-      "loss": 2.0803,
+      "epoch": 0.32,
+      "learning_rate": 2.1309555874837675e-05,
+      "loss": 2.5662,
       "step": 1695
     },
     {
-      "epoch": 3.94,
-      "learning_rate": 7.859566898685229e-06,
-      "loss": 2.0152,
-      "step": 1700
-    },
-    {
-      "epoch": 3.94,
-      "eval_accuracy": 0.596026279297097,
-      "eval_loss": 2.2315049171447754,
-      "eval_runtime": 58.2112,
-      "eval_samples_per_second": 7.387,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.32,
+      "learning_rate": 2.1297791738879393e-05,
+      "loss": 2.0166,
       "step": 1700
     },
     {
-      "epoch": 3.96,
-      "learning_rate": 7.815212683681361e-06,
-      "loss": 1.9705,
+      "epoch": 0.32,
+      "learning_rate": 2.1286027602921104e-05,
+      "loss": 1.9664,
       "step": 1705
     },
     {
-      "epoch": 3.97,
-      "learning_rate": 7.770858468677494e-06,
-      "loss": 1.9686,
+      "epoch": 0.32,
+      "learning_rate": 2.1274263466962822e-05,
+      "loss": 2.4688,
       "step": 1710
     },
     {
-      "epoch": 3.98,
-      "learning_rate": 7.726504253673627e-06,
-      "loss": 1.9508,
+      "epoch": 0.32,
+      "learning_rate": 2.1262499331004537e-05,
+      "loss": 2.5438,
       "step": 1715
     },
     {
-      "epoch": 3.99,
-      "learning_rate": 7.68215003866976e-06,
-      "loss": 1.9775,
+      "epoch": 0.32,
+      "learning_rate": 2.125073519504625e-05,
+      "loss": 2.3246,
       "step": 1720
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 7.637795823665893e-06,
-      "loss": 1.9657,
+      "epoch": 0.32,
+      "learning_rate": 2.1238971059087966e-05,
+      "loss": 2.3416,
       "step": 1725
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 7.593441608662027e-06,
-      "loss": 1.8765,
+      "epoch": 0.32,
+      "learning_rate": 2.122720692312968e-05,
+      "loss": 2.4248,
       "step": 1730
     },
     {
-      "epoch": 4.03,
-      "learning_rate": 7.549087393658159e-06,
-      "loss": 1.9449,
+      "epoch": 0.32,
+      "learning_rate": 2.1215442787171396e-05,
+      "loss": 2.0812,
       "step": 1735
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 7.5047331786542925e-06,
-      "loss": 1.9377,
+      "epoch": 0.32,
+      "learning_rate": 2.120367865121311e-05,
+      "loss": 2.336,
       "step": 1740
     },
     {
-      "epoch": 4.05,
-      "learning_rate": 7.460378963650426e-06,
-      "loss": 1.8993,
+      "epoch": 0.32,
+      "learning_rate": 2.1191914515254825e-05,
+      "loss": 2.5387,
       "step": 1745
     },
     {
-      "epoch": 4.06,
-      "learning_rate": 7.416024748646558e-06,
-      "loss": 1.9302,
+      "epoch": 0.33,
+      "learning_rate": 2.118015037929654e-05,
+      "loss": 2.3395,
       "step": 1750
     },
     {
-      "epoch": 4.07,
-      "learning_rate": 7.371670533642691e-06,
-      "loss": 1.9457,
+      "epoch": 0.33,
+      "learning_rate": 2.1168386243338255e-05,
+      "loss": 2.1717,
       "step": 1755
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 7.327316318638825e-06,
-      "loss": 1.9081,
+      "epoch": 0.33,
+      "learning_rate": 2.115662210737997e-05,
+      "loss": 2.3396,
       "step": 1760
     },
     {
-      "epoch": 4.1,
-      "learning_rate": 7.282962103634959e-06,
-      "loss": 1.906,
+      "epoch": 0.33,
+      "learning_rate": 2.1144857971421684e-05,
+      "loss": 2.3539,
       "step": 1765
     },
     {
-      "epoch": 4.11,
-      "learning_rate": 7.238607888631091e-06,
-      "loss": 1.9133,
+      "epoch": 0.33,
+      "learning_rate": 2.11330938354634e-05,
+      "loss": 2.2103,
       "step": 1770
     },
     {
-      "epoch": 4.12,
-      "learning_rate": 7.1942536736272235e-06,
-      "loss": 1.9495,
+      "epoch": 0.33,
+      "learning_rate": 2.1121329699505117e-05,
+      "loss": 2.471,
       "step": 1775
     },
     {
-      "epoch": 4.13,
-      "learning_rate": 7.149899458623357e-06,
-      "loss": 1.8968,
+      "epoch": 0.33,
+      "learning_rate": 2.110956556354683e-05,
+      "loss": 2.4741,
       "step": 1780
     },
     {
-      "epoch": 4.14,
-      "learning_rate": 7.105545243619489e-06,
-      "loss": 1.8823,
+      "epoch": 0.33,
+      "learning_rate": 2.1097801427588543e-05,
+      "loss": 2.0704,
       "step": 1785
     },
     {
-      "epoch": 4.15,
-      "learning_rate": 7.061191028615623e-06,
-      "loss": 1.9294,
+      "epoch": 0.33,
+      "learning_rate": 2.108603729163026e-05,
+      "loss": 2.3032,
       "step": 1790
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 7.016836813611756e-06,
-      "loss": 1.9581,
+      "epoch": 0.33,
+      "learning_rate": 2.1074273155671973e-05,
+      "loss": 2.1046,
       "step": 1795
     },
     {
-      "epoch": 4.18,
-      "learning_rate": 6.972482598607888e-06,
-      "loss": 1.886,
-      "step": 1800
-    },
-    {
-      "epoch": 4.18,
-      "eval_accuracy": 0.5957148378003592,
-      "eval_loss": 2.241971969604492,
-      "eval_runtime": 58.1995,
-      "eval_samples_per_second": 7.388,
-      "eval_steps_per_second": 1.856,
+      "epoch": 0.33,
+      "learning_rate": 2.106250901971369e-05,
+      "loss": 2.4872,
       "step": 1800
     },
     {
-      "epoch": 4.19,
-      "learning_rate": 6.928128383604022e-06,
-      "loss": 1.9006,
+      "epoch": 0.34,
+      "learning_rate": 2.1050744883755405e-05,
+      "loss": 2.4918,
       "step": 1805
     },
     {
-      "epoch": 4.2,
-      "learning_rate": 6.883774168600155e-06,
-      "loss": 1.9951,
+      "epoch": 0.34,
+      "learning_rate": 2.1038980747797117e-05,
+      "loss": 2.0083,
       "step": 1810
     },
     {
-      "epoch": 4.21,
-      "learning_rate": 6.839419953596288e-06,
-      "loss": 1.8455,
+      "epoch": 0.34,
+      "learning_rate": 2.1027216611838835e-05,
+      "loss": 2.3806,
       "step": 1815
     },
     {
-      "epoch": 4.22,
-      "learning_rate": 6.79506573859242e-06,
-      "loss": 1.9753,
+      "epoch": 0.34,
+      "learning_rate": 2.101545247588055e-05,
+      "loss": 2.3691,
       "step": 1820
     },
     {
-      "epoch": 4.23,
-      "learning_rate": 6.750711523588554e-06,
-      "loss": 1.9538,
+      "epoch": 0.34,
+      "learning_rate": 2.1003688339922264e-05,
+      "loss": 2.0162,
       "step": 1825
     },
     {
-      "epoch": 4.25,
-      "learning_rate": 6.706357308584687e-06,
-      "loss": 1.9943,
+      "epoch": 0.34,
+      "learning_rate": 2.099192420396398e-05,
+      "loss": 2.2573,
       "step": 1830
     },
     {
-      "epoch": 4.26,
-      "learning_rate": 6.66200309358082e-06,
-      "loss": 1.873,
+      "epoch": 0.34,
+      "learning_rate": 2.0980160068005694e-05,
+      "loss": 2.44,
       "step": 1835
     },
     {
-      "epoch": 4.27,
-      "learning_rate": 6.617648878576953e-06,
-      "loss": 1.9086,
+      "epoch": 0.34,
+      "learning_rate": 2.096839593204741e-05,
+      "loss": 2.2721,
       "step": 1840
     },
     {
-      "epoch": 4.28,
-      "learning_rate": 6.5732946635730864e-06,
-      "loss": 1.9429,
+      "epoch": 0.34,
+      "learning_rate": 2.0956631796089123e-05,
+      "loss": 2.1641,
       "step": 1845
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 6.528940448569219e-06,
-      "loss": 1.9145,
+      "epoch": 0.34,
+      "learning_rate": 2.0944867660130838e-05,
+      "loss": 2.4872,
       "step": 1850
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 6.484586233565351e-06,
-      "loss": 1.9733,
+      "epoch": 0.35,
+      "learning_rate": 2.0933103524172553e-05,
+      "loss": 2.254,
       "step": 1855
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 6.440232018561485e-06,
-      "loss": 1.9154,
+      "epoch": 0.35,
+      "learning_rate": 2.0921339388214267e-05,
+      "loss": 2.2815,
       "step": 1860
     },
     {
-      "epoch": 4.33,
-      "learning_rate": 6.395877803557619e-06,
-      "loss": 1.9516,
+      "epoch": 0.35,
+      "learning_rate": 2.0909575252255982e-05,
+      "loss": 2.1729,
       "step": 1865
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 6.351523588553751e-06,
-      "loss": 1.841,
+      "epoch": 0.35,
+      "learning_rate": 2.0897811116297697e-05,
+      "loss": 2.0859,
       "step": 1870
     },
     {
-      "epoch": 4.35,
-      "learning_rate": 6.307169373549884e-06,
-      "loss": 1.964,
+      "epoch": 0.35,
+      "learning_rate": 2.088604698033941e-05,
+      "loss": 2.276,
       "step": 1875
     },
     {
-      "epoch": 4.36,
-      "learning_rate": 6.2628151585460175e-06,
-      "loss": 1.897,
+      "epoch": 0.35,
+      "learning_rate": 2.087428284438113e-05,
+      "loss": 2.6566,
       "step": 1880
     },
     {
-      "epoch": 4.37,
-      "learning_rate": 6.2184609435421495e-06,
-      "loss": 1.9151,
+      "epoch": 0.35,
+      "learning_rate": 2.086251870842284e-05,
+      "loss": 2.344,
       "step": 1885
     },
     {
-      "epoch": 4.39,
-      "learning_rate": 6.174106728538283e-06,
-      "loss": 1.901,
+      "epoch": 0.35,
+      "learning_rate": 2.0850754572464556e-05,
+      "loss": 2.2243,
       "step": 1890
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 6.129752513534416e-06,
-      "loss": 1.9238,
+      "epoch": 0.35,
+      "learning_rate": 2.0838990436506274e-05,
+      "loss": 2.5325,
       "step": 1895
     },
     {
-      "epoch": 4.41,
-      "learning_rate": 6.08539829853055e-06,
-      "loss": 1.9151,
-      "step": 1900
-    },
-    {
-      "epoch": 4.41,
-      "eval_accuracy": 0.5966559821773625,
-      "eval_loss": 2.2408835887908936,
-      "eval_runtime": 58.2356,
-      "eval_samples_per_second": 7.384,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.35,
+      "learning_rate": 2.0827226300547985e-05,
+      "loss": 2.1057,
       "step": 1900
     },
     {
-      "epoch": 4.42,
-      "learning_rate": 6.041044083526682e-06,
-      "loss": 1.8906,
+      "epoch": 0.35,
+      "learning_rate": 2.0815462164589703e-05,
+      "loss": 2.4445,
       "step": 1905
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 5.996689868522816e-06,
-      "loss": 1.8945,
+      "epoch": 0.36,
+      "learning_rate": 2.0803698028631414e-05,
+      "loss": 2.1062,
       "step": 1910
     },
     {
-      "epoch": 4.44,
-      "learning_rate": 5.9523356535189485e-06,
-      "loss": 1.8905,
+      "epoch": 0.36,
+      "learning_rate": 2.079193389267313e-05,
+      "loss": 2.1631,
       "step": 1915
     },
     {
-      "epoch": 4.45,
-      "learning_rate": 5.9079814385150805e-06,
-      "loss": 1.9209,
+      "epoch": 0.36,
+      "learning_rate": 2.0780169756714847e-05,
+      "loss": 2.4469,
       "step": 1920
     },
     {
-      "epoch": 4.47,
-      "learning_rate": 5.863627223511214e-06,
-      "loss": 1.8885,
+      "epoch": 0.36,
+      "learning_rate": 2.076840562075656e-05,
+      "loss": 2.4799,
       "step": 1925
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 5.819273008507348e-06,
-      "loss": 1.8773,
+      "epoch": 0.36,
+      "learning_rate": 2.0756641484798277e-05,
+      "loss": 2.4734,
       "step": 1930
     },
     {
-      "epoch": 4.49,
-      "learning_rate": 5.774918793503481e-06,
-      "loss": 1.9252,
+      "epoch": 0.36,
+      "learning_rate": 2.074487734883999e-05,
+      "loss": 2.3981,
       "step": 1935
     },
     {
-      "epoch": 4.5,
-      "learning_rate": 5.730564578499614e-06,
-      "loss": 1.9655,
+      "epoch": 0.36,
+      "learning_rate": 2.0733113212881703e-05,
+      "loss": 2.1789,
       "step": 1940
     },
     {
-      "epoch": 4.51,
-      "learning_rate": 5.686210363495747e-06,
-      "loss": 1.9207,
+      "epoch": 0.36,
+      "learning_rate": 2.072134907692342e-05,
+      "loss": 2.4502,
       "step": 1945
     },
     {
-      "epoch": 4.52,
-      "learning_rate": 5.6418561484918795e-06,
-      "loss": 1.9147,
+      "epoch": 0.36,
+      "learning_rate": 2.0709584940965135e-05,
+      "loss": 1.9887,
       "step": 1950
     },
     {
-      "epoch": 4.54,
-      "learning_rate": 5.5975019334880124e-06,
-      "loss": 1.8857,
+      "epoch": 0.36,
+      "learning_rate": 2.069782080500685e-05,
+      "loss": 2.5355,
       "step": 1955
     },
     {
-      "epoch": 4.55,
-      "learning_rate": 5.553147718484145e-06,
-      "loss": 1.959,
+      "epoch": 0.36,
+      "learning_rate": 2.0686056669048565e-05,
+      "loss": 2.0708,
       "step": 1960
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 5.508793503480279e-06,
-      "loss": 1.8938,
+      "epoch": 0.37,
+      "learning_rate": 2.067429253309028e-05,
+      "loss": 2.2796,
       "step": 1965
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 5.464439288476412e-06,
-      "loss": 1.9368,
+      "epoch": 0.37,
+      "learning_rate": 2.0662528397131994e-05,
+      "loss": 2.4555,
       "step": 1970
     },
     {
-      "epoch": 4.58,
-      "learning_rate": 5.420085073472544e-06,
-      "loss": 1.9247,
+      "epoch": 0.37,
+      "learning_rate": 2.065076426117371e-05,
+      "loss": 1.98,
       "step": 1975
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 5.375730858468678e-06,
-      "loss": 1.9154,
+      "epoch": 0.37,
+      "learning_rate": 2.0639000125215424e-05,
+      "loss": 2.285,
       "step": 1980
     },
     {
-      "epoch": 4.61,
-      "learning_rate": 5.331376643464811e-06,
-      "loss": 1.8864,
+      "epoch": 0.37,
+      "learning_rate": 2.062723598925714e-05,
+      "loss": 2.4769,
       "step": 1985
     },
     {
-      "epoch": 4.62,
-      "learning_rate": 5.2870224284609435e-06,
-      "loss": 1.9635,
+      "epoch": 0.37,
+      "learning_rate": 2.0615471853298853e-05,
+      "loss": 2.4454,
       "step": 1990
     },
     {
-      "epoch": 4.63,
-      "learning_rate": 5.242668213457076e-06,
-      "loss": 1.891,
+      "epoch": 0.37,
+      "learning_rate": 2.0603707717340568e-05,
+      "loss": 2.2847,
       "step": 1995
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 5.19831399845321e-06,
-      "loss": 1.9538,
+      "epoch": 0.37,
+      "learning_rate": 2.0591943581382283e-05,
+      "loss": 2.7019,
       "step": 2000
     },
     {
-      "epoch": 4.64,
-      "eval_accuracy": 0.5971379208438473,
-      "eval_loss": 2.2378673553466797,
-      "eval_runtime": 58.2169,
-      "eval_samples_per_second": 7.386,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.37,
+      "eval_accuracy": 0.845649854021947,
+      "eval_loss": 2.3207740783691406,
+      "eval_runtime": 129.6225,
+      "eval_samples_per_second": 82.933,
+      "eval_steps_per_second": 10.369,
       "step": 2000
     },
     {
-      "epoch": 4.65,
-      "learning_rate": 5.153959783449343e-06,
-      "loss": 1.9781,
+      "epoch": 0.37,
+      "learning_rate": 2.0580179445423997e-05,
+      "loss": 2.2302,
       "step": 2005
     },
     {
-      "epoch": 4.66,
-      "learning_rate": 5.109605568445476e-06,
-      "loss": 1.9347,
+      "epoch": 0.37,
+      "learning_rate": 2.0568415309465715e-05,
+      "loss": 2.3256,
       "step": 2010
     },
     {
-      "epoch": 4.68,
-      "learning_rate": 5.065251353441609e-06,
-      "loss": 1.9551,
+      "epoch": 0.37,
+      "learning_rate": 2.0556651173507427e-05,
+      "loss": 2.243,
       "step": 2015
     },
     {
-      "epoch": 4.69,
-      "learning_rate": 5.020897138437742e-06,
-      "loss": 1.8674,
+      "epoch": 0.38,
+      "learning_rate": 2.0544887037549145e-05,
+      "loss": 2.4583,
       "step": 2020
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 4.976542923433875e-06,
-      "loss": 1.9526,
+      "epoch": 0.38,
+      "learning_rate": 2.053312290159086e-05,
+      "loss": 2.405,
       "step": 2025
     },
     {
-      "epoch": 4.71,
-      "learning_rate": 4.932188708430008e-06,
-      "loss": 1.9281,
+      "epoch": 0.38,
+      "learning_rate": 2.052135876563257e-05,
+      "loss": 2.4788,
       "step": 2030
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 4.88783449342614e-06,
-      "loss": 1.909,
+      "epoch": 0.38,
+      "learning_rate": 2.050959462967429e-05,
+      "loss": 2.4352,
       "step": 2035
     },
     {
-      "epoch": 4.73,
-      "learning_rate": 4.843480278422274e-06,
-      "loss": 1.9404,
+      "epoch": 0.38,
+      "learning_rate": 2.0497830493716004e-05,
+      "loss": 2.5007,
       "step": 2040
     },
     {
-      "epoch": 4.74,
-      "learning_rate": 4.799126063418407e-06,
-      "loss": 1.9413,
+      "epoch": 0.38,
+      "learning_rate": 2.048606635775772e-05,
+      "loss": 2.5601,
       "step": 2045
     },
     {
-      "epoch": 4.76,
-      "learning_rate": 4.75477184841454e-06,
-      "loss": 1.9653,
+      "epoch": 0.38,
+      "learning_rate": 2.0474302221799433e-05,
+      "loss": 2.3736,
       "step": 2050
     },
     {
-      "epoch": 4.77,
-      "learning_rate": 4.710417633410673e-06,
-      "loss": 1.931,
+      "epoch": 0.38,
+      "learning_rate": 2.0462538085841148e-05,
+      "loss": 2.2944,
       "step": 2055
     },
     {
-      "epoch": 4.78,
-      "learning_rate": 4.6660634184068055e-06,
-      "loss": 1.9733,
+      "epoch": 0.38,
+      "learning_rate": 2.0450773949882863e-05,
+      "loss": 2.746,
       "step": 2060
     },
     {
-      "epoch": 4.79,
-      "learning_rate": 4.621709203402939e-06,
-      "loss": 1.8921,
+      "epoch": 0.38,
+      "learning_rate": 2.0439009813924577e-05,
+      "loss": 2.4902,
       "step": 2065
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 4.577354988399072e-06,
-      "loss": 1.9374,
+      "epoch": 0.39,
+      "learning_rate": 2.0427245677966292e-05,
+      "loss": 2.7318,
       "step": 2070
     },
     {
-      "epoch": 4.81,
-      "learning_rate": 4.533000773395205e-06,
-      "loss": 1.9668,
+      "epoch": 0.39,
+      "learning_rate": 2.0415481542008007e-05,
+      "loss": 2.3585,
       "step": 2075
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 4.488646558391338e-06,
-      "loss": 1.8869,
+      "epoch": 0.39,
+      "learning_rate": 2.040371740604972e-05,
+      "loss": 2.5383,
       "step": 2080
     },
     {
-      "epoch": 4.84,
-      "learning_rate": 4.444292343387472e-06,
-      "loss": 1.9688,
+      "epoch": 0.39,
+      "learning_rate": 2.0391953270091436e-05,
+      "loss": 2.1141,
       "step": 2085
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 4.3999381283836045e-06,
-      "loss": 1.9266,
+      "epoch": 0.39,
+      "learning_rate": 2.038018913413315e-05,
+      "loss": 2.4491,
       "step": 2090
     },
     {
-      "epoch": 4.86,
-      "learning_rate": 4.3555839133797366e-06,
-      "loss": 1.907,
+      "epoch": 0.39,
+      "learning_rate": 2.0368424998174866e-05,
+      "loss": 2.4057,
       "step": 2095
     },
     {
-      "epoch": 4.87,
-      "learning_rate": 4.31122969837587e-06,
-      "loss": 1.8886,
-      "step": 2100
-    },
-    {
-      "epoch": 4.87,
-      "eval_accuracy": 0.5975880333719794,
-      "eval_loss": 2.2348711490631104,
-      "eval_runtime": 58.2184,
-      "eval_samples_per_second": 7.386,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.39,
+      "learning_rate": 2.0356660862216584e-05,
+      "loss": 2.1615,
       "step": 2100
     },
     {
-      "epoch": 4.88,
-      "learning_rate": 4.266875483372003e-06,
-      "loss": 1.9057,
+      "epoch": 0.39,
+      "learning_rate": 2.0344896726258295e-05,
+      "loss": 2.1614,
       "step": 2105
     },
     {
-      "epoch": 4.9,
-      "learning_rate": 4.222521268368136e-06,
-      "loss": 1.9208,
+      "epoch": 0.39,
+      "learning_rate": 2.033313259030001e-05,
+      "loss": 2.0801,
       "step": 2110
     },
     {
-      "epoch": 4.91,
-      "learning_rate": 4.178167053364269e-06,
-      "loss": 1.9621,
+      "epoch": 0.39,
+      "learning_rate": 2.0321368454341728e-05,
+      "loss": 2.3235,
       "step": 2115
     },
     {
-      "epoch": 4.92,
-      "learning_rate": 4.133812838360402e-06,
-      "loss": 1.9827,
+      "epoch": 0.39,
+      "learning_rate": 2.030960431838344e-05,
+      "loss": 2.4257,
       "step": 2120
     },
     {
-      "epoch": 4.93,
-      "learning_rate": 4.0894586233565356e-06,
-      "loss": 1.925,
+      "epoch": 0.4,
+      "learning_rate": 2.0297840182425157e-05,
+      "loss": 2.2655,
       "step": 2125
     },
     {
-      "epoch": 4.94,
-      "learning_rate": 4.0451044083526684e-06,
-      "loss": 1.958,
+      "epoch": 0.4,
+      "learning_rate": 2.0286076046466872e-05,
+      "loss": 2.1867,
       "step": 2130
     },
     {
-      "epoch": 4.95,
-      "learning_rate": 4.000750193348801e-06,
-      "loss": 1.9232,
+      "epoch": 0.4,
+      "learning_rate": 2.0274311910508583e-05,
+      "loss": 2.3719,
       "step": 2135
     },
     {
-      "epoch": 4.97,
-      "learning_rate": 3.956395978344934e-06,
-      "loss": 1.8868,
+      "epoch": 0.4,
+      "learning_rate": 2.02625477745503e-05,
+      "loss": 2.3344,
       "step": 2140
     },
     {
-      "epoch": 4.98,
-      "learning_rate": 3.912041763341068e-06,
-      "loss": 1.8728,
+      "epoch": 0.4,
+      "learning_rate": 2.0250783638592016e-05,
+      "loss": 2.3679,
       "step": 2145
     },
     {
-      "epoch": 4.99,
-      "learning_rate": 3.867687548337201e-06,
-      "loss": 1.9234,
+      "epoch": 0.4,
+      "learning_rate": 2.023901950263373e-05,
+      "loss": 2.3108,
       "step": 2150
     },
     {
-      "epoch": 5.0,
-      "learning_rate": 3.823333333333333e-06,
-      "loss": 1.9097,
+      "epoch": 0.4,
+      "learning_rate": 2.0227255366675446e-05,
+      "loss": 2.5112,
       "step": 2155
     },
     {
-      "epoch": 5.01,
-      "learning_rate": 3.7789791183294666e-06,
-      "loss": 1.9308,
+      "epoch": 0.4,
+      "learning_rate": 2.021549123071716e-05,
+      "loss": 2.4933,
       "step": 2160
     },
     {
-      "epoch": 5.02,
-      "learning_rate": 3.7346249033255995e-06,
-      "loss": 1.9104,
+      "epoch": 0.4,
+      "learning_rate": 2.0203727094758875e-05,
+      "loss": 2.4929,
       "step": 2165
     },
     {
-      "epoch": 5.03,
-      "learning_rate": 3.690270688321733e-06,
-      "loss": 1.8072,
+      "epoch": 0.4,
+      "learning_rate": 2.019196295880059e-05,
+      "loss": 2.3178,
       "step": 2170
     },
     {
-      "epoch": 5.05,
-      "learning_rate": 3.6459164733178657e-06,
-      "loss": 1.9154,
+      "epoch": 0.4,
+      "learning_rate": 2.0180198822842304e-05,
+      "loss": 2.339,
       "step": 2175
     },
     {
-      "epoch": 5.06,
-      "learning_rate": 3.601562258313998e-06,
-      "loss": 1.8595,
+      "epoch": 0.41,
+      "learning_rate": 2.016843468688402e-05,
+      "loss": 2.1715,
       "step": 2180
     },
     {
-      "epoch": 5.07,
-      "learning_rate": 3.557208043310132e-06,
-      "loss": 1.8349,
+      "epoch": 0.41,
+      "learning_rate": 2.0156670550925734e-05,
+      "loss": 2.3412,
       "step": 2185
     },
     {
-      "epoch": 5.08,
-      "learning_rate": 3.5128538283062643e-06,
-      "loss": 1.8689,
+      "epoch": 0.41,
+      "learning_rate": 2.014490641496745e-05,
+      "loss": 2.4255,
       "step": 2190
     },
     {
-      "epoch": 5.09,
-      "learning_rate": 3.4684996133023976e-06,
-      "loss": 1.8304,
+      "epoch": 0.41,
+      "learning_rate": 2.0133142279009163e-05,
+      "loss": 2.4197,
       "step": 2195
     },
     {
-      "epoch": 5.1,
-      "learning_rate": 3.4241453982985305e-06,
-      "loss": 1.9408,
-      "step": 2200
-    },
-    {
-      "epoch": 5.1,
-      "eval_accuracy": 0.5974516356361818,
-      "eval_loss": 2.2409677505493164,
-      "eval_runtime": 58.2193,
-      "eval_samples_per_second": 7.386,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.41,
+      "learning_rate": 2.0121378143050878e-05,
+      "loss": 2.1734,
       "step": 2200
     },
     {
-      "epoch": 5.12,
-      "learning_rate": 3.3797911832946634e-06,
-      "loss": 1.8922,
+      "epoch": 0.41,
+      "learning_rate": 2.0109614007092596e-05,
+      "loss": 2.4111,
       "step": 2205
     },
     {
-      "epoch": 5.13,
-      "learning_rate": 3.3354369682907967e-06,
-      "loss": 1.8482,
+      "epoch": 0.41,
+      "learning_rate": 2.0097849871134307e-05,
+      "loss": 2.2507,
       "step": 2210
     },
     {
-      "epoch": 5.14,
-      "learning_rate": 3.2910827532869296e-06,
-      "loss": 1.9108,
+      "epoch": 0.41,
+      "learning_rate": 2.0086085735176022e-05,
+      "loss": 2.0631,
       "step": 2215
     },
     {
-      "epoch": 5.15,
-      "learning_rate": 3.246728538283063e-06,
-      "loss": 1.8998,
+      "epoch": 0.41,
+      "learning_rate": 2.007432159921774e-05,
+      "loss": 2.4796,
       "step": 2220
     },
     {
-      "epoch": 5.16,
-      "learning_rate": 3.2023743232791958e-06,
-      "loss": 1.8573,
+      "epoch": 0.41,
+      "learning_rate": 2.006255746325945e-05,
+      "loss": 2.2022,
       "step": 2225
     },
     {
-      "epoch": 5.17,
-      "learning_rate": 3.158020108275329e-06,
-      "loss": 1.8576,
+      "epoch": 0.41,
+      "learning_rate": 2.005079332730117e-05,
+      "loss": 2.2455,
       "step": 2230
     },
     {
-      "epoch": 5.19,
-      "learning_rate": 3.113665893271462e-06,
-      "loss": 1.9237,
+      "epoch": 0.42,
+      "learning_rate": 2.0039029191342884e-05,
+      "loss": 2.6112,
       "step": 2235
     },
     {
-      "epoch": 5.2,
-      "learning_rate": 3.0693116782675944e-06,
-      "loss": 1.8795,
+      "epoch": 0.42,
+      "learning_rate": 2.0027265055384596e-05,
+      "loss": 2.3769,
       "step": 2240
     },
     {
-      "epoch": 5.21,
-      "learning_rate": 3.024957463263728e-06,
-      "loss": 1.8451,
+      "epoch": 0.42,
+      "learning_rate": 2.0015500919426314e-05,
+      "loss": 2.3387,
       "step": 2245
     },
     {
-      "epoch": 5.22,
-      "learning_rate": 2.9806032482598606e-06,
-      "loss": 1.8141,
+      "epoch": 0.42,
+      "learning_rate": 2.000373678346803e-05,
+      "loss": 2.2623,
       "step": 2250
     },
     {
-      "epoch": 5.23,
-      "learning_rate": 2.936249033255994e-06,
-      "loss": 1.8714,
+      "epoch": 0.42,
+      "learning_rate": 1.9991972647509743e-05,
+      "loss": 2.2732,
       "step": 2255
     },
     {
-      "epoch": 5.24,
-      "learning_rate": 2.891894818252127e-06,
-      "loss": 1.8939,
+      "epoch": 0.42,
+      "learning_rate": 1.9980208511551458e-05,
+      "loss": 2.022,
       "step": 2260
     },
     {
-      "epoch": 5.26,
-      "learning_rate": 2.84754060324826e-06,
-      "loss": 1.8502,
+      "epoch": 0.42,
+      "learning_rate": 1.996844437559317e-05,
+      "loss": 2.1274,
       "step": 2265
     },
     {
-      "epoch": 5.27,
-      "learning_rate": 2.803186388244393e-06,
-      "loss": 1.8796,
+      "epoch": 0.42,
+      "learning_rate": 1.9956680239634887e-05,
+      "loss": 2.2668,
       "step": 2270
     },
     {
-      "epoch": 5.28,
-      "learning_rate": 2.7588321732405263e-06,
-      "loss": 1.884,
+      "epoch": 0.42,
+      "learning_rate": 1.9944916103676602e-05,
+      "loss": 2.1882,
       "step": 2275
     },
     {
-      "epoch": 5.29,
-      "learning_rate": 2.7144779582366588e-06,
-      "loss": 1.9446,
+      "epoch": 0.42,
+      "learning_rate": 1.9933151967718317e-05,
+      "loss": 2.1965,
       "step": 2280
     },
     {
-      "epoch": 5.3,
-      "learning_rate": 2.670123743232792e-06,
-      "loss": 1.912,
+      "epoch": 0.43,
+      "learning_rate": 1.992138783176003e-05,
+      "loss": 2.7555,
       "step": 2285
     },
     {
-      "epoch": 5.31,
-      "learning_rate": 2.625769528228925e-06,
-      "loss": 1.9372,
+      "epoch": 0.43,
+      "learning_rate": 1.9909623695801746e-05,
+      "loss": 2.1214,
       "step": 2290
     },
     {
-      "epoch": 5.32,
-      "learning_rate": 2.5814153132250583e-06,
-      "loss": 1.873,
+      "epoch": 0.43,
+      "learning_rate": 1.989785955984346e-05,
+      "loss": 2.4507,
       "step": 2295
     },
     {
-      "epoch": 5.34,
-      "learning_rate": 2.537061098221191e-06,
-      "loss": 1.9168,
-      "step": 2300
-    },
-    {
-      "epoch": 5.34,
-      "eval_accuracy": 0.5975857600763828,
-      "eval_loss": 2.239448070526123,
-      "eval_runtime": 58.2108,
-      "eval_samples_per_second": 7.387,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.43,
+      "learning_rate": 1.9886095423885176e-05,
+      "loss": 2.747,
       "step": 2300
     },
     {
-      "epoch": 5.35,
-      "learning_rate": 2.492706883217324e-06,
-      "loss": 1.8444,
+      "epoch": 0.43,
+      "learning_rate": 1.987433128792689e-05,
+      "loss": 2.5539,
       "step": 2305
     },
     {
-      "epoch": 5.36,
-      "learning_rate": 2.448352668213457e-06,
-      "loss": 1.9179,
+      "epoch": 0.43,
+      "learning_rate": 1.9862567151968605e-05,
+      "loss": 2.249,
       "step": 2310
     },
     {
-      "epoch": 5.37,
-      "learning_rate": 2.4039984532095902e-06,
-      "loss": 1.8802,
+      "epoch": 0.43,
+      "learning_rate": 1.985080301601032e-05,
+      "loss": 2.3754,
       "step": 2315
     },
     {
-      "epoch": 5.38,
-      "learning_rate": 2.359644238205723e-06,
-      "loss": 1.8798,
+      "epoch": 0.43,
+      "learning_rate": 1.9839038880052038e-05,
+      "loss": 1.9462,
       "step": 2320
     },
     {
-      "epoch": 5.39,
-      "learning_rate": 2.3152900232018564e-06,
-      "loss": 1.922,
+      "epoch": 0.43,
+      "learning_rate": 1.982727474409375e-05,
+      "loss": 2.777,
       "step": 2325
     },
     {
-      "epoch": 5.41,
-      "learning_rate": 2.2709358081979893e-06,
-      "loss": 1.8982,
+      "epoch": 0.43,
+      "learning_rate": 1.9815510608135464e-05,
+      "loss": 2.6442,
       "step": 2330
     },
     {
-      "epoch": 5.42,
-      "learning_rate": 2.226581593194122e-06,
-      "loss": 1.853,
+      "epoch": 0.43,
+      "learning_rate": 1.9803746472177182e-05,
+      "loss": 2.3325,
       "step": 2335
     },
     {
-      "epoch": 5.43,
-      "learning_rate": 2.182227378190255e-06,
-      "loss": 1.8815,
+      "epoch": 0.44,
+      "learning_rate": 1.9791982336218894e-05,
+      "loss": 2.4372,
       "step": 2340
     },
     {
-      "epoch": 5.44,
-      "learning_rate": 2.1378731631863884e-06,
-      "loss": 1.8678,
+      "epoch": 0.44,
+      "learning_rate": 1.978021820026061e-05,
+      "loss": 2.4413,
       "step": 2345
     },
     {
-      "epoch": 5.45,
-      "learning_rate": 2.0935189481825213e-06,
-      "loss": 1.7827,
+      "epoch": 0.44,
+      "learning_rate": 1.9768454064302326e-05,
+      "loss": 2.2182,
       "step": 2350
     },
     {
-      "epoch": 5.46,
-      "learning_rate": 2.0491647331786546e-06,
-      "loss": 1.797,
+      "epoch": 0.44,
+      "learning_rate": 1.9756689928344038e-05,
+      "loss": 2.157,
       "step": 2355
     },
     {
-      "epoch": 5.48,
-      "learning_rate": 2.0048105181747875e-06,
-      "loss": 1.8769,
+      "epoch": 0.44,
+      "learning_rate": 1.9744925792385756e-05,
+      "loss": 2.1475,
       "step": 2360
     },
     {
-      "epoch": 5.49,
-      "learning_rate": 1.9604563031709203e-06,
-      "loss": 1.8924,
+      "epoch": 0.44,
+      "learning_rate": 1.973316165642747e-05,
+      "loss": 2.1303,
       "step": 2365
     },
     {
-      "epoch": 5.5,
-      "learning_rate": 1.9161020881670532e-06,
-      "loss": 1.9016,
+      "epoch": 0.44,
+      "learning_rate": 1.9721397520469185e-05,
+      "loss": 2.3693,
       "step": 2370
     },
     {
-      "epoch": 5.51,
-      "learning_rate": 1.8717478731631863e-06,
-      "loss": 1.838,
+      "epoch": 0.44,
+      "learning_rate": 1.97096333845109e-05,
+      "loss": 2.2634,
       "step": 2375
     },
     {
-      "epoch": 5.52,
-      "learning_rate": 1.8273936581593194e-06,
-      "loss": 1.8716,
+      "epoch": 0.44,
+      "learning_rate": 1.9697869248552615e-05,
+      "loss": 2.4722,
       "step": 2380
     },
     {
-      "epoch": 5.53,
-      "learning_rate": 1.7830394431554525e-06,
-      "loss": 1.8949,
+      "epoch": 0.44,
+      "learning_rate": 1.968610511259433e-05,
+      "loss": 2.0935,
       "step": 2385
     },
     {
-      "epoch": 5.55,
-      "learning_rate": 1.7386852281515856e-06,
-      "loss": 1.9195,
+      "epoch": 0.44,
+      "learning_rate": 1.9674340976636044e-05,
+      "loss": 2.546,
       "step": 2390
     },
     {
-      "epoch": 5.56,
-      "learning_rate": 1.6943310131477185e-06,
-      "loss": 1.917,
+      "epoch": 0.45,
+      "learning_rate": 1.966257684067776e-05,
+      "loss": 2.3,
       "step": 2395
     },
     {
-      "epoch": 5.57,
-      "learning_rate": 1.6499767981438514e-06,
-      "loss": 1.8002,
+      "epoch": 0.45,
+      "learning_rate": 1.9650812704719473e-05,
+      "loss": 2.6158,
       "step": 2400
     },
     {
-      "epoch": 5.57,
-      "eval_accuracy": 0.5977403441769533,
-      "eval_loss": 2.238140106201172,
-      "eval_runtime": 58.2064,
-      "eval_samples_per_second": 7.388,
-      "eval_steps_per_second": 1.855,
-      "step": 2400
+      "epoch": 0.45,
+      "learning_rate": 1.9639048568761188e-05,
+      "loss": 2.2332,
+      "step": 2405
     },
     {
-      "epoch": 5.58,
-      "learning_rate": 1.6056225831399845e-06,
-      "loss": 1.8851,
-      "step": 2405
-    },
-    {
-      "epoch": 5.59,
-      "learning_rate": 1.5612683681361176e-06,
-      "loss": 1.9006,
+      "epoch": 0.45,
+      "learning_rate": 1.9627284432802903e-05,
+      "loss": 2.6344,
       "step": 2410
     },
     {
-      "epoch": 5.6,
-      "learning_rate": 1.5169141531322507e-06,
-      "loss": 1.863,
+      "epoch": 0.45,
+      "learning_rate": 1.9615520296844618e-05,
+      "loss": 2.4277,
       "step": 2415
     },
     {
-      "epoch": 5.61,
-      "learning_rate": 1.4725599381283838e-06,
-      "loss": 1.9131,
+      "epoch": 0.45,
+      "learning_rate": 1.9603756160886332e-05,
+      "loss": 2.2472,
       "step": 2420
     },
     {
-      "epoch": 5.63,
-      "learning_rate": 1.4282057231245166e-06,
-      "loss": 1.8538,
+      "epoch": 0.45,
+      "learning_rate": 1.959199202492805e-05,
+      "loss": 2.6158,
       "step": 2425
     },
     {
-      "epoch": 5.64,
-      "learning_rate": 1.3838515081206495e-06,
-      "loss": 1.8868,
+      "epoch": 0.45,
+      "learning_rate": 1.9580227888969762e-05,
+      "loss": 2.5754,
       "step": 2430
     },
     {
-      "epoch": 5.65,
-      "learning_rate": 1.3394972931167826e-06,
-      "loss": 1.8671,
+      "epoch": 0.45,
+      "learning_rate": 1.9568463753011476e-05,
+      "loss": 2.3813,
       "step": 2435
     },
     {
-      "epoch": 5.66,
-      "learning_rate": 1.2951430781129157e-06,
-      "loss": 1.922,
+      "epoch": 0.45,
+      "learning_rate": 1.9556699617053195e-05,
+      "loss": 2.2544,
       "step": 2440
     },
     {
-      "epoch": 5.67,
-      "learning_rate": 1.2507888631090486e-06,
-      "loss": 1.816,
+      "epoch": 0.45,
+      "learning_rate": 1.9544935481094906e-05,
+      "loss": 2.2983,
       "step": 2445
     },
     {
-      "epoch": 5.68,
-      "learning_rate": 1.2064346481051817e-06,
-      "loss": 1.8375,
+      "epoch": 0.46,
+      "learning_rate": 1.9533171345136624e-05,
+      "loss": 2.3814,
       "step": 2450
     },
     {
-      "epoch": 5.7,
-      "learning_rate": 1.1620804331013148e-06,
-      "loss": 1.8747,
+      "epoch": 0.46,
+      "learning_rate": 1.952140720917834e-05,
+      "loss": 1.9954,
       "step": 2455
     },
     {
-      "epoch": 5.71,
-      "learning_rate": 1.1177262180974477e-06,
-      "loss": 1.8866,
+      "epoch": 0.46,
+      "learning_rate": 1.950964307322005e-05,
+      "loss": 2.1805,
       "step": 2460
     },
     {
-      "epoch": 5.72,
-      "learning_rate": 1.0733720030935808e-06,
-      "loss": 1.8335,
+      "epoch": 0.46,
+      "learning_rate": 1.9497878937261768e-05,
+      "loss": 2.852,
       "step": 2465
     },
     {
-      "epoch": 5.73,
-      "learning_rate": 1.0290177880897139e-06,
-      "loss": 1.8518,
+      "epoch": 0.46,
+      "learning_rate": 1.9486114801303483e-05,
+      "loss": 2.0869,
       "step": 2470
     },
     {
-      "epoch": 5.74,
-      "learning_rate": 9.846635730858468e-07,
-      "loss": 1.957,
+      "epoch": 0.46,
+      "learning_rate": 1.9474350665345198e-05,
+      "loss": 2.2437,
       "step": 2475
     },
     {
-      "epoch": 5.75,
-      "learning_rate": 9.4030935808198e-07,
-      "loss": 1.9421,
+      "epoch": 0.46,
+      "learning_rate": 1.9462586529386912e-05,
+      "loss": 2.3579,
       "step": 2480
     },
     {
-      "epoch": 5.77,
-      "learning_rate": 8.959551430781131e-07,
-      "loss": 1.8889,
+      "epoch": 0.46,
+      "learning_rate": 1.9450822393428627e-05,
+      "loss": 2.025,
       "step": 2485
     },
     {
-      "epoch": 5.78,
-      "learning_rate": 8.516009280742459e-07,
-      "loss": 1.8684,
+      "epoch": 0.46,
+      "learning_rate": 1.9439058257470342e-05,
+      "loss": 2.23,
       "step": 2490
     },
     {
-      "epoch": 5.79,
-      "learning_rate": 8.07246713070379e-07,
-      "loss": 1.8269,
+      "epoch": 0.46,
+      "learning_rate": 1.9427294121512056e-05,
+      "loss": 2.5128,
       "step": 2495
     },
     {
-      "epoch": 5.8,
-      "learning_rate": 7.628924980665121e-07,
-      "loss": 1.8888,
-      "step": 2500
-    },
-    {
-      "epoch": 5.8,
-      "eval_accuracy": 0.5977812634976926,
-      "eval_loss": 2.2366693019866943,
-      "eval_runtime": 58.2106,
-      "eval_samples_per_second": 7.387,
-      "eval_steps_per_second": 1.855,
+      "epoch": 0.47,
+      "learning_rate": 1.941552998555377e-05,
+      "loss": 2.2414,
       "step": 2500
     },
     {
-      "epoch": 5.81,
-      "learning_rate": 7.18538283062645e-07,
-      "loss": 1.8787,
+      "epoch": 0.47,
+      "learning_rate": 1.9403765849595486e-05,
+      "loss": 2.2478,
       "step": 2505
     },
     {
-      "epoch": 5.82,
-      "learning_rate": 6.741840680587781e-07,
-      "loss": 1.8871,
+      "epoch": 0.47,
+      "learning_rate": 1.93920017136372e-05,
+      "loss": 2.1582,
       "step": 2510
     },
     {
-      "epoch": 5.84,
-      "learning_rate": 6.298298530549111e-07,
-      "loss": 1.9249,
+      "epoch": 0.47,
+      "learning_rate": 1.9380237577678915e-05,
+      "loss": 2.1914,
       "step": 2515
     },
     {
-      "epoch": 5.85,
-      "learning_rate": 5.854756380510441e-07,
-      "loss": 1.8683,
+      "epoch": 0.47,
+      "learning_rate": 1.936847344172063e-05,
+      "loss": 2.3862,
       "step": 2520
     },
     {
-      "epoch": 5.86,
-      "learning_rate": 5.411214230471772e-07,
-      "loss": 1.8578,
+      "epoch": 0.47,
+      "learning_rate": 1.9356709305762345e-05,
+      "loss": 2.0638,
       "step": 2525
     },
     {
-      "epoch": 5.87,
-      "learning_rate": 4.967672080433102e-07,
-      "loss": 1.9025,
+      "epoch": 0.47,
+      "learning_rate": 1.9344945169804063e-05,
+      "loss": 2.3358,
       "step": 2530
     },
     {
-      "epoch": 5.88,
-      "learning_rate": 4.5241299303944316e-07,
-      "loss": 1.9059,
+      "epoch": 0.47,
+      "learning_rate": 1.9333181033845774e-05,
+      "loss": 2.28,
       "step": 2535
     },
     {
-      "epoch": 5.89,
-      "learning_rate": 4.080587780355762e-07,
-      "loss": 1.8879,
+      "epoch": 0.47,
+      "learning_rate": 1.932141689788749e-05,
+      "loss": 2.2137,
       "step": 2540
     },
     {
-      "epoch": 5.9,
-      "learning_rate": 3.6370456303170925e-07,
-      "loss": 1.8826,
+      "epoch": 0.47,
+      "learning_rate": 1.9309652761929207e-05,
+      "loss": 2.3894,
       "step": 2545
     },
     {
-      "epoch": 5.92,
-      "learning_rate": 3.1935034802784224e-07,
-      "loss": 1.8687,
+      "epoch": 0.47,
+      "learning_rate": 1.929788862597092e-05,
+      "loss": 2.4621,
       "step": 2550
     },
     {
-      "epoch": 5.93,
-      "learning_rate": 2.749961330239753e-07,
-      "loss": 1.8749,
+      "epoch": 0.48,
+      "learning_rate": 1.9286124490012636e-05,
+      "loss": 2.3493,
       "step": 2555
     },
     {
-      "epoch": 5.94,
-      "learning_rate": 2.306419180201083e-07,
-      "loss": 1.8967,
+      "epoch": 0.48,
+      "learning_rate": 1.927436035405435e-05,
+      "loss": 2.0996,
       "step": 2560
     },
     {
-      "epoch": 5.95,
-      "learning_rate": 1.862877030162413e-07,
-      "loss": 1.8327,
+      "epoch": 0.48,
+      "learning_rate": 1.9262596218096062e-05,
+      "loss": 2.5021,
       "step": 2565
     },
     {
-      "epoch": 5.96,
-      "learning_rate": 1.4193348801237433e-07,
-      "loss": 1.8679,
+      "epoch": 0.48,
+      "learning_rate": 1.925083208213778e-05,
+      "loss": 2.535,
       "step": 2570
     },
     {
-      "epoch": 5.97,
-      "learning_rate": 9.757927300850735e-08,
-      "loss": 1.9444,
+      "epoch": 0.48,
+      "learning_rate": 1.9239067946179495e-05,
+      "loss": 2.0711,
       "step": 2575
     },
     {
-      "epoch": 5.99,
-      "learning_rate": 5.322505800464037e-08,
-      "loss": 1.821,
+      "epoch": 0.48,
+      "learning_rate": 1.922730381022121e-05,
+      "loss": 2.2445,
       "step": 2580
     },
     {
-      "epoch": 6.0,
-      "learning_rate": 8.870843000773396e-09,
-      "loss": 1.9073,
+      "epoch": 0.48,
+      "learning_rate": 1.9215539674262925e-05,
+      "loss": 2.2462,
       "step": 2585
     },
     {
-      "epoch": 6.0,
-      "step": 2586,
-      "total_flos": 4.49946006847488e+16,
-      "train_loss": 2.103886123784088,
-      "train_runtime": 6222.0683,
-      "train_samples_per_second": 1.662,
-      "train_steps_per_second": 0.416
+      "epoch": 0.48,
+      "learning_rate": 1.920377553830464e-05,
+      "loss": 2.3344,
+      "step": 2590
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9192011402346354e-05,
+      "loss": 2.2154,
+      "step": 2595
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.918024726638807e-05,
+      "loss": 2.187,
+      "step": 2600
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9168483130429784e-05,
+      "loss": 2.3569,
+      "step": 2605
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.91567189944715e-05,
+      "loss": 2.3438,
+      "step": 2610
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9144954858513213e-05,
+      "loss": 2.2933,
+      "step": 2615
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9133190722554928e-05,
+      "loss": 2.0546,
+      "step": 2620
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9121426586596642e-05,
+      "loss": 2.1615,
+      "step": 2625
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9109662450638357e-05,
+      "loss": 2.1971,
+      "step": 2630
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9097898314680072e-05,
+      "loss": 2.2412,
+      "step": 2635
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9086134178721787e-05,
+      "loss": 2.4374,
+      "step": 2640
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9074370042763505e-05,
+      "loss": 2.2498,
+      "step": 2645
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9062605906805216e-05,
+      "loss": 2.1534,
+      "step": 2650
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.905084177084693e-05,
+      "loss": 2.371,
+      "step": 2655
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.903907763488865e-05,
+      "loss": 2.3063,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.902731349893036e-05,
+      "loss": 2.1823,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.901554936297208e-05,
+      "loss": 2.5433,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9003785227013793e-05,
+      "loss": 2.4324,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8992021091055504e-05,
+      "loss": 2.3933,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8980256955097222e-05,
+      "loss": 2.2444,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8968492819138937e-05,
+      "loss": 2.6491,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8956728683180652e-05,
+      "loss": 2.3012,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8944964547222367e-05,
+      "loss": 2.4084,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.893320041126408e-05,
+      "loss": 2.2014,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8921436275305796e-05,
+      "loss": 2.2808,
+      "step": 2710
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.890967213934751e-05,
+      "loss": 1.9891,
+      "step": 2715
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8897908003389225e-05,
+      "loss": 2.4147,
+      "step": 2720
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.888614386743094e-05,
+      "loss": 2.6616,
+      "step": 2725
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8874379731472655e-05,
+      "loss": 2.3897,
+      "step": 2730
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.886261559551437e-05,
+      "loss": 2.3019,
+      "step": 2735
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8850851459556084e-05,
+      "loss": 2.5422,
+      "step": 2740
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.88390873235978e-05,
+      "loss": 2.3852,
+      "step": 2745
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8827323187639517e-05,
+      "loss": 2.4082,
+      "step": 2750
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.881555905168123e-05,
+      "loss": 2.2455,
+      "step": 2755
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8803794915722943e-05,
+      "loss": 2.6003,
+      "step": 2760
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.879203077976466e-05,
+      "loss": 2.5991,
+      "step": 2765
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8780266643806373e-05,
+      "loss": 2.4768,
+      "step": 2770
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.876850250784809e-05,
+      "loss": 2.0744,
+      "step": 2775
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8756738371889805e-05,
+      "loss": 2.537,
+      "step": 2780
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8744974235931517e-05,
+      "loss": 2.4932,
+      "step": 2785
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8733210099973235e-05,
+      "loss": 2.519,
+      "step": 2790
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.872144596401495e-05,
+      "loss": 2.3585,
+      "step": 2795
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8709681828056664e-05,
+      "loss": 2.2269,
+      "step": 2800
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.869791769209838e-05,
+      "loss": 2.2978,
+      "step": 2805
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8686153556140094e-05,
+      "loss": 2.6597,
+      "step": 2810
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.867438942018181e-05,
+      "loss": 2.3325,
+      "step": 2815
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8662625284223523e-05,
+      "loss": 2.1164,
+      "step": 2820
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8650861148265238e-05,
+      "loss": 2.665,
+      "step": 2825
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8639097012306953e-05,
+      "loss": 2.3498,
+      "step": 2830
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8627332876348667e-05,
+      "loss": 2.5662,
+      "step": 2835
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8615568740390382e-05,
+      "loss": 2.1709,
+      "step": 2840
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8603804604432097e-05,
+      "loss": 2.2885,
+      "step": 2845
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.859204046847381e-05,
+      "loss": 2.1814,
+      "step": 2850
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.858027633251553e-05,
+      "loss": 2.4746,
+      "step": 2855
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.856851219655724e-05,
+      "loss": 2.433,
+      "step": 2860
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8556748060598956e-05,
+      "loss": 2.5462,
+      "step": 2865
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8544983924640674e-05,
+      "loss": 2.1056,
+      "step": 2870
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8533219788682385e-05,
+      "loss": 2.5366,
+      "step": 2875
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8521455652724103e-05,
+      "loss": 2.3272,
+      "step": 2880
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8509691516765818e-05,
+      "loss": 2.1142,
+      "step": 2885
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.849792738080753e-05,
+      "loss": 1.7842,
+      "step": 2890
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8486163244849247e-05,
+      "loss": 2.2089,
+      "step": 2895
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8474399108890962e-05,
+      "loss": 2.2893,
+      "step": 2900
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8462634972932677e-05,
+      "loss": 2.285,
+      "step": 2905
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.845087083697439e-05,
+      "loss": 2.1269,
+      "step": 2910
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8439106701016106e-05,
+      "loss": 2.5882,
+      "step": 2915
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.842734256505782e-05,
+      "loss": 2.6215,
+      "step": 2920
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8415578429099536e-05,
+      "loss": 2.5817,
+      "step": 2925
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.840381429314125e-05,
+      "loss": 2.2154,
+      "step": 2930
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8392050157182965e-05,
+      "loss": 2.3242,
+      "step": 2935
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.838028602122468e-05,
+      "loss": 1.8786,
+      "step": 2940
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8368521885266398e-05,
+      "loss": 2.2957,
+      "step": 2945
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.835675774930811e-05,
+      "loss": 2.5603,
+      "step": 2950
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8344993613349824e-05,
+      "loss": 2.4587,
+      "step": 2955
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8333229477391542e-05,
+      "loss": 1.9703,
+      "step": 2960
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8321465341433253e-05,
+      "loss": 2.2506,
+      "step": 2965
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.830970120547497e-05,
+      "loss": 2.2645,
+      "step": 2970
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8297937069516683e-05,
+      "loss": 2.2901,
+      "step": 2975
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8286172933558397e-05,
+      "loss": 2.113,
+      "step": 2980
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8274408797600116e-05,
+      "loss": 2.1493,
+      "step": 2985
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8262644661641827e-05,
+      "loss": 2.0246,
+      "step": 2990
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8250880525683545e-05,
+      "loss": 2.1576,
+      "step": 2995
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.823911638972526e-05,
+      "loss": 2.5278,
+      "step": 3000
+    },
+    {
+      "epoch": 0.56,
+      "eval_accuracy": 0.8463976643511527,
+      "eval_loss": 2.2469661235809326,
+      "eval_runtime": 129.4731,
+      "eval_samples_per_second": 83.029,
+      "eval_steps_per_second": 10.381,
+      "step": 3000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.822735225376697e-05,
+      "loss": 2.2716,
+      "step": 3005
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.821558811780869e-05,
+      "loss": 2.3098,
+      "step": 3010
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8203823981850404e-05,
+      "loss": 1.9717,
+      "step": 3015
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.819205984589212e-05,
+      "loss": 2.3056,
+      "step": 3020
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8180295709933833e-05,
+      "loss": 2.2244,
+      "step": 3025
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8168531573975548e-05,
+      "loss": 2.3269,
+      "step": 3030
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8156767438017263e-05,
+      "loss": 2.2463,
+      "step": 3035
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8145003302058977e-05,
+      "loss": 2.3383,
+      "step": 3040
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8133239166100692e-05,
+      "loss": 2.1296,
+      "step": 3045
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8121475030142407e-05,
+      "loss": 2.2593,
+      "step": 3050
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.810971089418412e-05,
+      "loss": 2.3584,
+      "step": 3055
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8097946758225836e-05,
+      "loss": 2.0471,
+      "step": 3060
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.808618262226755e-05,
+      "loss": 2.3271,
+      "step": 3065
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8074418486309266e-05,
+      "loss": 1.9052,
+      "step": 3070
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8062654350350984e-05,
+      "loss": 2.371,
+      "step": 3075
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8050890214392695e-05,
+      "loss": 2.5769,
+      "step": 3080
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.803912607843441e-05,
+      "loss": 2.128,
+      "step": 3085
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8027361942476128e-05,
+      "loss": 2.3772,
+      "step": 3090
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.801559780651784e-05,
+      "loss": 2.3192,
+      "step": 3095
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8003833670559557e-05,
+      "loss": 2.0984,
+      "step": 3100
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7992069534601272e-05,
+      "loss": 2.4784,
+      "step": 3105
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7980305398642983e-05,
+      "loss": 1.9928,
+      "step": 3110
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.79685412626847e-05,
+      "loss": 2.647,
+      "step": 3115
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7956777126726416e-05,
+      "loss": 2.1446,
+      "step": 3120
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.794501299076813e-05,
+      "loss": 2.2854,
+      "step": 3125
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7933248854809846e-05,
+      "loss": 2.1393,
+      "step": 3130
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.792148471885156e-05,
+      "loss": 2.0322,
+      "step": 3135
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7909720582893275e-05,
+      "loss": 2.4705,
+      "step": 3140
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.789795644693499e-05,
+      "loss": 2.189,
+      "step": 3145
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7886192310976705e-05,
+      "loss": 2.2703,
+      "step": 3150
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.787442817501842e-05,
+      "loss": 2.1293,
+      "step": 3155
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7862664039060134e-05,
+      "loss": 2.6898,
+      "step": 3160
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7850899903101852e-05,
+      "loss": 2.3431,
+      "step": 3165
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7839135767143563e-05,
+      "loss": 2.6859,
+      "step": 3170
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7827371631185278e-05,
+      "loss": 2.0361,
+      "step": 3175
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7815607495226996e-05,
+      "loss": 2.2326,
+      "step": 3180
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7803843359268708e-05,
+      "loss": 2.278,
+      "step": 3185
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7792079223310426e-05,
+      "loss": 2.6171,
+      "step": 3190
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.778031508735214e-05,
+      "loss": 2.323,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7768550951393852e-05,
+      "loss": 2.3452,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.775678681543557e-05,
+      "loss": 2.2284,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7745022679477285e-05,
+      "loss": 2.3707,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7733258543519e-05,
+      "loss": 2.209,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7721494407560714e-05,
+      "loss": 2.6077,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.770973027160243e-05,
+      "loss": 2.3008,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7697966135644143e-05,
+      "loss": 2.0932,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7686201999685858e-05,
+      "loss": 1.9965,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7674437863727573e-05,
+      "loss": 2.214,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7662673727769288e-05,
+      "loss": 2.3303,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7650909591811002e-05,
+      "loss": 2.1357,
+      "step": 3250
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7639145455852717e-05,
+      "loss": 2.6586,
+      "step": 3255
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7627381319894432e-05,
+      "loss": 2.1398,
+      "step": 3260
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7615617183936146e-05,
+      "loss": 2.2141,
+      "step": 3265
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7603853047977865e-05,
+      "loss": 2.4011,
+      "step": 3270
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7592088912019576e-05,
+      "loss": 2.1108,
+      "step": 3275
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.758032477606129e-05,
+      "loss": 1.9823,
+      "step": 3280
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.756856064010301e-05,
+      "loss": 2.3562,
+      "step": 3285
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.755679650414472e-05,
+      "loss": 2.202,
+      "step": 3290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7545032368186438e-05,
+      "loss": 2.3704,
+      "step": 3295
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7533268232228153e-05,
+      "loss": 2.2441,
+      "step": 3300
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7521504096269864e-05,
+      "loss": 2.183,
+      "step": 3305
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7509739960311582e-05,
+      "loss": 2.1835,
+      "step": 3310
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7497975824353297e-05,
+      "loss": 2.0659,
+      "step": 3315
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7486211688395012e-05,
+      "loss": 2.5222,
+      "step": 3320
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7474447552436726e-05,
+      "loss": 2.4026,
+      "step": 3325
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.746268341647844e-05,
+      "loss": 2.4423,
+      "step": 3330
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7450919280520156e-05,
+      "loss": 2.332,
+      "step": 3335
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.743915514456187e-05,
+      "loss": 2.3242,
+      "step": 3340
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7427391008603585e-05,
+      "loss": 2.4874,
+      "step": 3345
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.74156268726453e-05,
+      "loss": 2.1182,
+      "step": 3350
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7403862736687015e-05,
+      "loss": 2.3661,
+      "step": 3355
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.739209860072873e-05,
+      "loss": 1.9324,
+      "step": 3360
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7380334464770444e-05,
+      "loss": 2.2784,
+      "step": 3365
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.736857032881216e-05,
+      "loss": 2.4882,
+      "step": 3370
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7356806192853874e-05,
+      "loss": 2.5061,
+      "step": 3375
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.734504205689559e-05,
+      "loss": 1.9751,
+      "step": 3380
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7333277920937303e-05,
+      "loss": 1.9313,
+      "step": 3385
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7321513784979018e-05,
+      "loss": 2.3103,
+      "step": 3390
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7309749649020732e-05,
+      "loss": 1.7791,
+      "step": 3395
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.729798551306245e-05,
+      "loss": 2.0509,
+      "step": 3400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7286221377104162e-05,
+      "loss": 2.1514,
+      "step": 3405
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7274457241145877e-05,
+      "loss": 2.1919,
+      "step": 3410
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7262693105187595e-05,
+      "loss": 2.2812,
+      "step": 3415
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7250928969229306e-05,
+      "loss": 2.3251,
+      "step": 3420
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7239164833271024e-05,
+      "loss": 1.9922,
+      "step": 3425
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.722740069731274e-05,
+      "loss": 2.3279,
+      "step": 3430
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.721563656135445e-05,
+      "loss": 2.2647,
+      "step": 3435
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7203872425396168e-05,
+      "loss": 2.3735,
+      "step": 3440
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7192108289437883e-05,
+      "loss": 2.1398,
+      "step": 3445
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7180344153479598e-05,
+      "loss": 2.2288,
+      "step": 3450
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7168580017521312e-05,
+      "loss": 2.0255,
+      "step": 3455
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7156815881563027e-05,
+      "loss": 2.3582,
+      "step": 3460
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7145051745604742e-05,
+      "loss": 2.2322,
+      "step": 3465
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7133287609646457e-05,
+      "loss": 2.2703,
+      "step": 3470
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.712152347368817e-05,
+      "loss": 2.0195,
+      "step": 3475
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7109759337729886e-05,
+      "loss": 2.4042,
+      "step": 3480
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.70979952017716e-05,
+      "loss": 2.4916,
+      "step": 3485
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.708623106581332e-05,
+      "loss": 2.3128,
+      "step": 3490
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.707446692985503e-05,
+      "loss": 2.3604,
+      "step": 3495
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7062702793896745e-05,
+      "loss": 2.1941,
+      "step": 3500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7050938657938463e-05,
+      "loss": 2.0597,
+      "step": 3505
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7039174521980174e-05,
+      "loss": 2.309,
+      "step": 3510
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7027410386021892e-05,
+      "loss": 2.5635,
+      "step": 3515
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7015646250063607e-05,
+      "loss": 2.2153,
+      "step": 3520
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.700388211410532e-05,
+      "loss": 2.2763,
+      "step": 3525
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.6992117978147037e-05,
+      "loss": 2.5317,
+      "step": 3530
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.698035384218875e-05,
+      "loss": 2.1175,
+      "step": 3535
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.6968589706230466e-05,
+      "loss": 2.3361,
+      "step": 3540
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.695682557027218e-05,
+      "loss": 2.2494,
+      "step": 3545
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.6945061434313895e-05,
+      "loss": 2.388,
+      "step": 3550
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.693329729835561e-05,
+      "loss": 2.4341,
+      "step": 3555
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.6921533162397325e-05,
+      "loss": 2.0897,
+      "step": 3560
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.690976902643904e-05,
+      "loss": 2.4297,
+      "step": 3565
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.6898004890480754e-05,
+      "loss": 2.0843,
+      "step": 3570
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.688624075452247e-05,
+      "loss": 2.3291,
+      "step": 3575
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6874476618564184e-05,
+      "loss": 2.1834,
+      "step": 3580
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.68627124826059e-05,
+      "loss": 2.2828,
+      "step": 3585
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6850948346647613e-05,
+      "loss": 2.2881,
+      "step": 3590
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.683918421068933e-05,
+      "loss": 2.3394,
+      "step": 3595
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6827420074731043e-05,
+      "loss": 2.2082,
+      "step": 3600
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6815655938772757e-05,
+      "loss": 2.1282,
+      "step": 3605
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6803891802814475e-05,
+      "loss": 2.3299,
+      "step": 3610
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6792127666856187e-05,
+      "loss": 2.1723,
+      "step": 3615
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6780363530897905e-05,
+      "loss": 2.4329,
+      "step": 3620
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.676859939493962e-05,
+      "loss": 2.0853,
+      "step": 3625
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.675683525898133e-05,
+      "loss": 2.3464,
+      "step": 3630
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.674507112302305e-05,
+      "loss": 2.1622,
+      "step": 3635
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6733306987064764e-05,
+      "loss": 2.3685,
+      "step": 3640
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.672154285110648e-05,
+      "loss": 2.3802,
+      "step": 3645
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6709778715148193e-05,
+      "loss": 2.5158,
+      "step": 3650
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6698014579189908e-05,
+      "loss": 2.2212,
+      "step": 3655
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6686250443231623e-05,
+      "loss": 2.2426,
+      "step": 3660
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6674486307273337e-05,
+      "loss": 2.2132,
+      "step": 3665
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6662722171315052e-05,
+      "loss": 2.0926,
+      "step": 3670
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6650958035356767e-05,
+      "loss": 2.201,
+      "step": 3675
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.663919389939848e-05,
+      "loss": 2.3594,
+      "step": 3680
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6627429763440196e-05,
+      "loss": 2.2803,
+      "step": 3685
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.661566562748191e-05,
+      "loss": 2.5357,
+      "step": 3690
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6603901491523626e-05,
+      "loss": 2.2192,
+      "step": 3695
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.659213735556534e-05,
+      "loss": 2.2938,
+      "step": 3700
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6580373219607055e-05,
+      "loss": 2.0777,
+      "step": 3705
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.656860908364877e-05,
+      "loss": 2.4382,
+      "step": 3710
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6556844947690484e-05,
+      "loss": 2.104,
+      "step": 3715
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.65450808117322e-05,
+      "loss": 2.1332,
+      "step": 3720
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6533316675773917e-05,
+      "loss": 2.6409,
+      "step": 3725
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.652155253981563e-05,
+      "loss": 2.2173,
+      "step": 3730
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6509788403857343e-05,
+      "loss": 2.3015,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.649802426789906e-05,
+      "loss": 2.3413,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6486260131940773e-05,
+      "loss": 2.1272,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.647449599598249e-05,
+      "loss": 2.4673,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6462731860024206e-05,
+      "loss": 2.2785,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6450967724065917e-05,
+      "loss": 2.2714,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6439203588107635e-05,
+      "loss": 2.3506,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.642743945214935e-05,
+      "loss": 2.1574,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6415675316191064e-05,
+      "loss": 2.419,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.640391118023278e-05,
+      "loss": 2.299,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6392147044274494e-05,
+      "loss": 2.101,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.638038290831621e-05,
+      "loss": 2.6912,
+      "step": 3790
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6368618772357923e-05,
+      "loss": 2.2014,
+      "step": 3795
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6356854636399638e-05,
+      "loss": 2.3532,
+      "step": 3800
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6345090500441353e-05,
+      "loss": 2.0276,
+      "step": 3805
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6333326364483067e-05,
+      "loss": 2.2448,
+      "step": 3810
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6321562228524786e-05,
+      "loss": 2.0738,
+      "step": 3815
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6309798092566497e-05,
+      "loss": 1.9878,
+      "step": 3820
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.629803395660821e-05,
+      "loss": 2.067,
+      "step": 3825
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.628626982064993e-05,
+      "loss": 2.3584,
+      "step": 3830
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.627450568469164e-05,
+      "loss": 1.9388,
+      "step": 3835
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.626274154873336e-05,
+      "loss": 2.211,
+      "step": 3840
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6250977412775074e-05,
+      "loss": 2.2312,
+      "step": 3845
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6239213276816785e-05,
+      "loss": 2.1374,
+      "step": 3850
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6227449140858503e-05,
+      "loss": 2.3065,
+      "step": 3855
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6215685004900218e-05,
+      "loss": 2.3097,
+      "step": 3860
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6203920868941933e-05,
+      "loss": 2.2419,
+      "step": 3865
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6192156732983647e-05,
+      "loss": 2.3355,
+      "step": 3870
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6180392597025362e-05,
+      "loss": 2.2204,
+      "step": 3875
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6168628461067077e-05,
+      "loss": 2.1331,
+      "step": 3880
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.615686432510879e-05,
+      "loss": 2.4482,
+      "step": 3885
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6145100189150506e-05,
+      "loss": 2.4005,
+      "step": 3890
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.613333605319222e-05,
+      "loss": 2.0223,
+      "step": 3895
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6121571917233936e-05,
+      "loss": 2.2544,
+      "step": 3900
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.610980778127565e-05,
+      "loss": 2.3227,
+      "step": 3905
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6098043645317365e-05,
+      "loss": 2.1151,
+      "step": 3910
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.608627950935908e-05,
+      "loss": 2.0597,
+      "step": 3915
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6074515373400798e-05,
+      "loss": 2.4949,
+      "step": 3920
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.606275123744251e-05,
+      "loss": 2.0159,
+      "step": 3925
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6050987101484224e-05,
+      "loss": 2.2341,
+      "step": 3930
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6039222965525942e-05,
+      "loss": 2.1146,
+      "step": 3935
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6027458829567653e-05,
+      "loss": 1.7166,
+      "step": 3940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.601569469360937e-05,
+      "loss": 1.8895,
+      "step": 3945
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6003930557651086e-05,
+      "loss": 2.0268,
+      "step": 3950
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5992166421692798e-05,
+      "loss": 2.4976,
+      "step": 3955
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5980402285734516e-05,
+      "loss": 2.237,
+      "step": 3960
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.596863814977623e-05,
+      "loss": 2.293,
+      "step": 3965
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5956874013817945e-05,
+      "loss": 1.9507,
+      "step": 3970
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.594510987785966e-05,
+      "loss": 2.2886,
+      "step": 3975
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5933345741901375e-05,
+      "loss": 2.2152,
+      "step": 3980
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.592158160594309e-05,
+      "loss": 2.5009,
+      "step": 3985
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5909817469984804e-05,
+      "loss": 2.4498,
+      "step": 3990
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.589805333402652e-05,
+      "loss": 2.5374,
+      "step": 3995
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5886289198068233e-05,
+      "loss": 2.0687,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "eval_accuracy": 0.8467888855330716,
+      "eval_loss": 2.195303440093994,
+      "eval_runtime": 129.525,
+      "eval_samples_per_second": 82.996,
+      "eval_steps_per_second": 10.376,
+      "step": 4000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.5874525062109948e-05,
+      "loss": 1.9853,
+      "step": 4005
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5862760926151663e-05,
+      "loss": 2.1763,
+      "step": 4010
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5850996790193378e-05,
+      "loss": 2.3896,
+      "step": 4015
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5839232654235092e-05,
+      "loss": 2.1629,
+      "step": 4020
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.582746851827681e-05,
+      "loss": 2.2611,
+      "step": 4025
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5815704382318522e-05,
+      "loss": 2.2159,
+      "step": 4030
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5803940246360236e-05,
+      "loss": 2.4803,
+      "step": 4035
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5792176110401955e-05,
+      "loss": 2.2815,
+      "step": 4040
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5780411974443666e-05,
+      "loss": 2.0492,
+      "step": 4045
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5768647838485384e-05,
+      "loss": 2.0429,
+      "step": 4050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5756883702527095e-05,
+      "loss": 2.1306,
+      "step": 4055
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5745119566568813e-05,
+      "loss": 2.1961,
+      "step": 4060
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5733355430610528e-05,
+      "loss": 2.4967,
+      "step": 4065
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.572159129465224e-05,
+      "loss": 2.3935,
+      "step": 4070
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5709827158693958e-05,
+      "loss": 2.2648,
+      "step": 4075
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5698063022735672e-05,
+      "loss": 2.2732,
+      "step": 4080
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5686298886777387e-05,
+      "loss": 2.2012,
+      "step": 4085
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.56745347508191e-05,
+      "loss": 2.4076,
+      "step": 4090
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5662770614860816e-05,
+      "loss": 2.5336,
+      "step": 4095
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.565100647890253e-05,
+      "loss": 2.1548,
+      "step": 4100
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.5639242342944246e-05,
+      "loss": 2.262,
+      "step": 4105
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.562747820698596e-05,
+      "loss": 2.194,
+      "step": 4110
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5615714071027675e-05,
+      "loss": 2.2487,
+      "step": 4115
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.560394993506939e-05,
+      "loss": 2.294,
+      "step": 4120
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5592185799111105e-05,
+      "loss": 2.0672,
+      "step": 4125
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.558042166315282e-05,
+      "loss": 2.1674,
+      "step": 4130
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5568657527194534e-05,
+      "loss": 2.5388,
+      "step": 4135
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5556893391236252e-05,
+      "loss": 2.1462,
+      "step": 4140
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5545129255277964e-05,
+      "loss": 2.1283,
+      "step": 4145
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5533365119319678e-05,
+      "loss": 2.208,
+      "step": 4150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5521600983361396e-05,
+      "loss": 2.2296,
+      "step": 4155
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5509836847403108e-05,
+      "loss": 1.9881,
+      "step": 4160
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.5498072711444826e-05,
+      "loss": 1.9306,
+      "step": 4165
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.548630857548654e-05,
+      "loss": 1.8521,
+      "step": 4170
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5474544439528252e-05,
+      "loss": 1.9348,
+      "step": 4175
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.546278030356997e-05,
+      "loss": 2.2334,
+      "step": 4180
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5451016167611685e-05,
+      "loss": 2.1633,
+      "step": 4185
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.54392520316534e-05,
+      "loss": 2.1847,
+      "step": 4190
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5427487895695114e-05,
+      "loss": 1.6947,
+      "step": 4195
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.541572375973683e-05,
+      "loss": 2.2602,
+      "step": 4200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5403959623778544e-05,
+      "loss": 2.1107,
+      "step": 4205
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5392195487820258e-05,
+      "loss": 2.1561,
+      "step": 4210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5380431351861973e-05,
+      "loss": 2.1755,
+      "step": 4215
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.5368667215903688e-05,
+      "loss": 1.997,
+      "step": 4220
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5356903079945402e-05,
+      "loss": 2.2894,
+      "step": 4225
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5345138943987117e-05,
+      "loss": 2.3433,
+      "step": 4230
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5333374808028832e-05,
+      "loss": 2.4997,
+      "step": 4235
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5321610672070547e-05,
+      "loss": 2.1669,
+      "step": 4240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5309846536112265e-05,
+      "loss": 2.3885,
+      "step": 4245
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5298082400153976e-05,
+      "loss": 1.9504,
+      "step": 4250
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.528631826419569e-05,
+      "loss": 2.2695,
+      "step": 4255
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.527455412823741e-05,
+      "loss": 2.1053,
+      "step": 4260
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.526278999227912e-05,
+      "loss": 2.4957,
+      "step": 4265
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.5251025856320837e-05,
+      "loss": 2.022,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5239261720362553e-05,
+      "loss": 2.3187,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5227497584404266e-05,
+      "loss": 2.3057,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5215733448445982e-05,
+      "loss": 2.2161,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5203969312487697e-05,
+      "loss": 2.1756,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.519220517652941e-05,
+      "loss": 1.9107,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5180441040571127e-05,
+      "loss": 2.2568,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5168676904612841e-05,
+      "loss": 2.1279,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5156912768654556e-05,
+      "loss": 2.0274,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.514514863269627e-05,
+      "loss": 2.5754,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.5133384496737987e-05,
+      "loss": 2.1466,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.51216203607797e-05,
+      "loss": 2.1166,
+      "step": 4325
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5109856224821415e-05,
+      "loss": 2.3554,
+      "step": 4330
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5098092088863131e-05,
+      "loss": 2.0979,
+      "step": 4335
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5086327952904844e-05,
+      "loss": 2.3625,
+      "step": 4340
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.507456381694656e-05,
+      "loss": 2.1524,
+      "step": 4345
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5062799680988275e-05,
+      "loss": 2.2373,
+      "step": 4350
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5051035545029988e-05,
+      "loss": 2.1559,
+      "step": 4355
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5039271409071705e-05,
+      "loss": 2.1871,
+      "step": 4360
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5027507273113421e-05,
+      "loss": 2.153,
+      "step": 4365
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5015743137155134e-05,
+      "loss": 1.995,
+      "step": 4370
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.5003979001196849e-05,
+      "loss": 2.0231,
+      "step": 4375
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4992214865238565e-05,
+      "loss": 2.0107,
+      "step": 4380
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4980450729280278e-05,
+      "loss": 2.2434,
+      "step": 4385
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4968686593321995e-05,
+      "loss": 2.1751,
+      "step": 4390
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.495692245736371e-05,
+      "loss": 1.9206,
+      "step": 4395
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4945158321405423e-05,
+      "loss": 2.2718,
+      "step": 4400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4933394185447139e-05,
+      "loss": 2.192,
+      "step": 4405
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4921630049488852e-05,
+      "loss": 2.2962,
+      "step": 4410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4909865913530568e-05,
+      "loss": 1.9712,
+      "step": 4415
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4898101777572283e-05,
+      "loss": 2.3632,
+      "step": 4420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4886337641613996e-05,
+      "loss": 2.1704,
+      "step": 4425
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4874573505655713e-05,
+      "loss": 2.1859,
+      "step": 4430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4862809369697429e-05,
+      "loss": 2.6158,
+      "step": 4435
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4851045233739142e-05,
+      "loss": 2.2184,
+      "step": 4440
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4839281097780857e-05,
+      "loss": 2.2017,
+      "step": 4445
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4827516961822573e-05,
+      "loss": 2.2485,
+      "step": 4450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4815752825864286e-05,
+      "loss": 2.5214,
+      "step": 4455
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4803988689906003e-05,
+      "loss": 2.1791,
+      "step": 4460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4792224553947717e-05,
+      "loss": 2.1083,
+      "step": 4465
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.478046041798943e-05,
+      "loss": 1.8394,
+      "step": 4470
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4768696282031147e-05,
+      "loss": 2.3171,
+      "step": 4475
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4756932146072861e-05,
+      "loss": 2.3575,
+      "step": 4480
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4745168010114576e-05,
+      "loss": 2.2285,
+      "step": 4485
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.473340387415629e-05,
+      "loss": 2.2141,
+      "step": 4490
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4721639738198007e-05,
+      "loss": 2.4565,
+      "step": 4495
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.470987560223972e-05,
+      "loss": 2.5206,
+      "step": 4500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4698111466281435e-05,
+      "loss": 2.1014,
+      "step": 4505
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4686347330323151e-05,
+      "loss": 1.8273,
+      "step": 4510
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4674583194364864e-05,
+      "loss": 2.317,
+      "step": 4515
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.466281905840658e-05,
+      "loss": 2.2974,
+      "step": 4520
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4651054922448296e-05,
+      "loss": 2.2476,
+      "step": 4525
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4639290786490009e-05,
+      "loss": 2.0463,
+      "step": 4530
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4627526650531725e-05,
+      "loss": 2.0462,
+      "step": 4535
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4615762514573441e-05,
+      "loss": 2.2713,
+      "step": 4540
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4603998378615154e-05,
+      "loss": 2.1563,
+      "step": 4545
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4592234242656869e-05,
+      "loss": 2.1645,
+      "step": 4550
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4580470106698586e-05,
+      "loss": 2.3747,
+      "step": 4555
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4568705970740299e-05,
+      "loss": 2.2706,
+      "step": 4560
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4556941834782015e-05,
+      "loss": 2.4378,
+      "step": 4565
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.454517769882373e-05,
+      "loss": 2.2351,
+      "step": 4570
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4533413562865443e-05,
+      "loss": 1.9393,
+      "step": 4575
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4521649426907159e-05,
+      "loss": 2.2963,
+      "step": 4580
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4509885290948876e-05,
+      "loss": 2.2647,
+      "step": 4585
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4498121154990589e-05,
+      "loss": 2.385,
+      "step": 4590
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4486357019032303e-05,
+      "loss": 2.2644,
+      "step": 4595
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.447459288307402e-05,
+      "loss": 2.1553,
+      "step": 4600
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4462828747115733e-05,
+      "loss": 2.1284,
+      "step": 4605
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4451064611157449e-05,
+      "loss": 2.5055,
+      "step": 4610
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4439300475199164e-05,
+      "loss": 2.1191,
+      "step": 4615
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4427536339240877e-05,
+      "loss": 1.9399,
+      "step": 4620
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4415772203282593e-05,
+      "loss": 2.0715,
+      "step": 4625
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4404008067324308e-05,
+      "loss": 2.1948,
+      "step": 4630
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4392243931366023e-05,
+      "loss": 1.867,
+      "step": 4635
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4380479795407737e-05,
+      "loss": 1.9101,
+      "step": 4640
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4368715659449454e-05,
+      "loss": 2.1086,
+      "step": 4645
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4356951523491167e-05,
+      "loss": 2.2111,
+      "step": 4650
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4345187387532882e-05,
+      "loss": 2.1627,
+      "step": 4655
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4333423251574598e-05,
+      "loss": 2.4241,
+      "step": 4660
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4321659115616311e-05,
+      "loss": 2.3132,
+      "step": 4665
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4309894979658027e-05,
+      "loss": 2.1344,
+      "step": 4670
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4298130843699742e-05,
+      "loss": 2.2755,
+      "step": 4675
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4286366707741455e-05,
+      "loss": 2.1657,
+      "step": 4680
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4274602571783172e-05,
+      "loss": 1.8609,
+      "step": 4685
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4262838435824888e-05,
+      "loss": 2.1934,
+      "step": 4690
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4251074299866601e-05,
+      "loss": 2.0181,
+      "step": 4695
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4239310163908316e-05,
+      "loss": 2.1753,
+      "step": 4700
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4227546027950032e-05,
+      "loss": 2.4454,
+      "step": 4705
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4215781891991745e-05,
+      "loss": 1.9769,
+      "step": 4710
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4204017756033462e-05,
+      "loss": 2.1848,
+      "step": 4715
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4192253620075176e-05,
+      "loss": 2.0797,
+      "step": 4720
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.418048948411689e-05,
+      "loss": 2.225,
+      "step": 4725
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4168725348158606e-05,
+      "loss": 2.4235,
+      "step": 4730
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4156961212200322e-05,
+      "loss": 2.1811,
+      "step": 4735
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4145197076242035e-05,
+      "loss": 1.8886,
+      "step": 4740
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.413343294028375e-05,
+      "loss": 2.438,
+      "step": 4745
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4121668804325466e-05,
+      "loss": 2.3625,
+      "step": 4750
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.410990466836718e-05,
+      "loss": 2.1022,
+      "step": 4755
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4098140532408896e-05,
+      "loss": 2.256,
+      "step": 4760
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4086376396450609e-05,
+      "loss": 2.3552,
+      "step": 4765
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4074612260492323e-05,
+      "loss": 2.2754,
+      "step": 4770
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.406284812453404e-05,
+      "loss": 2.0031,
+      "step": 4775
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4051083988575753e-05,
+      "loss": 2.3274,
+      "step": 4780
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.403931985261747e-05,
+      "loss": 2.1057,
+      "step": 4785
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4027555716659184e-05,
+      "loss": 2.411,
+      "step": 4790
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4015791580700897e-05,
+      "loss": 2.1887,
+      "step": 4795
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4004027444742613e-05,
+      "loss": 1.8688,
+      "step": 4800
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3992263308784328e-05,
+      "loss": 2.4514,
+      "step": 4805
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3980499172826043e-05,
+      "loss": 2.2174,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3968735036867758e-05,
+      "loss": 2.1479,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3956970900909474e-05,
+      "loss": 2.1097,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3945206764951187e-05,
+      "loss": 1.9096,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3933442628992902e-05,
+      "loss": 2.3103,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3921678493034618e-05,
+      "loss": 1.8826,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3909914357076331e-05,
+      "loss": 2.2167,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3898150221118048e-05,
+      "loss": 2.0824,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3886386085159762e-05,
+      "loss": 2.224,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3874621949201475e-05,
+      "loss": 2.0534,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3862857813243192e-05,
+      "loss": 2.3489,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.3851093677284908e-05,
+      "loss": 2.5083,
+      "step": 4865
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3839329541326621e-05,
+      "loss": 2.109,
+      "step": 4870
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3827565405368336e-05,
+      "loss": 2.408,
+      "step": 4875
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3815801269410052e-05,
+      "loss": 2.2965,
+      "step": 4880
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3804037133451765e-05,
+      "loss": 2.2492,
+      "step": 4885
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3792272997493482e-05,
+      "loss": 2.146,
+      "step": 4890
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3780508861535196e-05,
+      "loss": 2.6407,
+      "step": 4895
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.376874472557691e-05,
+      "loss": 2.1612,
+      "step": 4900
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3756980589618626e-05,
+      "loss": 2.3467,
+      "step": 4905
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3745216453660342e-05,
+      "loss": 2.1353,
+      "step": 4910
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.3733452317702055e-05,
+      "loss": 2.5929,
+      "step": 4915
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.372168818174377e-05,
+      "loss": 2.3245,
+      "step": 4920
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3709924045785486e-05,
+      "loss": 1.8362,
+      "step": 4925
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.36981599098272e-05,
+      "loss": 2.1472,
+      "step": 4930
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3686395773868916e-05,
+      "loss": 2.381,
+      "step": 4935
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.367463163791063e-05,
+      "loss": 2.2104,
+      "step": 4940
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3662867501952344e-05,
+      "loss": 2.4152,
+      "step": 4945
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.365110336599406e-05,
+      "loss": 2.5162,
+      "step": 4950
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3639339230035775e-05,
+      "loss": 2.2522,
+      "step": 4955
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.362757509407749e-05,
+      "loss": 2.0108,
+      "step": 4960
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3615810958119204e-05,
+      "loss": 2.3476,
+      "step": 4965
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.360404682216092e-05,
+      "loss": 2.0804,
+      "step": 4970
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3592282686202634e-05,
+      "loss": 1.9018,
+      "step": 4975
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3580518550244348e-05,
+      "loss": 2.6082,
+      "step": 4980
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3568754414286065e-05,
+      "loss": 2.032,
+      "step": 4985
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3556990278327778e-05,
+      "loss": 2.1587,
+      "step": 4990
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3545226142369494e-05,
+      "loss": 2.4488,
+      "step": 4995
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3533462006411209e-05,
+      "loss": 2.1738,
+      "step": 5000
+    },
+    {
+      "epoch": 0.93,
+      "eval_accuracy": 0.8472137320044297,
+      "eval_loss": 2.1542866230010986,
+      "eval_runtime": 129.4009,
+      "eval_samples_per_second": 83.075,
+      "eval_steps_per_second": 10.386,
+      "step": 5000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3521697870452922e-05,
+      "loss": 2.0456,
+      "step": 5005
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3509933734494638e-05,
+      "loss": 2.1973,
+      "step": 5010
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3498169598536355e-05,
+      "loss": 2.2501,
+      "step": 5015
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3486405462578068e-05,
+      "loss": 2.1929,
+      "step": 5020
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3474641326619782e-05,
+      "loss": 2.1211,
+      "step": 5025
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3462877190661499e-05,
+      "loss": 2.2028,
+      "step": 5030
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3451113054703212e-05,
+      "loss": 2.0214,
+      "step": 5035
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3439348918744928e-05,
+      "loss": 2.2788,
+      "step": 5040
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3427584782786643e-05,
+      "loss": 2.1209,
+      "step": 5045
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3415820646828356e-05,
+      "loss": 2.0796,
+      "step": 5050
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3404056510870072e-05,
+      "loss": 2.0959,
+      "step": 5055
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3392292374911789e-05,
+      "loss": 2.2969,
+      "step": 5060
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3380528238953502e-05,
+      "loss": 1.9857,
+      "step": 5065
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3368764102995217e-05,
+      "loss": 2.0833,
+      "step": 5070
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3356999967036933e-05,
+      "loss": 2.543,
+      "step": 5075
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.3345235831078646e-05,
+      "loss": 2.4396,
+      "step": 5080
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3333471695120362e-05,
+      "loss": 2.249,
+      "step": 5085
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3321707559162077e-05,
+      "loss": 1.6987,
+      "step": 5090
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.330994342320379e-05,
+      "loss": 2.0587,
+      "step": 5095
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3298179287245507e-05,
+      "loss": 2.0922,
+      "step": 5100
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3286415151287223e-05,
+      "loss": 2.1312,
+      "step": 5105
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3274651015328936e-05,
+      "loss": 2.2167,
+      "step": 5110
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.326288687937065e-05,
+      "loss": 2.0439,
+      "step": 5115
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3251122743412367e-05,
+      "loss": 1.6765,
+      "step": 5120
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.323935860745408e-05,
+      "loss": 2.1156,
+      "step": 5125
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3227594471495796e-05,
+      "loss": 2.0455,
+      "step": 5130
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.321583033553751e-05,
+      "loss": 1.8213,
+      "step": 5135
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3204066199579224e-05,
+      "loss": 2.2903,
+      "step": 5140
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.319230206362094e-05,
+      "loss": 2.183,
+      "step": 5145
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3180537927662654e-05,
+      "loss": 2.2773,
+      "step": 5150
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.316877379170437e-05,
+      "loss": 2.1562,
+      "step": 5155
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3157009655746085e-05,
+      "loss": 2.406,
+      "step": 5160
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3145245519787798e-05,
+      "loss": 2.1582,
+      "step": 5165
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3133481383829514e-05,
+      "loss": 2.528,
+      "step": 5170
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3121717247871229e-05,
+      "loss": 2.4469,
+      "step": 5175
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3109953111912944e-05,
+      "loss": 2.1881,
+      "step": 5180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3098188975954658e-05,
+      "loss": 2.461,
+      "step": 5185
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3086424839996375e-05,
+      "loss": 2.353,
+      "step": 5190
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3074660704038088e-05,
+      "loss": 1.8327,
+      "step": 5195
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3062896568079803e-05,
+      "loss": 2.2302,
+      "step": 5200
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3051132432121519e-05,
+      "loss": 2.3622,
+      "step": 5205
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3039368296163232e-05,
+      "loss": 2.2509,
+      "step": 5210
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3027604160204948e-05,
+      "loss": 2.0527,
+      "step": 5215
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3015840024246663e-05,
+      "loss": 2.086,
+      "step": 5220
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3004075888288376e-05,
+      "loss": 2.5221,
+      "step": 5225
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2992311752330093e-05,
+      "loss": 2.0695,
+      "step": 5230
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2980547616371809e-05,
+      "loss": 2.272,
+      "step": 5235
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2968783480413522e-05,
+      "loss": 2.019,
+      "step": 5240
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2957019344455237e-05,
+      "loss": 1.8066,
+      "step": 5245
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2945255208496953e-05,
+      "loss": 2.0143,
+      "step": 5250
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2933491072538666e-05,
+      "loss": 2.1565,
+      "step": 5255
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2921726936580383e-05,
+      "loss": 2.3011,
+      "step": 5260
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2909962800622097e-05,
+      "loss": 1.9746,
+      "step": 5265
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.289819866466381e-05,
+      "loss": 2.0396,
+      "step": 5270
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2886434528705527e-05,
+      "loss": 2.1294,
+      "step": 5275
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2874670392747243e-05,
+      "loss": 2.1141,
+      "step": 5280
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2862906256788956e-05,
+      "loss": 2.0758,
+      "step": 5285
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.285114212083067e-05,
+      "loss": 2.1814,
+      "step": 5290
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2839377984872387e-05,
+      "loss": 1.8827,
+      "step": 5295
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.28276138489141e-05,
+      "loss": 1.8691,
+      "step": 5300
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2815849712955817e-05,
+      "loss": 2.1813,
+      "step": 5305
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2804085576997531e-05,
+      "loss": 2.123,
+      "step": 5310
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2792321441039244e-05,
+      "loss": 2.2108,
+      "step": 5315
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.278055730508096e-05,
+      "loss": 1.9969,
+      "step": 5320
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2768793169122676e-05,
+      "loss": 2.3264,
+      "step": 5325
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.275702903316439e-05,
+      "loss": 1.977,
+      "step": 5330
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2745264897206105e-05,
+      "loss": 2.4235,
+      "step": 5335
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2733500761247821e-05,
+      "loss": 2.0733,
+      "step": 5340
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2721736625289534e-05,
+      "loss": 2.0025,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2709972489331249e-05,
+      "loss": 2.2249,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2698208353372965e-05,
+      "loss": 2.2994,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2686444217414679e-05,
+      "loss": 2.0879,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2674680081456395e-05,
+      "loss": 2.1515,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.266291594549811e-05,
+      "loss": 2.0858,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2651151809539823e-05,
+      "loss": 2.4817,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2639387673581539e-05,
+      "loss": 1.8792,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2627623537623255e-05,
+      "loss": 1.9015,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2615859401664969e-05,
+      "loss": 1.8139,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2604095265706683e-05,
+      "loss": 2.2508,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.25923311297484e-05,
+      "loss": 1.8315,
+      "step": 5400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2580566993790113e-05,
+      "loss": 2.0378,
+      "step": 5405
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2568802857831829e-05,
+      "loss": 2.0258,
+      "step": 5410
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2557038721873544e-05,
+      "loss": 1.9814,
+      "step": 5415
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2545274585915257e-05,
+      "loss": 1.9256,
+      "step": 5420
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2533510449956973e-05,
+      "loss": 1.8259,
+      "step": 5425
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.252174631399869e-05,
+      "loss": 1.8801,
+      "step": 5430
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2509982178040403e-05,
+      "loss": 1.8796,
+      "step": 5435
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2498218042082117e-05,
+      "loss": 1.9875,
+      "step": 5440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2486453906123834e-05,
+      "loss": 1.8781,
+      "step": 5445
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2474689770165547e-05,
+      "loss": 1.7595,
+      "step": 5450
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.2462925634207263e-05,
+      "loss": 2.2036,
+      "step": 5455
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2451161498248978e-05,
+      "loss": 1.8098,
+      "step": 5460
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2439397362290691e-05,
+      "loss": 2.2482,
+      "step": 5465
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2427633226332407e-05,
+      "loss": 1.962,
+      "step": 5470
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2415869090374122e-05,
+      "loss": 1.9131,
+      "step": 5475
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2404104954415837e-05,
+      "loss": 1.7542,
+      "step": 5480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2392340818457551e-05,
+      "loss": 2.259,
+      "step": 5485
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2380576682499265e-05,
+      "loss": 1.7425,
+      "step": 5490
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2368812546540981e-05,
+      "loss": 1.6552,
+      "step": 5495
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2357048410582696e-05,
+      "loss": 1.9705,
+      "step": 5500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.234528427462441e-05,
+      "loss": 1.9675,
+      "step": 5505
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2333520138666125e-05,
+      "loss": 1.9024,
+      "step": 5510
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2321756002707841e-05,
+      "loss": 1.8468,
+      "step": 5515
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2309991866749555e-05,
+      "loss": 1.843,
+      "step": 5520
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.229822773079127e-05,
+      "loss": 2.0675,
+      "step": 5525
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2286463594832986e-05,
+      "loss": 1.6987,
+      "step": 5530
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2274699458874699e-05,
+      "loss": 2.0907,
+      "step": 5535
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2262935322916415e-05,
+      "loss": 1.8166,
+      "step": 5540
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.225117118695813e-05,
+      "loss": 1.8721,
+      "step": 5545
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2239407050999843e-05,
+      "loss": 1.9928,
+      "step": 5550
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.222764291504156e-05,
+      "loss": 2.0307,
+      "step": 5555
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.2215878779083276e-05,
+      "loss": 1.8114,
+      "step": 5560
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2204114643124989e-05,
+      "loss": 1.8102,
+      "step": 5565
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2192350507166703e-05,
+      "loss": 2.0205,
+      "step": 5570
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.218058637120842e-05,
+      "loss": 1.63,
+      "step": 5575
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2168822235250133e-05,
+      "loss": 1.9191,
+      "step": 5580
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.215705809929185e-05,
+      "loss": 1.9026,
+      "step": 5585
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2145293963333564e-05,
+      "loss": 1.7594,
+      "step": 5590
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2133529827375277e-05,
+      "loss": 1.6511,
+      "step": 5595
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2121765691416993e-05,
+      "loss": 1.5948,
+      "step": 5600
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.211000155545871e-05,
+      "loss": 1.8899,
+      "step": 5605
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2098237419500423e-05,
+      "loss": 1.8666,
+      "step": 5610
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.2086473283542137e-05,
+      "loss": 1.9487,
+      "step": 5615
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2074709147583854e-05,
+      "loss": 1.7876,
+      "step": 5620
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2062945011625567e-05,
+      "loss": 1.947,
+      "step": 5625
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2051180875667283e-05,
+      "loss": 1.8845,
+      "step": 5630
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2039416739708998e-05,
+      "loss": 1.9375,
+      "step": 5635
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2027652603750711e-05,
+      "loss": 1.5105,
+      "step": 5640
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2015888467792427e-05,
+      "loss": 1.6819,
+      "step": 5645
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.2004124331834142e-05,
+      "loss": 1.8186,
+      "step": 5650
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.1992360195875857e-05,
+      "loss": 1.992,
+      "step": 5655
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.1980596059917572e-05,
+      "loss": 1.7257,
+      "step": 5660
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.1968831923959288e-05,
+      "loss": 2.0831,
+      "step": 5665
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.1957067788001001e-05,
+      "loss": 1.9934,
+      "step": 5670
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1945303652042716e-05,
+      "loss": 1.8708,
+      "step": 5675
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1933539516084432e-05,
+      "loss": 2.0009,
+      "step": 5680
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1921775380126145e-05,
+      "loss": 1.9864,
+      "step": 5685
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1910011244167862e-05,
+      "loss": 2.0763,
+      "step": 5690
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1898247108209576e-05,
+      "loss": 2.0896,
+      "step": 5695
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.188648297225129e-05,
+      "loss": 1.804,
+      "step": 5700
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1874718836293006e-05,
+      "loss": 1.9783,
+      "step": 5705
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1862954700334722e-05,
+      "loss": 1.9844,
+      "step": 5710
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1851190564376435e-05,
+      "loss": 1.904,
+      "step": 5715
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.183942642841815e-05,
+      "loss": 1.7514,
+      "step": 5720
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.1827662292459866e-05,
+      "loss": 1.8182,
+      "step": 5725
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.181589815650158e-05,
+      "loss": 1.8074,
+      "step": 5730
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1804134020543296e-05,
+      "loss": 1.9018,
+      "step": 5735
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.179236988458501e-05,
+      "loss": 1.8714,
+      "step": 5740
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1780605748626724e-05,
+      "loss": 2.0276,
+      "step": 5745
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.176884161266844e-05,
+      "loss": 1.8631,
+      "step": 5750
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1757077476710156e-05,
+      "loss": 1.9271,
+      "step": 5755
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.174531334075187e-05,
+      "loss": 2.0708,
+      "step": 5760
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1733549204793584e-05,
+      "loss": 1.8673,
+      "step": 5765
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.17217850688353e-05,
+      "loss": 1.7587,
+      "step": 5770
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.1710020932877013e-05,
+      "loss": 2.1394,
+      "step": 5775
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.169825679691873e-05,
+      "loss": 1.7978,
+      "step": 5780
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1686492660960445e-05,
+      "loss": 2.0853,
+      "step": 5785
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1674728525002158e-05,
+      "loss": 1.6652,
+      "step": 5790
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1662964389043874e-05,
+      "loss": 1.7521,
+      "step": 5795
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1651200253085589e-05,
+      "loss": 1.9535,
+      "step": 5800
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1639436117127303e-05,
+      "loss": 2.1948,
+      "step": 5805
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1627671981169018e-05,
+      "loss": 1.6979,
+      "step": 5810
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1615907845210735e-05,
+      "loss": 1.5995,
+      "step": 5815
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1604143709252448e-05,
+      "loss": 1.6516,
+      "step": 5820
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1592379573294162e-05,
+      "loss": 1.8618,
+      "step": 5825
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.1580615437335879e-05,
+      "loss": 1.9039,
+      "step": 5830
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1568851301377592e-05,
+      "loss": 1.9106,
+      "step": 5835
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1557087165419308e-05,
+      "loss": 1.9609,
+      "step": 5840
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1545323029461021e-05,
+      "loss": 2.186,
+      "step": 5845
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1533558893502736e-05,
+      "loss": 1.8374,
+      "step": 5850
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1521794757544452e-05,
+      "loss": 2.1121,
+      "step": 5855
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1510030621586165e-05,
+      "loss": 2.0299,
+      "step": 5860
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1498266485627882e-05,
+      "loss": 1.9002,
+      "step": 5865
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1486502349669596e-05,
+      "loss": 1.7937,
+      "step": 5870
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.147473821371131e-05,
+      "loss": 1.8247,
+      "step": 5875
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1462974077753026e-05,
+      "loss": 1.9874,
+      "step": 5880
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.1451209941794742e-05,
+      "loss": 1.7359,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1439445805836457e-05,
+      "loss": 1.7051,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.142768166987817e-05,
+      "loss": 2.0265,
+      "step": 5895
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1415917533919886e-05,
+      "loss": 1.947,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1404153397961601e-05,
+      "loss": 2.0248,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1392389262003316e-05,
+      "loss": 1.6476,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.138062512604503e-05,
+      "loss": 2.0258,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1368860990086745e-05,
+      "loss": 1.6575,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.135709685412846e-05,
+      "loss": 1.9802,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1345332718170175e-05,
+      "loss": 2.0276,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.133356858221189e-05,
+      "loss": 1.812,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.1321804446253604e-05,
+      "loss": 2.0398,
+      "step": 5940
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1310040310295319e-05,
+      "loss": 2.2775,
+      "step": 5945
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1298276174337035e-05,
+      "loss": 2.0812,
+      "step": 5950
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.128651203837875e-05,
+      "loss": 1.9775,
+      "step": 5955
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1274747902420463e-05,
+      "loss": 1.717,
+      "step": 5960
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.126298376646218e-05,
+      "loss": 1.932,
+      "step": 5965
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1251219630503894e-05,
+      "loss": 1.9041,
+      "step": 5970
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1239455494545609e-05,
+      "loss": 1.8285,
+      "step": 5975
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1227691358587324e-05,
+      "loss": 1.8585,
+      "step": 5980
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1215927222629038e-05,
+      "loss": 1.8462,
+      "step": 5985
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.1204163086670753e-05,
+      "loss": 1.7823,
+      "step": 5990
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.119239895071247e-05,
+      "loss": 1.9703,
+      "step": 5995
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1180634814754182e-05,
+      "loss": 1.8554,
+      "step": 6000
+    },
+    {
+      "epoch": 1.12,
+      "eval_accuracy": 0.8474708547266687,
+      "eval_loss": 2.1500444412231445,
+      "eval_runtime": 129.3864,
+      "eval_samples_per_second": 83.084,
+      "eval_steps_per_second": 10.387,
+      "step": 6000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1168870678795897e-05,
+      "loss": 2.1431,
+      "step": 6005
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1157106542837614e-05,
+      "loss": 1.6554,
+      "step": 6010
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1145342406879328e-05,
+      "loss": 2.0275,
+      "step": 6015
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1133578270921043e-05,
+      "loss": 2.057,
+      "step": 6020
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1121814134962758e-05,
+      "loss": 1.7979,
+      "step": 6025
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1110049999004472e-05,
+      "loss": 1.4648,
+      "step": 6030
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1098285863046187e-05,
+      "loss": 1.8252,
+      "step": 6035
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1086521727087904e-05,
+      "loss": 1.8758,
+      "step": 6040
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.1074757591129617e-05,
+      "loss": 1.9125,
+      "step": 6045
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1062993455171331e-05,
+      "loss": 1.7944,
+      "step": 6050
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1051229319213048e-05,
+      "loss": 1.9629,
+      "step": 6055
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1039465183254762e-05,
+      "loss": 1.9488,
+      "step": 6060
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1027701047296477e-05,
+      "loss": 1.7122,
+      "step": 6065
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1015936911338192e-05,
+      "loss": 1.8631,
+      "step": 6070
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.1004172775379907e-05,
+      "loss": 1.993,
+      "step": 6075
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.0992408639421621e-05,
+      "loss": 1.8524,
+      "step": 6080
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.0980644503463336e-05,
+      "loss": 2.1574,
+      "step": 6085
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.096888036750505e-05,
+      "loss": 1.8759,
+      "step": 6090
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.0957116231546765e-05,
+      "loss": 1.8705,
+      "step": 6095
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.0945352095588482e-05,
+      "loss": 1.8946,
+      "step": 6100
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0933587959630197e-05,
+      "loss": 1.8043,
+      "step": 6105
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.092182382367191e-05,
+      "loss": 1.6761,
+      "step": 6110
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0910059687713624e-05,
+      "loss": 1.7802,
+      "step": 6115
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.089829555175534e-05,
+      "loss": 2.0636,
+      "step": 6120
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0886531415797055e-05,
+      "loss": 1.7088,
+      "step": 6125
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.087476727983877e-05,
+      "loss": 2.1411,
+      "step": 6130
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0863003143880485e-05,
+      "loss": 1.9638,
+      "step": 6135
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.08512390079222e-05,
+      "loss": 2.0859,
+      "step": 6140
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0839474871963914e-05,
+      "loss": 1.9132,
+      "step": 6145
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.082771073600563e-05,
+      "loss": 1.7481,
+      "step": 6150
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.0815946600047344e-05,
+      "loss": 1.767,
+      "step": 6155
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0804182464089058e-05,
+      "loss": 1.9778,
+      "step": 6160
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0792418328130775e-05,
+      "loss": 1.8963,
+      "step": 6165
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.078065419217249e-05,
+      "loss": 2.1058,
+      "step": 6170
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0768890056214204e-05,
+      "loss": 1.9526,
+      "step": 6175
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0757125920255919e-05,
+      "loss": 1.9574,
+      "step": 6180
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0745361784297634e-05,
+      "loss": 1.9474,
+      "step": 6185
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0733597648339348e-05,
+      "loss": 1.7678,
+      "step": 6190
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0721833512381063e-05,
+      "loss": 2.1175,
+      "step": 6195
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0710069376422778e-05,
+      "loss": 1.8637,
+      "step": 6200
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.0698305240464493e-05,
+      "loss": 1.9491,
+      "step": 6205
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0686541104506209e-05,
+      "loss": 1.9199,
+      "step": 6210
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0674776968547924e-05,
+      "loss": 1.7572,
+      "step": 6215
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0663012832589637e-05,
+      "loss": 1.9684,
+      "step": 6220
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0651248696631353e-05,
+      "loss": 1.8952,
+      "step": 6225
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0639484560673068e-05,
+      "loss": 1.7841,
+      "step": 6230
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0627720424714783e-05,
+      "loss": 1.6059,
+      "step": 6235
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0615956288756497e-05,
+      "loss": 1.8697,
+      "step": 6240
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0604192152798212e-05,
+      "loss": 2.1396,
+      "step": 6245
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0592428016839927e-05,
+      "loss": 2.0393,
+      "step": 6250
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0580663880881643e-05,
+      "loss": 2.0431,
+      "step": 6255
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.0568899744923356e-05,
+      "loss": 1.8958,
+      "step": 6260
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0557135608965071e-05,
+      "loss": 1.8377,
+      "step": 6265
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0545371473006787e-05,
+      "loss": 1.7989,
+      "step": 6270
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0533607337048502e-05,
+      "loss": 1.9754,
+      "step": 6275
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0521843201090217e-05,
+      "loss": 1.7924,
+      "step": 6280
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0510079065131931e-05,
+      "loss": 1.8734,
+      "step": 6285
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0498314929173646e-05,
+      "loss": 2.1106,
+      "step": 6290
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0486550793215361e-05,
+      "loss": 2.041,
+      "step": 6295
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0474786657257076e-05,
+      "loss": 1.6627,
+      "step": 6300
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.046302252129879e-05,
+      "loss": 1.724,
+      "step": 6305
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.0451258385340505e-05,
+      "loss": 1.9233,
+      "step": 6310
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.043949424938222e-05,
+      "loss": 2.2463,
+      "step": 6315
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0427730113423936e-05,
+      "loss": 1.8343,
+      "step": 6320
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0415965977465651e-05,
+      "loss": 2.011,
+      "step": 6325
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0404201841507364e-05,
+      "loss": 1.7403,
+      "step": 6330
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.039243770554908e-05,
+      "loss": 1.9348,
+      "step": 6335
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0380673569590795e-05,
+      "loss": 1.6817,
+      "step": 6340
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.036890943363251e-05,
+      "loss": 1.7877,
+      "step": 6345
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0357145297674224e-05,
+      "loss": 1.9522,
+      "step": 6350
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.034538116171594e-05,
+      "loss": 2.1021,
+      "step": 6355
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0333617025757654e-05,
+      "loss": 1.8498,
+      "step": 6360
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.032185288979937e-05,
+      "loss": 2.0173,
+      "step": 6365
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.0310088753841083e-05,
+      "loss": 1.71,
+      "step": 6370
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0298324617882798e-05,
+      "loss": 2.0036,
+      "step": 6375
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0286560481924514e-05,
+      "loss": 2.0275,
+      "step": 6380
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.027479634596623e-05,
+      "loss": 2.194,
+      "step": 6385
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0263032210007944e-05,
+      "loss": 1.9075,
+      "step": 6390
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0251268074049659e-05,
+      "loss": 1.8197,
+      "step": 6395
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0239503938091373e-05,
+      "loss": 1.8786,
+      "step": 6400
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0227739802133088e-05,
+      "loss": 1.8362,
+      "step": 6405
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0215975666174803e-05,
+      "loss": 1.8807,
+      "step": 6410
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0204211530216517e-05,
+      "loss": 1.915,
+      "step": 6415
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0192447394258232e-05,
+      "loss": 1.8205,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0180683258299949e-05,
+      "loss": 1.6785,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0168919122341663e-05,
+      "loss": 2.096,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0157154986383376e-05,
+      "loss": 2.1267,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0145390850425093e-05,
+      "loss": 2.026,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0133626714466807e-05,
+      "loss": 1.9932,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0121862578508522e-05,
+      "loss": 1.7378,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0110098442550237e-05,
+      "loss": 2.033,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0098334306591952e-05,
+      "loss": 1.8207,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0086570170633666e-05,
+      "loss": 1.8621,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0074806034675381e-05,
+      "loss": 1.8703,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.0063041898717097e-05,
+      "loss": 2.0857,
+      "step": 6475
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.005127776275881e-05,
+      "loss": 1.7986,
+      "step": 6480
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.0039513626800525e-05,
+      "loss": 1.8944,
+      "step": 6485
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.0027749490842242e-05,
+      "loss": 1.9103,
+      "step": 6490
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.0015985354883956e-05,
+      "loss": 1.9713,
+      "step": 6495
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.0004221218925671e-05,
+      "loss": 1.701,
+      "step": 6500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.992457082967386e-06,
+      "loss": 1.5407,
+      "step": 6505
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.9806929470091e-06,
+      "loss": 1.9869,
+      "step": 6510
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.968928811050815e-06,
+      "loss": 2.098,
+      "step": 6515
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.95716467509253e-06,
+      "loss": 1.9601,
+      "step": 6520
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.945400539134245e-06,
+      "loss": 1.6523,
+      "step": 6525
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.93363640317596e-06,
+      "loss": 1.6735,
+      "step": 6530
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.921872267217676e-06,
+      "loss": 2.119,
+      "step": 6535
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.91010813125939e-06,
+      "loss": 1.8703,
+      "step": 6540
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.898343995301103e-06,
+      "loss": 1.8844,
+      "step": 6545
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.88657985934282e-06,
+      "loss": 1.7218,
+      "step": 6550
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.874815723384535e-06,
+      "loss": 1.7885,
+      "step": 6555
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.86305158742625e-06,
+      "loss": 1.8457,
+      "step": 6560
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.851287451467964e-06,
+      "loss": 1.8258,
+      "step": 6565
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.839523315509679e-06,
+      "loss": 1.9701,
+      "step": 6570
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.827759179551393e-06,
+      "loss": 1.9895,
+      "step": 6575
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.81599504359311e-06,
+      "loss": 1.9944,
+      "step": 6580
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.804230907634823e-06,
+      "loss": 1.9196,
+      "step": 6585
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.792466771676538e-06,
+      "loss": 1.9936,
+      "step": 6590
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.780702635718254e-06,
+      "loss": 1.5199,
+      "step": 6595
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.768938499759969e-06,
+      "loss": 1.8009,
+      "step": 6600
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.757174363801683e-06,
+      "loss": 1.7893,
+      "step": 6605
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.745410227843398e-06,
+      "loss": 1.7771,
+      "step": 6610
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.733646091885113e-06,
+      "loss": 1.8076,
+      "step": 6615
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.721881955926828e-06,
+      "loss": 1.9219,
+      "step": 6620
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.710117819968544e-06,
+      "loss": 1.999,
+      "step": 6625
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.698353684010257e-06,
+      "loss": 1.8125,
+      "step": 6630
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.686589548051972e-06,
+      "loss": 2.2106,
+      "step": 6635
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.674825412093688e-06,
+      "loss": 1.7235,
+      "step": 6640
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.663061276135403e-06,
+      "loss": 2.0446,
+      "step": 6645
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.651297140177118e-06,
+      "loss": 1.7209,
+      "step": 6650
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.63953300421883e-06,
+      "loss": 2.0344,
+      "step": 6655
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.627768868260547e-06,
+      "loss": 1.8583,
+      "step": 6660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.616004732302262e-06,
+      "loss": 1.6872,
+      "step": 6665
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.604240596343976e-06,
+      "loss": 2.1454,
+      "step": 6670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.592476460385691e-06,
+      "loss": 1.8174,
+      "step": 6675
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.580712324427406e-06,
+      "loss": 1.818,
+      "step": 6680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.56894818846912e-06,
+      "loss": 1.9371,
+      "step": 6685
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.557184052510837e-06,
+      "loss": 1.4802,
+      "step": 6690
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.54541991655255e-06,
+      "loss": 1.5377,
+      "step": 6695
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.533655780594265e-06,
+      "loss": 2.0516,
+      "step": 6700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.521891644635981e-06,
+      "loss": 1.7354,
+      "step": 6705
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.510127508677696e-06,
+      "loss": 1.5009,
+      "step": 6710
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.49836337271941e-06,
+      "loss": 2.0768,
+      "step": 6715
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.486599236761125e-06,
+      "loss": 1.7429,
+      "step": 6720
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.47483510080284e-06,
+      "loss": 1.7665,
+      "step": 6725
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.463070964844555e-06,
+      "loss": 1.8592,
+      "step": 6730
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.451306828886271e-06,
+      "loss": 1.9791,
+      "step": 6735
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.439542692927984e-06,
+      "loss": 2.1448,
+      "step": 6740
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.427778556969699e-06,
+      "loss": 1.9291,
+      "step": 6745
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.416014421011415e-06,
+      "loss": 1.9738,
+      "step": 6750
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.40425028505313e-06,
+      "loss": 1.7188,
+      "step": 6755
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.392486149094845e-06,
+      "loss": 1.8967,
+      "step": 6760
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.38072201313656e-06,
+      "loss": 1.9697,
+      "step": 6765
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.368957877178274e-06,
+      "loss": 2.0377,
+      "step": 6770
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.357193741219989e-06,
+      "loss": 1.8753,
+      "step": 6775
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.345429605261704e-06,
+      "loss": 2.2034,
+      "step": 6780
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.333665469303418e-06,
+      "loss": 1.886,
+      "step": 6785
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.321901333345133e-06,
+      "loss": 1.7994,
+      "step": 6790
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.31013719738685e-06,
+      "loss": 1.7304,
+      "step": 6795
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.298373061428564e-06,
+      "loss": 1.9934,
+      "step": 6800
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.286608925470277e-06,
+      "loss": 2.0714,
+      "step": 6805
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.274844789511994e-06,
+      "loss": 1.9513,
+      "step": 6810
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.263080653553708e-06,
+      "loss": 2.176,
+      "step": 6815
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.251316517595423e-06,
+      "loss": 1.9228,
+      "step": 6820
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.239552381637138e-06,
+      "loss": 1.9366,
+      "step": 6825
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.227788245678852e-06,
+      "loss": 1.6386,
+      "step": 6830
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.216024109720567e-06,
+      "loss": 1.617,
+      "step": 6835
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.204259973762282e-06,
+      "loss": 1.8727,
+      "step": 6840
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.192495837803997e-06,
+      "loss": 2.0035,
+      "step": 6845
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.180731701845711e-06,
+      "loss": 1.6326,
+      "step": 6850
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.168967565887426e-06,
+      "loss": 1.7872,
+      "step": 6855
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.157203429929142e-06,
+      "loss": 1.9379,
+      "step": 6860
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.145439293970857e-06,
+      "loss": 1.6053,
+      "step": 6865
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.13367515801257e-06,
+      "loss": 1.8292,
+      "step": 6870
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.121911022054287e-06,
+      "loss": 1.857,
+      "step": 6875
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.110146886096001e-06,
+      "loss": 1.7884,
+      "step": 6880
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.098382750137716e-06,
+      "loss": 1.6977,
+      "step": 6885
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.08661861417943e-06,
+      "loss": 2.0604,
+      "step": 6890
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.074854478221145e-06,
+      "loss": 1.9084,
+      "step": 6895
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.06309034226286e-06,
+      "loss": 1.7692,
+      "step": 6900
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.051326206304577e-06,
+      "loss": 1.8571,
+      "step": 6905
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.039562070346291e-06,
+      "loss": 1.8225,
+      "step": 6910
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.027797934388004e-06,
+      "loss": 1.8273,
+      "step": 6915
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.01603379842972e-06,
+      "loss": 2.0332,
+      "step": 6920
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.004269662471435e-06,
+      "loss": 1.7937,
+      "step": 6925
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.99250552651315e-06,
+      "loss": 1.7206,
+      "step": 6930
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.980741390554865e-06,
+      "loss": 1.688,
+      "step": 6935
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.96897725459658e-06,
+      "loss": 1.9035,
+      "step": 6940
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.957213118638294e-06,
+      "loss": 1.6544,
+      "step": 6945
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.94544898268001e-06,
+      "loss": 1.8483,
+      "step": 6950
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.933684846721724e-06,
+      "loss": 1.8354,
+      "step": 6955
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 8.921920710763438e-06,
+      "loss": 1.8915,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.910156574805155e-06,
+      "loss": 2.0755,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.89839243884687e-06,
+      "loss": 2.0566,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.886628302888584e-06,
+      "loss": 1.8168,
+      "step": 6975
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.874864166930299e-06,
+      "loss": 1.7823,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.863100030972014e-06,
+      "loss": 1.9992,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.851335895013728e-06,
+      "loss": 1.8741,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.839571759055443e-06,
+      "loss": 1.8838,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.827807623097158e-06,
+      "loss": 1.9276,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3,
+      "eval_accuracy": 0.8476983791402396,
+      "eval_loss": 2.1223256587982178,
+      "eval_runtime": 129.4929,
+      "eval_samples_per_second": 83.016,
+      "eval_steps_per_second": 10.379,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.816043487138873e-06,
+      "loss": 1.7621,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.804279351180587e-06,
+      "loss": 2.0429,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 8.792515215222304e-06,
+      "loss": 2.2024,
+      "step": 7015
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.780751079264017e-06,
+      "loss": 1.9041,
+      "step": 7020
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.768986943305731e-06,
+      "loss": 1.724,
+      "step": 7025
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.757222807347448e-06,
+      "loss": 1.8509,
+      "step": 7030
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.745458671389163e-06,
+      "loss": 1.9666,
+      "step": 7035
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.733694535430877e-06,
+      "loss": 1.6957,
+      "step": 7040
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.721930399472592e-06,
+      "loss": 1.8726,
+      "step": 7045
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.710166263514307e-06,
+      "loss": 2.0265,
+      "step": 7050
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.698402127556021e-06,
+      "loss": 1.7862,
+      "step": 7055
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.686637991597738e-06,
+      "loss": 1.5634,
+      "step": 7060
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 8.674873855639451e-06,
+      "loss": 2.0236,
+      "step": 7065
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.663109719681166e-06,
+      "loss": 1.8506,
+      "step": 7070
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.651345583722882e-06,
+      "loss": 1.6327,
+      "step": 7075
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.639581447764597e-06,
+      "loss": 1.8199,
+      "step": 7080
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.627817311806311e-06,
+      "loss": 1.6521,
+      "step": 7085
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.616053175848026e-06,
+      "loss": 1.6226,
+      "step": 7090
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.604289039889741e-06,
+      "loss": 2.0402,
+      "step": 7095
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.592524903931456e-06,
+      "loss": 2.0009,
+      "step": 7100
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.58076076797317e-06,
+      "loss": 1.7264,
+      "step": 7105
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.568996632014885e-06,
+      "loss": 2.0781,
+      "step": 7110
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.5572324960566e-06,
+      "loss": 1.7499,
+      "step": 7115
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 8.545468360098316e-06,
+      "loss": 1.942,
+      "step": 7120
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.533704224140031e-06,
+      "loss": 1.8795,
+      "step": 7125
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.521940088181744e-06,
+      "loss": 1.7453,
+      "step": 7130
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.51017595222346e-06,
+      "loss": 2.1427,
+      "step": 7135
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.498411816265175e-06,
+      "loss": 1.9748,
+      "step": 7140
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.48664768030689e-06,
+      "loss": 1.6533,
+      "step": 7145
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.474883544348604e-06,
+      "loss": 1.4142,
+      "step": 7150
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.463119408390319e-06,
+      "loss": 1.9294,
+      "step": 7155
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.451355272432034e-06,
+      "loss": 1.7086,
+      "step": 7160
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.43959113647375e-06,
+      "loss": 1.7291,
+      "step": 7165
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.427827000515465e-06,
+      "loss": 1.8013,
+      "step": 7170
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.416062864557178e-06,
+      "loss": 1.9423,
+      "step": 7175
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.404298728598894e-06,
+      "loss": 1.8555,
+      "step": 7180
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.392534592640609e-06,
+      "loss": 1.8039,
+      "step": 7185
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.380770456682324e-06,
+      "loss": 1.8363,
+      "step": 7190
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.369006320724039e-06,
+      "loss": 1.8829,
+      "step": 7195
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.357242184765753e-06,
+      "loss": 1.7913,
+      "step": 7200
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.345478048807468e-06,
+      "loss": 1.7596,
+      "step": 7205
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.333713912849183e-06,
+      "loss": 1.9339,
+      "step": 7210
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.321949776890897e-06,
+      "loss": 2.0406,
+      "step": 7215
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.310185640932612e-06,
+      "loss": 1.7827,
+      "step": 7220
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.298421504974327e-06,
+      "loss": 1.9488,
+      "step": 7225
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.286657369016043e-06,
+      "loss": 1.9942,
+      "step": 7230
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.274893233057758e-06,
+      "loss": 1.9188,
+      "step": 7235
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.263129097099471e-06,
+      "loss": 1.5903,
+      "step": 7240
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.251364961141187e-06,
+      "loss": 1.8215,
+      "step": 7245
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.239600825182902e-06,
+      "loss": 1.8615,
+      "step": 7250
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.227836689224617e-06,
+      "loss": 1.7496,
+      "step": 7255
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.216072553266332e-06,
+      "loss": 1.9657,
+      "step": 7260
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.204308417308046e-06,
+      "loss": 1.6968,
+      "step": 7265
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.192544281349761e-06,
+      "loss": 1.7301,
+      "step": 7270
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.180780145391477e-06,
+      "loss": 1.8773,
+      "step": 7275
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 8.16901600943319e-06,
+      "loss": 1.6158,
+      "step": 7280
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.157251873474905e-06,
+      "loss": 1.9048,
+      "step": 7285
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.145487737516622e-06,
+      "loss": 1.7323,
+      "step": 7290
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.133723601558336e-06,
+      "loss": 1.977,
+      "step": 7295
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.121959465600051e-06,
+      "loss": 2.0594,
+      "step": 7300
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.110195329641766e-06,
+      "loss": 2.0709,
+      "step": 7305
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.09843119368348e-06,
+      "loss": 1.8705,
+      "step": 7310
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.086667057725195e-06,
+      "loss": 1.6715,
+      "step": 7315
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.074902921766912e-06,
+      "loss": 1.7166,
+      "step": 7320
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.063138785808625e-06,
+      "loss": 1.781,
+      "step": 7325
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.05137464985034e-06,
+      "loss": 2.0216,
+      "step": 7330
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 8.039610513892056e-06,
+      "loss": 1.9923,
+      "step": 7335
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 8.02784637793377e-06,
+      "loss": 1.8937,
+      "step": 7340
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 8.016082241975485e-06,
+      "loss": 1.9217,
+      "step": 7345
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 8.0043181060172e-06,
+      "loss": 1.8002,
+      "step": 7350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.992553970058915e-06,
+      "loss": 1.8248,
+      "step": 7355
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.98078983410063e-06,
+      "loss": 1.6303,
+      "step": 7360
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.969025698142344e-06,
+      "loss": 1.6364,
+      "step": 7365
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.957261562184059e-06,
+      "loss": 1.5364,
+      "step": 7370
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.945497426225773e-06,
+      "loss": 1.7259,
+      "step": 7375
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.933733290267488e-06,
+      "loss": 2.0054,
+      "step": 7380
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.921969154309205e-06,
+      "loss": 1.9358,
+      "step": 7385
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.910205018350918e-06,
+      "loss": 1.677,
+      "step": 7390
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.898440882392632e-06,
+      "loss": 1.8709,
+      "step": 7395
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.886676746434349e-06,
+      "loss": 1.8622,
+      "step": 7400
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.874912610476063e-06,
+      "loss": 1.9117,
+      "step": 7405
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.863148474517778e-06,
+      "loss": 1.7819,
+      "step": 7410
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.851384338559493e-06,
+      "loss": 1.9088,
+      "step": 7415
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.839620202601208e-06,
+      "loss": 1.8226,
+      "step": 7420
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.827856066642922e-06,
+      "loss": 1.7416,
+      "step": 7425
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.816091930684637e-06,
+      "loss": 1.729,
+      "step": 7430
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.804327794726352e-06,
+      "loss": 1.7448,
+      "step": 7435
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.792563658768066e-06,
+      "loss": 1.9784,
+      "step": 7440
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.780799522809783e-06,
+      "loss": 1.8046,
+      "step": 7445
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.769035386851498e-06,
+      "loss": 1.9487,
+      "step": 7450
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.75727125089321e-06,
+      "loss": 1.9343,
+      "step": 7455
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.745507114934927e-06,
+      "loss": 2.4002,
+      "step": 7460
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.733742978976642e-06,
+      "loss": 1.6184,
+      "step": 7465
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.721978843018356e-06,
+      "loss": 1.996,
+      "step": 7470
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.710214707060071e-06,
+      "loss": 1.9318,
+      "step": 7475
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.698450571101786e-06,
+      "loss": 1.7346,
+      "step": 7480
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.6866864351435e-06,
+      "loss": 1.837,
+      "step": 7485
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.674922299185217e-06,
+      "loss": 1.5632,
+      "step": 7490
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 7.663158163226932e-06,
+      "loss": 1.658,
+      "step": 7495
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.651394027268645e-06,
+      "loss": 1.6994,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.639629891310361e-06,
+      "loss": 1.7616,
+      "step": 7505
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.627865755352076e-06,
+      "loss": 2.037,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.61610161939379e-06,
+      "loss": 1.7281,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.604337483435506e-06,
+      "loss": 1.7088,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.59257334747722e-06,
+      "loss": 1.9751,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.580809211518935e-06,
+      "loss": 1.8002,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.56904507556065e-06,
+      "loss": 1.7723,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.557280939602365e-06,
+      "loss": 1.6989,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.54551680364408e-06,
+      "loss": 1.9106,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 7.5337526676857936e-06,
+      "loss": 1.5693,
+      "step": 7550
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.521988531727509e-06,
+      "loss": 1.6974,
+      "step": 7555
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.510224395769224e-06,
+      "loss": 1.8855,
+      "step": 7560
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.4984602598109386e-06,
+      "loss": 1.5739,
+      "step": 7565
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.486696123852654e-06,
+      "loss": 1.9979,
+      "step": 7570
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.474931987894369e-06,
+      "loss": 1.8717,
+      "step": 7575
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.463167851936083e-06,
+      "loss": 1.9176,
+      "step": 7580
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.451403715977799e-06,
+      "loss": 1.5578,
+      "step": 7585
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.439639580019513e-06,
+      "loss": 2.2408,
+      "step": 7590
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.427875444061228e-06,
+      "loss": 1.7498,
+      "step": 7595
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.416111308102943e-06,
+      "loss": 1.9328,
+      "step": 7600
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 7.404347172144658e-06,
+      "loss": 2.1611,
+      "step": 7605
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.392583036186373e-06,
+      "loss": 1.9141,
+      "step": 7610
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.380818900228088e-06,
+      "loss": 1.6278,
+      "step": 7615
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.369054764269803e-06,
+      "loss": 1.777,
+      "step": 7620
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.357290628311517e-06,
+      "loss": 2.0055,
+      "step": 7625
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.345526492353233e-06,
+      "loss": 1.743,
+      "step": 7630
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.333762356394947e-06,
+      "loss": 1.9667,
+      "step": 7635
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.321998220436662e-06,
+      "loss": 1.8485,
+      "step": 7640
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.310234084478377e-06,
+      "loss": 1.8966,
+      "step": 7645
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.298469948520092e-06,
+      "loss": 1.9837,
+      "step": 7650
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.286705812561807e-06,
+      "loss": 1.7558,
+      "step": 7655
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 7.274941676603522e-06,
+      "loss": 2.0033,
+      "step": 7660
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.263177540645236e-06,
+      "loss": 1.8117,
+      "step": 7665
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.251413404686951e-06,
+      "loss": 1.6726,
+      "step": 7670
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.2396492687286665e-06,
+      "loss": 1.7312,
+      "step": 7675
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.227885132770381e-06,
+      "loss": 1.8003,
+      "step": 7680
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.216120996812096e-06,
+      "loss": 1.573,
+      "step": 7685
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.2043568608538115e-06,
+      "loss": 1.8955,
+      "step": 7690
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.192592724895526e-06,
+      "loss": 1.6835,
+      "step": 7695
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.18082858893724e-06,
+      "loss": 1.9213,
+      "step": 7700
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.1690644529789565e-06,
+      "loss": 1.9134,
+      "step": 7705
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 7.15730031702067e-06,
+      "loss": 1.8141,
+      "step": 7710
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.145536181062385e-06,
+      "loss": 1.947,
+      "step": 7715
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.133772045104101e-06,
+      "loss": 1.4904,
+      "step": 7720
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.122007909145815e-06,
+      "loss": 1.4855,
+      "step": 7725
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.11024377318753e-06,
+      "loss": 1.7716,
+      "step": 7730
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.098479637229244e-06,
+      "loss": 1.7952,
+      "step": 7735
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.0867155012709596e-06,
+      "loss": 1.8193,
+      "step": 7740
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.074951365312674e-06,
+      "loss": 1.9532,
+      "step": 7745
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.063187229354389e-06,
+      "loss": 1.6561,
+      "step": 7750
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.0514230933961045e-06,
+      "loss": 1.7668,
+      "step": 7755
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.039658957437819e-06,
+      "loss": 1.6245,
+      "step": 7760
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 7.027894821479533e-06,
+      "loss": 1.8736,
+      "step": 7765
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 7.0161306855212495e-06,
+      "loss": 1.3439,
+      "step": 7770
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 7.004366549562963e-06,
+      "loss": 1.7836,
+      "step": 7775
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.992602413604678e-06,
+      "loss": 1.8209,
+      "step": 7780
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.980838277646394e-06,
+      "loss": 1.8775,
+      "step": 7785
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.969074141688108e-06,
+      "loss": 1.8956,
+      "step": 7790
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.957310005729823e-06,
+      "loss": 1.8403,
+      "step": 7795
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.945545869771539e-06,
+      "loss": 1.7407,
+      "step": 7800
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.933781733813253e-06,
+      "loss": 1.7399,
+      "step": 7805
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.922017597854967e-06,
+      "loss": 1.9348,
+      "step": 7810
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.910253461896683e-06,
+      "loss": 2.1156,
+      "step": 7815
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 6.8984893259383975e-06,
+      "loss": 1.9742,
+      "step": 7820
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.886725189980112e-06,
+      "loss": 1.8624,
+      "step": 7825
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.874961054021828e-06,
+      "loss": 1.9871,
+      "step": 7830
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.8631969180635425e-06,
+      "loss": 1.6763,
+      "step": 7835
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.851432782105256e-06,
+      "loss": 1.7253,
+      "step": 7840
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.839668646146973e-06,
+      "loss": 2.0094,
+      "step": 7845
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.827904510188687e-06,
+      "loss": 1.9092,
+      "step": 7850
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.816140374230401e-06,
+      "loss": 1.7728,
+      "step": 7855
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.804376238272117e-06,
+      "loss": 1.9439,
+      "step": 7860
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.792612102313832e-06,
+      "loss": 2.2257,
+      "step": 7865
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.780847966355546e-06,
+      "loss": 2.111,
+      "step": 7870
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 6.769083830397262e-06,
+      "loss": 1.8134,
+      "step": 7875
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.757319694438977e-06,
+      "loss": 1.7535,
+      "step": 7880
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.7455555584806905e-06,
+      "loss": 1.9242,
+      "step": 7885
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.733791422522406e-06,
+      "loss": 1.993,
+      "step": 7890
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.722027286564121e-06,
+      "loss": 1.8782,
+      "step": 7895
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.7102631506058355e-06,
+      "loss": 1.7447,
+      "step": 7900
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.69849901464755e-06,
+      "loss": 1.9785,
+      "step": 7905
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.686734878689266e-06,
+      "loss": 1.6696,
+      "step": 7910
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.67497074273098e-06,
+      "loss": 1.9958,
+      "step": 7915
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.663206606772694e-06,
+      "loss": 1.8955,
+      "step": 7920
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 6.65144247081441e-06,
+      "loss": 1.8291,
+      "step": 7925
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.639678334856125e-06,
+      "loss": 1.8951,
+      "step": 7930
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.627914198897839e-06,
+      "loss": 2.2041,
+      "step": 7935
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.616150062939555e-06,
+      "loss": 1.7803,
+      "step": 7940
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.60438592698127e-06,
+      "loss": 1.864,
+      "step": 7945
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.5926217910229836e-06,
+      "loss": 1.8535,
+      "step": 7950
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.5808576550647e-06,
+      "loss": 1.8598,
+      "step": 7955
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.569093519106414e-06,
+      "loss": 1.7623,
+      "step": 7960
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.5573293831481285e-06,
+      "loss": 1.7875,
+      "step": 7965
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.545565247189844e-06,
+      "loss": 1.8891,
+      "step": 7970
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.533801111231559e-06,
+      "loss": 2.0535,
+      "step": 7975
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 6.5220369752732735e-06,
+      "loss": 1.8859,
+      "step": 7980
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.510272839314989e-06,
+      "loss": 1.7183,
+      "step": 7985
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.498508703356703e-06,
+      "loss": 1.659,
+      "step": 7990
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.486744567398418e-06,
+      "loss": 1.9087,
+      "step": 7995
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.474980431440133e-06,
+      "loss": 1.7988,
+      "step": 8000
+    },
+    {
+      "epoch": 1.49,
+      "eval_accuracy": 0.8478570421826236,
+      "eval_loss": 2.1119508743286133,
+      "eval_runtime": 129.4307,
+      "eval_samples_per_second": 83.056,
+      "eval_steps_per_second": 10.384,
+      "step": 8000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.463216295481848e-06,
+      "loss": 1.9162,
+      "step": 8005
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.451452159523563e-06,
+      "loss": 2.0717,
+      "step": 8010
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.439688023565278e-06,
+      "loss": 1.6121,
+      "step": 8015
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.427923887606993e-06,
+      "loss": 1.7125,
+      "step": 8020
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.416159751648707e-06,
+      "loss": 2.0044,
+      "step": 8025
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.404395615690423e-06,
+      "loss": 2.1258,
+      "step": 8030
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 6.392631479732137e-06,
+      "loss": 1.6559,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.380867343773852e-06,
+      "loss": 1.8379,
+      "step": 8040
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.369103207815567e-06,
+      "loss": 2.0337,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.357339071857282e-06,
+      "loss": 1.7674,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.345574935898997e-06,
+      "loss": 2.0009,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.333810799940712e-06,
+      "loss": 1.6838,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.322046663982426e-06,
+      "loss": 1.6851,
+      "step": 8065
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.310282528024141e-06,
+      "loss": 1.8202,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.2985183920658565e-06,
+      "loss": 1.8773,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.286754256107571e-06,
+      "loss": 1.6838,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.274990120149286e-06,
+      "loss": 1.8643,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 6.263225984191e-06,
+      "loss": 1.8226,
+      "step": 8090
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.251461848232716e-06,
+      "loss": 2.0646,
+      "step": 8095
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.23969771227443e-06,
+      "loss": 1.636,
+      "step": 8100
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.227933576316145e-06,
+      "loss": 1.6926,
+      "step": 8105
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.21616944035786e-06,
+      "loss": 1.7297,
+      "step": 8110
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.204405304399575e-06,
+      "loss": 2.206,
+      "step": 8115
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.19264116844129e-06,
+      "loss": 2.0745,
+      "step": 8120
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.180877032483005e-06,
+      "loss": 2.0228,
+      "step": 8125
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.16911289652472e-06,
+      "loss": 1.9831,
+      "step": 8130
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.157348760566434e-06,
+      "loss": 1.5342,
+      "step": 8135
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 6.14558462460815e-06,
+      "loss": 1.9515,
+      "step": 8140
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.133820488649864e-06,
+      "loss": 1.9011,
+      "step": 8145
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.122056352691579e-06,
+      "loss": 1.8775,
+      "step": 8150
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.1102922167332945e-06,
+      "loss": 1.9665,
+      "step": 8155
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.098528080775009e-06,
+      "loss": 1.8151,
+      "step": 8160
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.086763944816724e-06,
+      "loss": 2.343,
+      "step": 8165
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.0749998088584395e-06,
+      "loss": 1.7105,
+      "step": 8170
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.063235672900153e-06,
+      "loss": 2.1118,
+      "step": 8175
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.051471536941868e-06,
+      "loss": 1.7032,
+      "step": 8180
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.039707400983584e-06,
+      "loss": 1.7681,
+      "step": 8185
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.027943265025298e-06,
+      "loss": 1.6487,
+      "step": 8190
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 6.016179129067013e-06,
+      "loss": 1.8716,
+      "step": 8195
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 6.004414993108729e-06,
+      "loss": 1.838,
+      "step": 8200
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.992650857150443e-06,
+      "loss": 1.8428,
+      "step": 8205
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.980886721192157e-06,
+      "loss": 1.8634,
+      "step": 8210
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.969122585233874e-06,
+      "loss": 1.9455,
+      "step": 8215
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.9573584492755875e-06,
+      "loss": 1.7631,
+      "step": 8220
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.945594313317302e-06,
+      "loss": 1.891,
+      "step": 8225
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.933830177359018e-06,
+      "loss": 1.8222,
+      "step": 8230
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.9220660414007325e-06,
+      "loss": 1.7573,
+      "step": 8235
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.910301905442447e-06,
+      "loss": 1.8815,
+      "step": 8240
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.898537769484163e-06,
+      "loss": 1.6142,
+      "step": 8245
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.886773633525877e-06,
+      "loss": 1.7924,
+      "step": 8250
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.875009497567591e-06,
+      "loss": 1.8616,
+      "step": 8255
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.863245361609307e-06,
+      "loss": 1.9623,
+      "step": 8260
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.851481225651022e-06,
+      "loss": 1.8305,
+      "step": 8265
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.839717089692736e-06,
+      "loss": 1.8184,
+      "step": 8270
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.82795295373445e-06,
+      "loss": 1.8152,
+      "step": 8275
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.816188817776167e-06,
+      "loss": 1.8384,
+      "step": 8280
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.8044246818178805e-06,
+      "loss": 1.4474,
+      "step": 8285
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.792660545859595e-06,
+      "loss": 1.7699,
+      "step": 8290
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.780896409901311e-06,
+      "loss": 1.9611,
+      "step": 8295
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.7691322739430255e-06,
+      "loss": 1.7652,
+      "step": 8300
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.75736813798474e-06,
+      "loss": 2.1775,
+      "step": 8305
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.745604002026456e-06,
+      "loss": 1.8659,
+      "step": 8310
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.7338398660681705e-06,
+      "loss": 1.7278,
+      "step": 8315
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.722075730109885e-06,
+      "loss": 1.7779,
+      "step": 8320
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.7103115941516e-06,
+      "loss": 1.719,
+      "step": 8325
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.698547458193315e-06,
+      "loss": 2.0371,
+      "step": 8330
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.68678332223503e-06,
+      "loss": 1.4501,
+      "step": 8335
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.675019186276744e-06,
+      "loss": 1.5519,
+      "step": 8340
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.66325505031846e-06,
+      "loss": 1.8991,
+      "step": 8345
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.651490914360174e-06,
+      "loss": 1.9612,
+      "step": 8350
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.639726778401889e-06,
+      "loss": 1.6888,
+      "step": 8355
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.627962642443604e-06,
+      "loss": 2.2227,
+      "step": 8360
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.616198506485319e-06,
+      "loss": 2.1402,
+      "step": 8365
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.604434370527034e-06,
+      "loss": 1.9068,
+      "step": 8370
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.592670234568749e-06,
+      "loss": 2.0604,
+      "step": 8375
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5809060986104635e-06,
+      "loss": 1.8777,
+      "step": 8380
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.569141962652178e-06,
+      "loss": 1.8356,
+      "step": 8385
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.557377826693894e-06,
+      "loss": 1.9724,
+      "step": 8390
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.545613690735608e-06,
+      "loss": 1.7764,
+      "step": 8395
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.533849554777323e-06,
+      "loss": 1.9299,
+      "step": 8400
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.522085418819038e-06,
+      "loss": 1.6917,
+      "step": 8405
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.510321282860753e-06,
+      "loss": 1.8237,
+      "step": 8410
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.498557146902467e-06,
+      "loss": 2.048,
+      "step": 8415
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.486793010944183e-06,
+      "loss": 1.8278,
+      "step": 8420
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.475028874985897e-06,
+      "loss": 1.6927,
+      "step": 8425
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.463264739027612e-06,
+      "loss": 1.7185,
+      "step": 8430
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.451500603069327e-06,
+      "loss": 1.6067,
+      "step": 8435
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.439736467111042e-06,
+      "loss": 2.1339,
+      "step": 8440
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.427972331152757e-06,
+      "loss": 1.9512,
+      "step": 8445
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.416208195194472e-06,
+      "loss": 1.9791,
+      "step": 8450
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.404444059236187e-06,
+      "loss": 1.6586,
+      "step": 8455
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.3926799232779015e-06,
+      "loss": 1.8016,
+      "step": 8460
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.380915787319617e-06,
+      "loss": 2.0724,
+      "step": 8465
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.369151651361331e-06,
+      "loss": 2.0399,
+      "step": 8470
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.3573875154030465e-06,
+      "loss": 1.8087,
+      "step": 8475
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.345623379444761e-06,
+      "loss": 1.9003,
+      "step": 8480
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.333859243486476e-06,
+      "loss": 1.6732,
+      "step": 8485
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.322095107528191e-06,
+      "loss": 1.9033,
+      "step": 8490
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.310330971569905e-06,
+      "loss": 1.8053,
+      "step": 8495
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.29856683561162e-06,
+      "loss": 1.9723,
+      "step": 8500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.286802699653336e-06,
+      "loss": 1.9261,
+      "step": 8505
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.27503856369505e-06,
+      "loss": 1.8459,
+      "step": 8510
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.263274427736765e-06,
+      "loss": 1.9758,
+      "step": 8515
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.251510291778481e-06,
+      "loss": 1.8425,
+      "step": 8520
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.2397461558201945e-06,
+      "loss": 1.8219,
+      "step": 8525
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.22798201986191e-06,
+      "loss": 1.4872,
+      "step": 8530
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.216217883903625e-06,
+      "loss": 1.7925,
+      "step": 8535
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.2044537479453395e-06,
+      "loss": 2.1242,
+      "step": 8540
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.192689611987054e-06,
+      "loss": 1.8437,
+      "step": 8545
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.18092547602877e-06,
+      "loss": 1.7188,
+      "step": 8550
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.169161340070484e-06,
+      "loss": 1.8353,
+      "step": 8555
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.157397204112199e-06,
+      "loss": 1.8662,
+      "step": 8560
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.145633068153914e-06,
+      "loss": 1.7548,
+      "step": 8565
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.133868932195629e-06,
+      "loss": 1.7481,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.122104796237343e-06,
+      "loss": 2.047,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.110340660279058e-06,
+      "loss": 1.7196,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.098576524320774e-06,
+      "loss": 2.1188,
+      "step": 8585
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.086812388362488e-06,
+      "loss": 1.5506,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.075048252404203e-06,
+      "loss": 1.7931,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.063284116445918e-06,
+      "loss": 1.751,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.051519980487633e-06,
+      "loss": 2.0697,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.039755844529347e-06,
+      "loss": 1.9149,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.027991708571063e-06,
+      "loss": 1.7989,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.0162275726127775e-06,
+      "loss": 1.6959,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.004463436654492e-06,
+      "loss": 1.9891,
+      "step": 8625
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.992699300696207e-06,
+      "loss": 1.929,
+      "step": 8630
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.9809351647379225e-06,
+      "loss": 2.1285,
+      "step": 8635
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.969171028779637e-06,
+      "loss": 1.8805,
+      "step": 8640
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.957406892821352e-06,
+      "loss": 1.9716,
+      "step": 8645
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.9456427568630675e-06,
+      "loss": 1.721,
+      "step": 8650
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.933878620904781e-06,
+      "loss": 1.8971,
+      "step": 8655
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.922114484946497e-06,
+      "loss": 1.8352,
+      "step": 8660
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.910350348988211e-06,
+      "loss": 1.9398,
+      "step": 8665
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.898586213029926e-06,
+      "loss": 1.6944,
+      "step": 8670
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.886822077071641e-06,
+      "loss": 2.0851,
+      "step": 8675
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.875057941113356e-06,
+      "loss": 1.6685,
+      "step": 8680
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.8632938051550705e-06,
+      "loss": 1.8765,
+      "step": 8685
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.851529669196786e-06,
+      "loss": 1.8284,
+      "step": 8690
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.839765533238501e-06,
+      "loss": 1.8727,
+      "step": 8695
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.8280013972802155e-06,
+      "loss": 1.7254,
+      "step": 8700
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.81623726132193e-06,
+      "loss": 1.8444,
+      "step": 8705
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.804473125363645e-06,
+      "loss": 1.7878,
+      "step": 8710
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.7927089894053605e-06,
+      "loss": 2.1156,
+      "step": 8715
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.780944853447075e-06,
+      "loss": 1.8702,
+      "step": 8720
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.76918071748879e-06,
+      "loss": 1.9305,
+      "step": 8725
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.757416581530505e-06,
+      "loss": 1.9285,
+      "step": 8730
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.74565244557222e-06,
+      "loss": 2.067,
+      "step": 8735
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.733888309613934e-06,
+      "loss": 1.8003,
+      "step": 8740
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.72212417365565e-06,
+      "loss": 1.7325,
+      "step": 8745
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.710360037697364e-06,
+      "loss": 2.1279,
+      "step": 8750
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.698595901739079e-06,
+      "loss": 1.7887,
+      "step": 8755
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.686831765780794e-06,
+      "loss": 1.7587,
+      "step": 8760
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.6750676298225085e-06,
+      "loss": 2.1356,
+      "step": 8765
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.663303493864224e-06,
+      "loss": 1.9799,
+      "step": 8770
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.651539357905939e-06,
+      "loss": 1.8402,
+      "step": 8775
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.6397752219476535e-06,
+      "loss": 1.8269,
+      "step": 8780
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.628011085989368e-06,
+      "loss": 1.5749,
+      "step": 8785
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.616246950031084e-06,
+      "loss": 1.7726,
+      "step": 8790
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.604482814072798e-06,
+      "loss": 1.6373,
+      "step": 8795
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.592718678114513e-06,
+      "loss": 1.8965,
+      "step": 8800
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.580954542156228e-06,
+      "loss": 1.7276,
+      "step": 8805
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.569190406197943e-06,
+      "loss": 1.7034,
+      "step": 8810
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.557426270239657e-06,
+      "loss": 1.4389,
+      "step": 8815
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.545662134281373e-06,
+      "loss": 1.76,
+      "step": 8820
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.533897998323088e-06,
+      "loss": 1.762,
+      "step": 8825
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.522133862364802e-06,
+      "loss": 1.5353,
+      "step": 8830
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.510369726406517e-06,
+      "loss": 2.0668,
+      "step": 8835
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.498605590448232e-06,
+      "loss": 1.908,
+      "step": 8840
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.486841454489947e-06,
+      "loss": 1.9192,
+      "step": 8845
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.475077318531661e-06,
+      "loss": 1.8406,
+      "step": 8850
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.463313182573377e-06,
+      "loss": 2.0242,
+      "step": 8855
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.4515490466150915e-06,
+      "loss": 1.7966,
+      "step": 8860
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.439784910656806e-06,
+      "loss": 1.7064,
+      "step": 8865
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.428020774698521e-06,
+      "loss": 1.6817,
+      "step": 8870
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.4162566387402365e-06,
+      "loss": 1.6622,
+      "step": 8875
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.404492502781951e-06,
+      "loss": 2.1309,
+      "step": 8880
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.392728366823666e-06,
+      "loss": 1.8312,
+      "step": 8885
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.380964230865381e-06,
+      "loss": 1.5634,
+      "step": 8890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.369200094907095e-06,
+      "loss": 1.7106,
+      "step": 8895
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.357435958948811e-06,
+      "loss": 1.5431,
+      "step": 8900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.345671822990526e-06,
+      "loss": 1.7091,
+      "step": 8905
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.33390768703224e-06,
+      "loss": 1.9698,
+      "step": 8910
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.322143551073955e-06,
+      "loss": 1.7843,
+      "step": 8915
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.310379415115671e-06,
+      "loss": 1.6524,
+      "step": 8920
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.2986152791573845e-06,
+      "loss": 2.0059,
+      "step": 8925
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.2868511431991e-06,
+      "loss": 2.2107,
+      "step": 8930
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.275087007240814e-06,
+      "loss": 1.8346,
+      "step": 8935
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.2633228712825295e-06,
+      "loss": 1.7397,
+      "step": 8940
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.251558735324244e-06,
+      "loss": 1.9724,
+      "step": 8945
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.239794599365959e-06,
+      "loss": 1.5896,
+      "step": 8950
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.2280304634076745e-06,
+      "loss": 1.9525,
+      "step": 8955
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.216266327449389e-06,
+      "loss": 1.876,
+      "step": 8960
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.204502191491104e-06,
+      "loss": 1.6996,
+      "step": 8965
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.192738055532819e-06,
+      "loss": 1.6365,
+      "step": 8970
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.180973919574534e-06,
+      "loss": 1.8295,
+      "step": 8975
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.169209783616248e-06,
+      "loss": 1.8042,
+      "step": 8980
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.157445647657964e-06,
+      "loss": 1.7111,
+      "step": 8985
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.145681511699678e-06,
+      "loss": 1.888,
+      "step": 8990
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.133917375741393e-06,
+      "loss": 1.9463,
+      "step": 8995
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.122153239783108e-06,
+      "loss": 2.0632,
+      "step": 9000
+    },
+    {
+      "epoch": 1.67,
+      "eval_accuracy": 0.848024161884627,
+      "eval_loss": 2.0973165035247803,
+      "eval_runtime": 129.5041,
+      "eval_samples_per_second": 83.009,
+      "eval_steps_per_second": 10.378,
+      "step": 9000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.110389103824823e-06,
+      "loss": 1.7713,
+      "step": 9005
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.098624967866537e-06,
+      "loss": 1.6065,
+      "step": 9010
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.086860831908253e-06,
+      "loss": 1.6697,
+      "step": 9015
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.0750966959499675e-06,
+      "loss": 1.9018,
+      "step": 9020
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.063332559991682e-06,
+      "loss": 2.0076,
+      "step": 9025
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.051568424033398e-06,
+      "loss": 1.842,
+      "step": 9030
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.039804288075112e-06,
+      "loss": 1.6151,
+      "step": 9035
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.028040152116827e-06,
+      "loss": 1.7152,
+      "step": 9040
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.016276016158542e-06,
+      "loss": 1.6855,
+      "step": 9045
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.004511880200257e-06,
+      "loss": 1.6469,
+      "step": 9050
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.992747744241971e-06,
+      "loss": 1.6682,
+      "step": 9055
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.980983608283687e-06,
+      "loss": 1.9617,
+      "step": 9060
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.969219472325401e-06,
+      "loss": 1.7199,
+      "step": 9065
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.957455336367116e-06,
+      "loss": 2.1332,
+      "step": 9070
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.945691200408831e-06,
+      "loss": 1.8218,
+      "step": 9075
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.933927064450546e-06,
+      "loss": 1.6435,
+      "step": 9080
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.9221629284922605e-06,
+      "loss": 1.8045,
+      "step": 9085
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.910398792533976e-06,
+      "loss": 1.7026,
+      "step": 9090
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.898634656575691e-06,
+      "loss": 1.642,
+      "step": 9095
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.8868705206174055e-06,
+      "loss": 1.7302,
+      "step": 9100
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.875106384659121e-06,
+      "loss": 1.981,
+      "step": 9105
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.863342248700835e-06,
+      "loss": 1.9044,
+      "step": 9110
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.8515781127425505e-06,
+      "loss": 1.6887,
+      "step": 9115
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.839813976784264e-06,
+      "loss": 1.805,
+      "step": 9120
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.82804984082598e-06,
+      "loss": 1.8013,
+      "step": 9125
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.816285704867695e-06,
+      "loss": 1.8952,
+      "step": 9130
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.8045215689094094e-06,
+      "loss": 1.8892,
+      "step": 9135
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7927574329511245e-06,
+      "loss": 1.9537,
+      "step": 9140
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7809932969928396e-06,
+      "loss": 1.9912,
+      "step": 9145
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.769229161034554e-06,
+      "loss": 1.9312,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.757465025076269e-06,
+      "loss": 1.7422,
+      "step": 9155
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.745700889117984e-06,
+      "loss": 1.8821,
+      "step": 9160
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7339367531596985e-06,
+      "loss": 2.054,
+      "step": 9165
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.7221726172014136e-06,
+      "loss": 1.948,
+      "step": 9170
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.7104084812431288e-06,
+      "loss": 1.9095,
+      "step": 9175
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6986443452848435e-06,
+      "loss": 1.9925,
+      "step": 9180
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6868802093265586e-06,
+      "loss": 1.8864,
+      "step": 9185
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6751160733682733e-06,
+      "loss": 1.5928,
+      "step": 9190
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.663351937409988e-06,
+      "loss": 2.2111,
+      "step": 9195
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.651587801451703e-06,
+      "loss": 2.0508,
+      "step": 9200
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6398236654934175e-06,
+      "loss": 1.8638,
+      "step": 9205
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6280595295351326e-06,
+      "loss": 1.8192,
+      "step": 9210
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6162953935768478e-06,
+      "loss": 1.9464,
+      "step": 9215
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.604531257618562e-06,
+      "loss": 1.9671,
+      "step": 9220
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.592767121660277e-06,
+      "loss": 1.8535,
+      "step": 9225
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.5810029857019923e-06,
+      "loss": 1.8808,
+      "step": 9230
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.569238849743707e-06,
+      "loss": 1.7286,
+      "step": 9235
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.5574747137854218e-06,
+      "loss": 1.6662,
+      "step": 9240
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.545710577827137e-06,
+      "loss": 1.8211,
+      "step": 9245
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.5339464418688516e-06,
+      "loss": 1.4916,
+      "step": 9250
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.5221823059105668e-06,
+      "loss": 1.7472,
+      "step": 9255
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.510418169952282e-06,
+      "loss": 1.7452,
+      "step": 9260
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.498654033993996e-06,
+      "loss": 1.6811,
+      "step": 9265
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.4868898980357113e-06,
+      "loss": 1.6402,
+      "step": 9270
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.4751257620774265e-06,
+      "loss": 1.9137,
+      "step": 9275
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.4633616261191408e-06,
+      "loss": 1.948,
+      "step": 9280
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.451597490160856e-06,
+      "loss": 1.7804,
+      "step": 9285
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.439833354202571e-06,
+      "loss": 1.7908,
+      "step": 9290
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.4280692182442853e-06,
+      "loss": 1.7568,
+      "step": 9295
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.4163050822860005e-06,
+      "loss": 1.6,
+      "step": 9300
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.404540946327715e-06,
+      "loss": 1.8525,
+      "step": 9305
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.3927768103694303e-06,
+      "loss": 2.2561,
+      "step": 9310
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.381012674411145e-06,
+      "loss": 1.7422,
+      "step": 9315
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.3692485384528598e-06,
+      "loss": 1.8654,
+      "step": 9320
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.357484402494575e-06,
+      "loss": 1.9354,
+      "step": 9325
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.34572026653629e-06,
+      "loss": 1.6806,
+      "step": 9330
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.3339561305780043e-06,
+      "loss": 1.8392,
+      "step": 9335
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.3221919946197195e-06,
+      "loss": 1.9343,
+      "step": 9340
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.3104278586614346e-06,
+      "loss": 1.8175,
+      "step": 9345
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.298663722703149e-06,
+      "loss": 1.5308,
+      "step": 9350
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.286899586744864e-06,
+      "loss": 1.774,
+      "step": 9355
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.275135450786579e-06,
+      "loss": 1.8929,
+      "step": 9360
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.2633713148282935e-06,
+      "loss": 1.7822,
+      "step": 9365
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.2516071788700086e-06,
+      "loss": 1.9426,
+      "step": 9370
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.2398430429117238e-06,
+      "loss": 1.8915,
+      "step": 9375
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.2280789069534385e-06,
+      "loss": 1.7575,
+      "step": 9380
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.2163147709951536e-06,
+      "loss": 2.075,
+      "step": 9385
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.204550635036868e-06,
+      "loss": 2.0316,
+      "step": 9390
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.192786499078583e-06,
+      "loss": 1.6989,
+      "step": 9395
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.181022363120298e-06,
+      "loss": 2.0262,
+      "step": 9400
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.1692582271620125e-06,
+      "loss": 1.6655,
+      "step": 9405
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.1574940912037276e-06,
+      "loss": 1.8843,
+      "step": 9410
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.1457299552454428e-06,
+      "loss": 1.8461,
+      "step": 9415
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.133965819287157e-06,
+      "loss": 1.7459,
+      "step": 9420
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.122201683328872e-06,
+      "loss": 1.637,
+      "step": 9425
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.1104375473705873e-06,
+      "loss": 1.7036,
+      "step": 9430
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.098673411412302e-06,
+      "loss": 1.497,
+      "step": 9435
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.086909275454017e-06,
+      "loss": 1.911,
+      "step": 9440
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.075145139495732e-06,
+      "loss": 1.9572,
+      "step": 9445
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.0633810035374466e-06,
+      "loss": 1.6736,
+      "step": 9450
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.0516168675791618e-06,
+      "loss": 1.8196,
+      "step": 9455
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.039852731620877e-06,
+      "loss": 1.6948,
+      "step": 9460
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.028088595662591e-06,
+      "loss": 1.9161,
+      "step": 9465
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.0163244597043063e-06,
+      "loss": 1.7115,
+      "step": 9470
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.0045603237460206e-06,
+      "loss": 1.8408,
+      "step": 9475
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.9927961877877358e-06,
+      "loss": 1.9541,
+      "step": 9480
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.981032051829451e-06,
+      "loss": 1.8345,
+      "step": 9485
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9692679158711656e-06,
+      "loss": 1.7361,
+      "step": 9490
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9575037799128803e-06,
+      "loss": 2.0323,
+      "step": 9495
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9457396439545955e-06,
+      "loss": 1.7042,
+      "step": 9500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.93397550799631e-06,
+      "loss": 1.841,
+      "step": 9505
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9222113720380253e-06,
+      "loss": 1.8559,
+      "step": 9510
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.9104472360797405e-06,
+      "loss": 1.6402,
+      "step": 9515
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.8986831001214548e-06,
+      "loss": 2.0415,
+      "step": 9520
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.88691896416317e-06,
+      "loss": 1.6592,
+      "step": 9525
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.875154828204885e-06,
+      "loss": 1.89,
+      "step": 9530
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.8633906922465998e-06,
+      "loss": 1.7824,
+      "step": 9535
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.8516265562883145e-06,
+      "loss": 2.0037,
+      "step": 9540
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.839862420330029e-06,
+      "loss": 2.1169,
+      "step": 9545
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.828098284371744e-06,
+      "loss": 2.017,
+      "step": 9550
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.816334148413459e-06,
+      "loss": 1.4411,
+      "step": 9555
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.8045700124551738e-06,
+      "loss": 1.7493,
+      "step": 9560
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.792805876496889e-06,
+      "loss": 1.8857,
+      "step": 9565
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7810417405386036e-06,
+      "loss": 1.6793,
+      "step": 9570
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7692776045803188e-06,
+      "loss": 1.8343,
+      "step": 9575
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7575134686220335e-06,
+      "loss": 1.9257,
+      "step": 9580
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.745749332663748e-06,
+      "loss": 1.8629,
+      "step": 9585
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7339851967054633e-06,
+      "loss": 2.0206,
+      "step": 9590
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.722221060747178e-06,
+      "loss": 1.9768,
+      "step": 9595
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.7104569247888928e-06,
+      "loss": 1.7844,
+      "step": 9600
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.698692788830608e-06,
+      "loss": 1.7483,
+      "step": 9605
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.6869286528723226e-06,
+      "loss": 1.5277,
+      "step": 9610
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.6751645169140373e-06,
+      "loss": 2.0711,
+      "step": 9615
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.6634003809557525e-06,
+      "loss": 1.8575,
+      "step": 9620
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.651636244997467e-06,
+      "loss": 1.9961,
+      "step": 9625
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.6398721090391823e-06,
+      "loss": 1.9177,
+      "step": 9630
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.628107973080897e-06,
+      "loss": 1.9108,
+      "step": 9635
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.616343837122612e-06,
+      "loss": 1.8047,
+      "step": 9640
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.604579701164327e-06,
+      "loss": 1.7475,
+      "step": 9645
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5928155652060416e-06,
+      "loss": 1.9709,
+      "step": 9650
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5810514292477568e-06,
+      "loss": 1.6721,
+      "step": 9655
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5692872932894715e-06,
+      "loss": 1.6876,
+      "step": 9660
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.557523157331186e-06,
+      "loss": 1.9196,
+      "step": 9665
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5457590213729013e-06,
+      "loss": 1.8111,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.533994885414616e-06,
+      "loss": 2.0223,
+      "step": 9675
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5222307494563308e-06,
+      "loss": 1.9981,
+      "step": 9680
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.5104666134980455e-06,
+      "loss": 1.7671,
+      "step": 9685
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.4987024775397606e-06,
+      "loss": 1.8583,
+      "step": 9690
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.4869383415814758e-06,
+      "loss": 1.6488,
+      "step": 9695
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.4751742056231905e-06,
+      "loss": 2.1679,
+      "step": 9700
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.4634100696649056e-06,
+      "loss": 1.6889,
+      "step": 9705
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.4516459337066203e-06,
+      "loss": 1.6899,
+      "step": 9710
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.439881797748335e-06,
+      "loss": 1.7223,
+      "step": 9715
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.4281176617900498e-06,
+      "loss": 2.1167,
+      "step": 9720
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.416353525831765e-06,
+      "loss": 1.7187,
+      "step": 9725
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.4045893898734796e-06,
+      "loss": 1.9707,
+      "step": 9730
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.3928252539151943e-06,
+      "loss": 1.922,
+      "step": 9735
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.3810611179569095e-06,
+      "loss": 1.823,
+      "step": 9740
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.369296981998624e-06,
+      "loss": 1.7315,
+      "step": 9745
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.357532846040339e-06,
+      "loss": 1.8811,
+      "step": 9750
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.345768710082054e-06,
+      "loss": 1.6516,
+      "step": 9755
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.3340045741237688e-06,
+      "loss": 1.7688,
+      "step": 9760
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.322240438165484e-06,
+      "loss": 1.6986,
+      "step": 9765
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.3104763022071986e-06,
+      "loss": 1.9485,
+      "step": 9770
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.2987121662489138e-06,
+      "loss": 1.9216,
+      "step": 9775
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.2869480302906285e-06,
+      "loss": 1.9966,
+      "step": 9780
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.275183894332343e-06,
+      "loss": 1.7666,
+      "step": 9785
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.2634197583740583e-06,
+      "loss": 2.0442,
+      "step": 9790
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.251655622415773e-06,
+      "loss": 1.83,
+      "step": 9795
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.2398914864574878e-06,
+      "loss": 1.8298,
+      "step": 9800
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.228127350499203e-06,
+      "loss": 1.807,
+      "step": 9805
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.2163632145409176e-06,
+      "loss": 1.7401,
+      "step": 9810
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.2045990785826323e-06,
+      "loss": 1.742,
+      "step": 9815
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1928349426243475e-06,
+      "loss": 2.0319,
+      "step": 9820
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.181070806666062e-06,
+      "loss": 1.7051,
+      "step": 9825
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1693066707077773e-06,
+      "loss": 1.688,
+      "step": 9830
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.157542534749492e-06,
+      "loss": 1.7206,
+      "step": 9835
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.145778398791207e-06,
+      "loss": 1.8728,
+      "step": 9840
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.134014262832922e-06,
+      "loss": 1.7658,
+      "step": 9845
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1222501268746366e-06,
+      "loss": 1.7693,
+      "step": 9850
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.1104859909163513e-06,
+      "loss": 1.8815,
+      "step": 9855
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0987218549580665e-06,
+      "loss": 1.6237,
+      "step": 9860
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.086957718999781e-06,
+      "loss": 1.544,
+      "step": 9865
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.075193583041496e-06,
+      "loss": 1.759,
+      "step": 9870
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.063429447083211e-06,
+      "loss": 2.0515,
+      "step": 9875
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0516653111249258e-06,
+      "loss": 1.7567,
+      "step": 9880
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.039901175166641e-06,
+      "loss": 1.7844,
+      "step": 9885
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0281370392083556e-06,
+      "loss": 1.9163,
+      "step": 9890
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0163729032500707e-06,
+      "loss": 2.0447,
+      "step": 9895
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0046087672917855e-06,
+      "loss": 2.074,
+      "step": 9900
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9928446313335e-06,
+      "loss": 1.7544,
+      "step": 9905
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9810804953752153e-06,
+      "loss": 1.7648,
+      "step": 9910
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.96931635941693e-06,
+      "loss": 1.9007,
+      "step": 9915
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9575522234586447e-06,
+      "loss": 2.205,
+      "step": 9920
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.94578808750036e-06,
+      "loss": 1.7543,
+      "step": 9925
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9340239515420746e-06,
+      "loss": 2.0118,
+      "step": 9930
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9222598155837893e-06,
+      "loss": 1.9625,
+      "step": 9935
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9104956796255045e-06,
+      "loss": 1.5751,
+      "step": 9940
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.8987315436672194e-06,
+      "loss": 1.8749,
+      "step": 9945
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.886967407708934e-06,
+      "loss": 2.0136,
+      "step": 9950
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.8752032717506488e-06,
+      "loss": 1.9296,
+      "step": 9955
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.863439135792364e-06,
+      "loss": 2.0822,
+      "step": 9960
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.8516749998340787e-06,
+      "loss": 2.1667,
+      "step": 9965
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.8399108638757936e-06,
+      "loss": 1.7356,
+      "step": 9970
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.8281467279175087e-06,
+      "loss": 1.9915,
+      "step": 9975
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.8163825919592235e-06,
+      "loss": 1.9713,
+      "step": 9980
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.8046184560009382e-06,
+      "loss": 1.6892,
+      "step": 9985
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7928543200426529e-06,
+      "loss": 1.8452,
+      "step": 9990
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.781090184084368e-06,
+      "loss": 1.5651,
+      "step": 9995
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.769326048126083e-06,
+      "loss": 1.9586,
+      "step": 10000
+    },
+    {
+      "epoch": 1.86,
+      "eval_accuracy": 0.8481397362327595,
+      "eval_loss": 2.0825693607330322,
+      "eval_runtime": 129.4866,
+      "eval_samples_per_second": 83.02,
+      "eval_steps_per_second": 10.379,
+      "step": 10000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7575619121677977e-06,
+      "loss": 2.3005,
+      "step": 10005
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7457977762095128e-06,
+      "loss": 1.8285,
+      "step": 10010
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7340336402512275e-06,
+      "loss": 1.9501,
+      "step": 10015
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7222695042929422e-06,
+      "loss": 1.8749,
+      "step": 10020
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7105053683346574e-06,
+      "loss": 1.9548,
+      "step": 10025
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.698741232376372e-06,
+      "loss": 1.9391,
+      "step": 10030
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.686977096418087e-06,
+      "loss": 1.8646,
+      "step": 10035
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6752129604598017e-06,
+      "loss": 1.6386,
+      "step": 10040
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6634488245015169e-06,
+      "loss": 1.9579,
+      "step": 10045
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6516846885432316e-06,
+      "loss": 1.8709,
+      "step": 10050
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6399205525849463e-06,
+      "loss": 1.7001,
+      "step": 10055
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6281564166266615e-06,
+      "loss": 1.8714,
+      "step": 10060
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6163922806683764e-06,
+      "loss": 1.8629,
+      "step": 10065
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.604628144710091e-06,
+      "loss": 1.6322,
+      "step": 10070
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.5928640087518062e-06,
+      "loss": 2.0094,
+      "step": 10075
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.581099872793521e-06,
+      "loss": 1.8591,
+      "step": 10080
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5693357368352357e-06,
+      "loss": 1.7914,
+      "step": 10085
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5575716008769506e-06,
+      "loss": 2.0819,
+      "step": 10090
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5458074649186655e-06,
+      "loss": 1.7338,
+      "step": 10095
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5340433289603805e-06,
+      "loss": 2.0772,
+      "step": 10100
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5222791930020952e-06,
+      "loss": 1.863,
+      "step": 10105
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5105150570438103e-06,
+      "loss": 1.8443,
+      "step": 10110
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.498750921085525e-06,
+      "loss": 1.7636,
+      "step": 10115
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.4869867851272397e-06,
+      "loss": 1.7533,
+      "step": 10120
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.4752226491689547e-06,
+      "loss": 1.7555,
+      "step": 10125
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.4634585132106696e-06,
+      "loss": 2.017,
+      "step": 10130
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.4516943772523845e-06,
+      "loss": 1.7163,
+      "step": 10135
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.4399302412940992e-06,
+      "loss": 1.9118,
+      "step": 10140
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.4281661053358142e-06,
+      "loss": 1.8934,
+      "step": 10145
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.416401969377529e-06,
+      "loss": 2.0631,
+      "step": 10150
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.404637833419244e-06,
+      "loss": 1.8838,
+      "step": 10155
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.392873697460959e-06,
+      "loss": 1.9034,
+      "step": 10160
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.3811095615026739e-06,
+      "loss": 1.6808,
+      "step": 10165
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.3693454255443886e-06,
+      "loss": 1.6575,
+      "step": 10170
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.3575812895861035e-06,
+      "loss": 1.4221,
+      "step": 10175
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.3458171536278182e-06,
+      "loss": 1.6231,
+      "step": 10180
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.3340530176695332e-06,
+      "loss": 1.8556,
+      "step": 10185
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.322288881711248e-06,
+      "loss": 1.83,
+      "step": 10190
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.310524745752963e-06,
+      "loss": 1.9004,
+      "step": 10195
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.298760609794678e-06,
+      "loss": 1.8312,
+      "step": 10200
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2869964738363927e-06,
+      "loss": 1.9444,
+      "step": 10205
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2752323378781076e-06,
+      "loss": 1.55,
+      "step": 10210
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2634682019198225e-06,
+      "loss": 1.9109,
+      "step": 10215
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2517040659615372e-06,
+      "loss": 1.7743,
+      "step": 10220
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2399399300032524e-06,
+      "loss": 1.9077,
+      "step": 10225
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.228175794044967e-06,
+      "loss": 1.7546,
+      "step": 10230
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.216411658086682e-06,
+      "loss": 2.1645,
+      "step": 10235
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.204647522128397e-06,
+      "loss": 2.1366,
+      "step": 10240
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1928833861701117e-06,
+      "loss": 1.6449,
+      "step": 10245
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1811192502118266e-06,
+      "loss": 1.9739,
+      "step": 10250
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1693551142535415e-06,
+      "loss": 1.6492,
+      "step": 10255
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1575909782952564e-06,
+      "loss": 1.9898,
+      "step": 10260
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1458268423369714e-06,
+      "loss": 1.8451,
+      "step": 10265
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.134062706378686e-06,
+      "loss": 2.1758,
+      "step": 10270
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.122298570420401e-06,
+      "loss": 2.0065,
+      "step": 10275
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.1105344344621157e-06,
+      "loss": 1.774,
+      "step": 10280
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0987702985038307e-06,
+      "loss": 1.9924,
+      "step": 10285
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0870061625455456e-06,
+      "loss": 1.848,
+      "step": 10290
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0752420265872605e-06,
+      "loss": 1.9898,
+      "step": 10295
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0634778906289754e-06,
+      "loss": 2.1111,
+      "step": 10300
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0517137546706902e-06,
+      "loss": 1.8931,
+      "step": 10305
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.039949618712405e-06,
+      "loss": 2.0328,
+      "step": 10310
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0281854827541198e-06,
+      "loss": 1.8924,
+      "step": 10315
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.016421346795835e-06,
+      "loss": 2.1472,
+      "step": 10320
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0046572108375499e-06,
+      "loss": 1.6744,
+      "step": 10325
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.928930748792646e-07,
+      "loss": 1.6584,
+      "step": 10330
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.811289389209795e-07,
+      "loss": 1.914,
+      "step": 10335
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.693648029626942e-07,
+      "loss": 2.0228,
+      "step": 10340
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.576006670044092e-07,
+      "loss": 1.6295,
+      "step": 10345
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.458365310461242e-07,
+      "loss": 1.91,
+      "step": 10350
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.340723950878389e-07,
+      "loss": 1.7157,
+      "step": 10355
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.223082591295538e-07,
+      "loss": 2.0007,
+      "step": 10360
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.105441231712687e-07,
+      "loss": 1.8112,
+      "step": 10365
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.987799872129836e-07,
+      "loss": 1.6855,
+      "step": 10370
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.870158512546985e-07,
+      "loss": 1.4989,
+      "step": 10375
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.752517152964133e-07,
+      "loss": 1.8539,
+      "step": 10380
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.634875793381283e-07,
+      "loss": 1.6471,
+      "step": 10385
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.517234433798431e-07,
+      "loss": 1.7335,
+      "step": 10390
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.39959307421558e-07,
+      "loss": 1.8246,
+      "step": 10395
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.281951714632729e-07,
+      "loss": 1.4332,
+      "step": 10400
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.164310355049877e-07,
+      "loss": 1.8263,
+      "step": 10405
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.046668995467027e-07,
+      "loss": 2.0113,
+      "step": 10410
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.929027635884174e-07,
+      "loss": 1.7151,
+      "step": 10415
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.811386276301323e-07,
+      "loss": 1.9404,
+      "step": 10420
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.693744916718472e-07,
+      "loss": 1.9382,
+      "step": 10425
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.576103557135621e-07,
+      "loss": 1.9345,
+      "step": 10430
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.45846219755277e-07,
+      "loss": 1.7227,
+      "step": 10435
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.340820837969918e-07,
+      "loss": 1.9156,
+      "step": 10440
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.223179478387068e-07,
+      "loss": 1.5789,
+      "step": 10445
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 7.105538118804216e-07,
+      "loss": 1.5983,
+      "step": 10450
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.987896759221364e-07,
+      "loss": 1.9769,
+      "step": 10455
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.870255399638513e-07,
+      "loss": 2.2449,
+      "step": 10460
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.752614040055662e-07,
+      "loss": 1.7278,
+      "step": 10465
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.634972680472811e-07,
+      "loss": 1.9207,
+      "step": 10470
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.51733132088996e-07,
+      "loss": 1.8335,
+      "step": 10475
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.399689961307108e-07,
+      "loss": 1.577,
+      "step": 10480
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.282048601724257e-07,
+      "loss": 1.6799,
+      "step": 10485
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.164407242141406e-07,
+      "loss": 1.8577,
+      "step": 10490
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 6.046765882558555e-07,
+      "loss": 2.0126,
+      "step": 10495
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.929124522975703e-07,
+      "loss": 1.6797,
+      "step": 10500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.811483163392853e-07,
+      "loss": 1.8692,
+      "step": 10505
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.693841803810001e-07,
+      "loss": 1.6246,
+      "step": 10510
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.576200444227149e-07,
+      "loss": 2.066,
+      "step": 10515
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.458559084644298e-07,
+      "loss": 1.7753,
+      "step": 10520
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.340917725061448e-07,
+      "loss": 2.0079,
+      "step": 10525
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.223276365478596e-07,
+      "loss": 1.9543,
+      "step": 10530
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.105635005895744e-07,
+      "loss": 1.9597,
+      "step": 10535
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.987993646312893e-07,
+      "loss": 1.7472,
+      "step": 10540
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.870352286730042e-07,
+      "loss": 1.871,
+      "step": 10545
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.7527109271471913e-07,
+      "loss": 1.7862,
+      "step": 10550
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.6350695675643395e-07,
+      "loss": 2.006,
+      "step": 10555
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.5174282079814883e-07,
+      "loss": 1.5212,
+      "step": 10560
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.399786848398637e-07,
+      "loss": 1.8134,
+      "step": 10565
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.282145488815785e-07,
+      "loss": 1.6593,
+      "step": 10570
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.164504129232934e-07,
+      "loss": 1.9933,
+      "step": 10575
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.0468627696500833e-07,
+      "loss": 2.0238,
+      "step": 10580
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.929221410067232e-07,
+      "loss": 1.8263,
+      "step": 10585
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.811580050484381e-07,
+      "loss": 1.9154,
+      "step": 10590
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.6939386909015295e-07,
+      "loss": 1.8784,
+      "step": 10595
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.5762973313186783e-07,
+      "loss": 2.0301,
+      "step": 10600
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.458655971735827e-07,
+      "loss": 1.9524,
+      "step": 10605
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.341014612152976e-07,
+      "loss": 1.7684,
+      "step": 10610
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.2233732525701245e-07,
+      "loss": 2.1926,
+      "step": 10615
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.1057318929872733e-07,
+      "loss": 1.5457,
+      "step": 10620
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.988090533404422e-07,
+      "loss": 1.8775,
+      "step": 10625
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.870449173821571e-07,
+      "loss": 2.1269,
+      "step": 10630
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.7528078142387195e-07,
+      "loss": 1.7646,
+      "step": 10635
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.635166454655868e-07,
+      "loss": 1.891,
+      "step": 10640
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.517525095073017e-07,
+      "loss": 1.9147,
+      "step": 10645
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.399883735490165e-07,
+      "loss": 1.9656,
+      "step": 10650
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.2822423759073145e-07,
+      "loss": 1.7951,
+      "step": 10655
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.164601016324463e-07,
+      "loss": 1.7507,
+      "step": 10660
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.046959656741612e-07,
+      "loss": 1.9079,
+      "step": 10665
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.9293182971587607e-07,
+      "loss": 2.0217,
+      "step": 10670
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.8116769375759092e-07,
+      "loss": 1.7537,
+      "step": 10675
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.694035577993058e-07,
+      "loss": 1.6031,
+      "step": 10680
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.5763942184102067e-07,
+      "loss": 1.7569,
+      "step": 10685
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.4587528588273555e-07,
+      "loss": 2.1558,
+      "step": 10690
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3411114992445045e-07,
+      "loss": 1.8888,
+      "step": 10695
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.223470139661653e-07,
+      "loss": 1.9908,
+      "step": 10700
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.1058287800788018e-07,
+      "loss": 1.6977,
+      "step": 10705
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.881874204959506e-08,
+      "loss": 1.7936,
+      "step": 10710
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 8.705460609130993e-08,
+      "loss": 1.5857,
+      "step": 10715
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.529047013302481e-08,
+      "loss": 1.9536,
+      "step": 10720
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.352633417473968e-08,
+      "loss": 1.6279,
+      "step": 10725
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.176219821645456e-08,
+      "loss": 1.9257,
+      "step": 10730
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.999806225816943e-08,
+      "loss": 1.91,
+      "step": 10735
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.8233926299884302e-08,
+      "loss": 2.0231,
+      "step": 10740
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6469790341599177e-08,
+      "loss": 1.8488,
+      "step": 10745
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.7056543833140506e-09,
+      "loss": 1.6943,
+      "step": 10750
+    },
+    {
+      "epoch": 2.0,
+      "step": 10752,
+      "total_flos": 3.18339467359488e+16,
+      "train_loss": 2.1184611846914603,
+      "train_runtime": 4663.4223,
+      "train_samples_per_second": 18.443,
+      "train_steps_per_second": 2.306
     }
   ],
-  "max_steps": 2586,
-  "num_train_epochs": 6,
-  "total_flos": 4.49946006847488e+16,
+  "max_steps": 10752,
+  "num_train_epochs": 2,
+  "total_flos": 3.18339467359488e+16,
   "trial_name": null,
   "trial_params": null
 }