diff --git "a/checkpoint-378/trainer_state.json" "b/checkpoint-378/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-378/trainer_state.json"
@@ -0,0 +1,5313 @@
+{
+  "best_metric": 0.9350641369819641,
+  "best_model_checkpoint": "llava-v1.5-7b-chess/checkpoint-254",
+  "epoch": 3.0,
+  "eval_steps": 1,
+  "global_step": 378,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.973544973544973e-05,
+      "loss": 1.9416,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 1.8246577978134155,
+      "eval_runtime": 11.1051,
+      "eval_samples_per_second": 5.133,
+      "eval_steps_per_second": 1.711,
+      "step": 1
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.947089947089946e-05,
+      "loss": 1.8254,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 1.7272577285766602,
+      "eval_runtime": 11.142,
+      "eval_samples_per_second": 5.116,
+      "eval_steps_per_second": 1.705,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.920634920634922e-05,
+      "loss": 1.726,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 1.630748987197876,
+      "eval_runtime": 11.2506,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 3
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.894179894179895e-05,
+      "loss": 1.5798,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 1.5509014129638672,
+      "eval_runtime": 11.2587,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 4
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.867724867724868e-05,
+      "loss": 1.4559,
+      "step": 5
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.487467885017395,
+      "eval_runtime": 11.3884,
+      "eval_samples_per_second": 5.005,
+      "eval_steps_per_second": 1.668,
+      "step": 5
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.841269841269841e-05,
+      "loss": 1.4914,
+      "step": 6
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 1.4393059015274048,
+      "eval_runtime": 11.2599,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 6
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.814814814814815e-05,
+      "loss": 1.4471,
+      "step": 7
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.4020471572875977,
+      "eval_runtime": 11.2524,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 7
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.78835978835979e-05,
+      "loss": 1.2306,
+      "step": 8
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.3740724325180054,
+      "eval_runtime": 11.2867,
+      "eval_samples_per_second": 5.05,
+      "eval_steps_per_second": 1.683,
+      "step": 8
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.761904761904762e-05,
+      "loss": 1.4704,
+      "step": 9
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 1.3502081632614136,
+      "eval_runtime": 11.1146,
+      "eval_samples_per_second": 5.128,
+      "eval_steps_per_second": 1.709,
+      "step": 9
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.735449735449735e-05,
+      "loss": 1.3146,
+      "step": 10
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 1.3297895193099976,
+      "eval_runtime": 11.1575,
+      "eval_samples_per_second": 5.109,
+      "eval_steps_per_second": 1.703,
+      "step": 10
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.70899470899471e-05,
+      "loss": 1.3019,
+      "step": 11
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 1.3097199201583862,
+      "eval_runtime": 11.2593,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 11
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.682539682539682e-05,
+      "loss": 1.3221,
+      "step": 12
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.2873374223709106,
+      "eval_runtime": 11.3312,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 12
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.656084656084657e-05,
+      "loss": 1.1772,
+      "step": 13
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.266847014427185,
+      "eval_runtime": 11.1617,
+      "eval_samples_per_second": 5.107,
+      "eval_steps_per_second": 1.702,
+      "step": 13
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.62962962962963e-05,
+      "loss": 1.3181,
+      "step": 14
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 1.2443625926971436,
+      "eval_runtime": 11.1487,
+      "eval_samples_per_second": 5.113,
+      "eval_steps_per_second": 1.704,
+      "step": 14
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.603174603174604e-05,
+      "loss": 1.3618,
+      "step": 15
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 1.2243196964263916,
+      "eval_runtime": 11.1938,
+      "eval_samples_per_second": 5.092,
+      "eval_steps_per_second": 1.697,
+      "step": 15
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.576719576719577e-05,
+      "loss": 1.1889,
+      "step": 16
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.2069342136383057,
+      "eval_runtime": 11.4694,
+      "eval_samples_per_second": 4.97,
+      "eval_steps_per_second": 1.657,
+      "step": 16
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.550264550264551e-05,
+      "loss": 1.1392,
+      "step": 17
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.1917818784713745,
+      "eval_runtime": 11.2495,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 17
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 1.0116,
+      "step": 18
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 1.1815506219863892,
+      "eval_runtime": 11.242,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 18
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.497354497354498e-05,
+      "loss": 1.1372,
+      "step": 19
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 1.1708626747131348,
+      "eval_runtime": 11.2192,
+      "eval_samples_per_second": 5.081,
+      "eval_steps_per_second": 1.694,
+      "step": 19
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.470899470899471e-05,
+      "loss": 1.1948,
+      "step": 20
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 1.1638435125350952,
+      "eval_runtime": 11.366,
+      "eval_samples_per_second": 5.015,
+      "eval_steps_per_second": 1.672,
+      "step": 20
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.444444444444444e-05,
+      "loss": 1.0286,
+      "step": 21
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.1577672958374023,
+      "eval_runtime": 11.1403,
+      "eval_samples_per_second": 5.117,
+      "eval_steps_per_second": 1.706,
+      "step": 21
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.417989417989419e-05,
+      "loss": 1.1499,
+      "step": 22
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.151979923248291,
+      "eval_runtime": 11.2244,
+      "eval_samples_per_second": 5.078,
+      "eval_steps_per_second": 1.693,
+      "step": 22
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.391534391534393e-05,
+      "loss": 1.1372,
+      "step": 23
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 1.144721269607544,
+      "eval_runtime": 11.2159,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 23
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.365079365079366e-05,
+      "loss": 1.1184,
+      "step": 24
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 1.1370524168014526,
+      "eval_runtime": 11.6226,
+      "eval_samples_per_second": 4.904,
+      "eval_steps_per_second": 1.635,
+      "step": 24
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.338624338624339e-05,
+      "loss": 0.9123,
+      "step": 25
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 1.130428671836853,
+      "eval_runtime": 11.5048,
+      "eval_samples_per_second": 4.954,
+      "eval_steps_per_second": 1.651,
+      "step": 25
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.312169312169312e-05,
+      "loss": 1.0534,
+      "step": 26
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 1.125217318534851,
+      "eval_runtime": 11.2392,
+      "eval_samples_per_second": 5.072,
+      "eval_steps_per_second": 1.691,
+      "step": 26
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.285714285714286e-05,
+      "loss": 0.9502,
+      "step": 27
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 1.118015170097351,
+      "eval_runtime": 11.3322,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 27
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.25925925925926e-05,
+      "loss": 1.0064,
+      "step": 28
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 1.1118570566177368,
+      "eval_runtime": 11.5282,
+      "eval_samples_per_second": 4.944,
+      "eval_steps_per_second": 1.648,
+      "step": 28
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.232804232804233e-05,
+      "loss": 1.1374,
+      "step": 29
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 1.1059463024139404,
+      "eval_runtime": 11.448,
+      "eval_samples_per_second": 4.979,
+      "eval_steps_per_second": 1.66,
+      "step": 29
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.206349206349206e-05,
+      "loss": 0.9704,
+      "step": 30
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.1008621454238892,
+      "eval_runtime": 11.2328,
+      "eval_samples_per_second": 5.074,
+      "eval_steps_per_second": 1.691,
+      "step": 30
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.17989417989418e-05,
+      "loss": 1.1065,
+      "step": 31
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0981618165969849,
+      "eval_runtime": 11.2834,
+      "eval_samples_per_second": 5.052,
+      "eval_steps_per_second": 1.684,
+      "step": 31
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.153439153439155e-05,
+      "loss": 1.0171,
+      "step": 32
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0968972444534302,
+      "eval_runtime": 11.3954,
+      "eval_samples_per_second": 5.002,
+      "eval_steps_per_second": 1.667,
+      "step": 32
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.126984126984128e-05,
+      "loss": 1.1673,
+      "step": 33
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.0943207740783691,
+      "eval_runtime": 11.2123,
+      "eval_samples_per_second": 5.084,
+      "eval_steps_per_second": 1.695,
+      "step": 33
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.1005291005291e-05,
+      "loss": 1.0381,
+      "step": 34
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 1.09196937084198,
+      "eval_runtime": 11.2142,
+      "eval_samples_per_second": 5.083,
+      "eval_steps_per_second": 1.694,
+      "step": 34
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.074074074074075e-05,
+      "loss": 1.0529,
+      "step": 35
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 1.091015338897705,
+      "eval_runtime": 11.2585,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 35
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 1.0849,
+      "step": 36
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.0885584354400635,
+      "eval_runtime": 11.3567,
+      "eval_samples_per_second": 5.019,
+      "eval_steps_per_second": 1.673,
+      "step": 36
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.021164021164022e-05,
+      "loss": 1.0472,
+      "step": 37
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.0872443914413452,
+      "eval_runtime": 11.2417,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 37
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.994708994708995e-05,
+      "loss": 1.0949,
+      "step": 38
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.0832560062408447,
+      "eval_runtime": 11.8516,
+      "eval_samples_per_second": 4.809,
+      "eval_steps_per_second": 1.603,
+      "step": 38
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.968253968253969e-05,
+      "loss": 1.0909,
+      "step": 39
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.0786099433898926,
+      "eval_runtime": 11.432,
+      "eval_samples_per_second": 4.986,
+      "eval_steps_per_second": 1.662,
+      "step": 39
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.941798941798942e-05,
+      "loss": 0.9484,
+      "step": 40
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 1.0725367069244385,
+      "eval_runtime": 11.4482,
+      "eval_samples_per_second": 4.979,
+      "eval_steps_per_second": 1.66,
+      "step": 40
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.915343915343915e-05,
+      "loss": 1.0805,
+      "step": 41
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.0665147304534912,
+      "eval_runtime": 11.259,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 41
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.9088,
+      "step": 42
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.0613280534744263,
+      "eval_runtime": 11.4272,
+      "eval_samples_per_second": 4.988,
+      "eval_steps_per_second": 1.663,
+      "step": 42
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 8.862433862433864e-05,
+      "loss": 0.91,
+      "step": 43
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 1.057435154914856,
+      "eval_runtime": 11.2452,
+      "eval_samples_per_second": 5.069,
+      "eval_steps_per_second": 1.69,
+      "step": 43
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 8.835978835978837e-05,
+      "loss": 1.0241,
+      "step": 44
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.0551300048828125,
+      "eval_runtime": 11.2229,
+      "eval_samples_per_second": 5.079,
+      "eval_steps_per_second": 1.693,
+      "step": 44
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 8.80952380952381e-05,
+      "loss": 1.0352,
+      "step": 45
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 1.0539908409118652,
+      "eval_runtime": 11.3835,
+      "eval_samples_per_second": 5.007,
+      "eval_steps_per_second": 1.669,
+      "step": 45
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 8.783068783068782e-05,
+      "loss": 1.0175,
+      "step": 46
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.0538028478622437,
+      "eval_runtime": 11.3312,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 46
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 8.756613756613758e-05,
+      "loss": 0.9668,
+      "step": 47
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.0534106492996216,
+      "eval_runtime": 11.2714,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 47
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.730158730158731e-05,
+      "loss": 1.0426,
+      "step": 48
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.052772045135498,
+      "eval_runtime": 11.2543,
+      "eval_samples_per_second": 5.065,
+      "eval_steps_per_second": 1.688,
+      "step": 48
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 8.703703703703704e-05,
+      "loss": 1.0044,
+      "step": 49
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.0515741109848022,
+      "eval_runtime": 11.3376,
+      "eval_samples_per_second": 5.028,
+      "eval_steps_per_second": 1.676,
+      "step": 49
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 8.677248677248677e-05,
+      "loss": 1.0367,
+      "step": 50
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.049968957901001,
+      "eval_runtime": 11.2567,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 50
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 8.650793650793651e-05,
+      "loss": 1.1123,
+      "step": 51
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.0470366477966309,
+      "eval_runtime": 11.1206,
+      "eval_samples_per_second": 5.126,
+      "eval_steps_per_second": 1.709,
+      "step": 51
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 8.624338624338625e-05,
+      "loss": 1.0706,
+      "step": 52
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.0442218780517578,
+      "eval_runtime": 11.2711,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 52
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 8.597883597883598e-05,
+      "loss": 0.9999,
+      "step": 53
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.0409353971481323,
+      "eval_runtime": 11.2516,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 53
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 1.0905,
+      "step": 54
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.0385661125183105,
+      "eval_runtime": 11.328,
+      "eval_samples_per_second": 5.032,
+      "eval_steps_per_second": 1.677,
+      "step": 54
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 8.544973544973546e-05,
+      "loss": 0.8978,
+      "step": 55
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.0356653928756714,
+      "eval_runtime": 11.3452,
+      "eval_samples_per_second": 5.024,
+      "eval_steps_per_second": 1.675,
+      "step": 55
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 8.518518518518518e-05,
+      "loss": 1.0951,
+      "step": 56
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.0337508916854858,
+      "eval_runtime": 11.2553,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 56
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 8.492063492063493e-05,
+      "loss": 0.978,
+      "step": 57
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.031426191329956,
+      "eval_runtime": 11.3396,
+      "eval_samples_per_second": 5.027,
+      "eval_steps_per_second": 1.676,
+      "step": 57
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 8.465608465608466e-05,
+      "loss": 0.9974,
+      "step": 58
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.0299843549728394,
+      "eval_runtime": 11.3892,
+      "eval_samples_per_second": 5.005,
+      "eval_steps_per_second": 1.668,
+      "step": 58
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 8.43915343915344e-05,
+      "loss": 0.9869,
+      "step": 59
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.0282464027404785,
+      "eval_runtime": 11.2739,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 59
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 8.412698412698413e-05,
+      "loss": 1.0325,
+      "step": 60
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.0266135931015015,
+      "eval_runtime": 11.2367,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 60
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 8.386243386243386e-05,
+      "loss": 1.0886,
+      "step": 61
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.0231937170028687,
+      "eval_runtime": 11.3652,
+      "eval_samples_per_second": 5.015,
+      "eval_steps_per_second": 1.672,
+      "step": 61
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 8.35978835978836e-05,
+      "loss": 1.1033,
+      "step": 62
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.0222442150115967,
+      "eval_runtime": 11.2493,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 62
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.0087,
+      "step": 63
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.020636796951294,
+      "eval_runtime": 11.2865,
+      "eval_samples_per_second": 5.05,
+      "eval_steps_per_second": 1.683,
+      "step": 63
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 8.306878306878307e-05,
+      "loss": 0.8476,
+      "step": 64
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.0187979936599731,
+      "eval_runtime": 11.289,
+      "eval_samples_per_second": 5.049,
+      "eval_steps_per_second": 1.683,
+      "step": 64
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 8.28042328042328e-05,
+      "loss": 0.9543,
+      "step": 65
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.0155143737792969,
+      "eval_runtime": 11.2374,
+      "eval_samples_per_second": 5.072,
+      "eval_steps_per_second": 1.691,
+      "step": 65
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 8.253968253968255e-05,
+      "loss": 0.933,
+      "step": 66
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.0131887197494507,
+      "eval_runtime": 11.2775,
+      "eval_samples_per_second": 5.054,
+      "eval_steps_per_second": 1.685,
+      "step": 66
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 8.227513227513229e-05,
+      "loss": 1.0186,
+      "step": 67
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.0090850591659546,
+      "eval_runtime": 11.3677,
+      "eval_samples_per_second": 5.014,
+      "eval_steps_per_second": 1.671,
+      "step": 67
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 8.201058201058202e-05,
+      "loss": 0.9425,
+      "step": 68
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.0056878328323364,
+      "eval_runtime": 11.0888,
+      "eval_samples_per_second": 5.14,
+      "eval_steps_per_second": 1.713,
+      "step": 68
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.174603174603175e-05,
+      "loss": 0.9943,
+      "step": 69
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.0025955438613892,
+      "eval_runtime": 11.2572,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 69
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 1.1287,
+      "step": 70
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.0005890130996704,
+      "eval_runtime": 11.2495,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 70
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.121693121693122e-05,
+      "loss": 1.0688,
+      "step": 71
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.9989040493965149,
+      "eval_runtime": 11.2643,
+      "eval_samples_per_second": 5.06,
+      "eval_steps_per_second": 1.687,
+      "step": 71
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 1.0818,
+      "step": 72
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.9982772469520569,
+      "eval_runtime": 11.121,
+      "eval_samples_per_second": 5.125,
+      "eval_steps_per_second": 1.708,
+      "step": 72
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.068783068783069e-05,
+      "loss": 0.9234,
+      "step": 73
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.9961081147193909,
+      "eval_runtime": 11.268,
+      "eval_samples_per_second": 5.059,
+      "eval_steps_per_second": 1.686,
+      "step": 73
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.042328042328042e-05,
+      "loss": 0.9169,
+      "step": 74
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.9951977729797363,
+      "eval_runtime": 11.3425,
+      "eval_samples_per_second": 5.025,
+      "eval_steps_per_second": 1.675,
+      "step": 74
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.015873015873016e-05,
+      "loss": 0.8732,
+      "step": 75
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.9947168827056885,
+      "eval_runtime": 11.4976,
+      "eval_samples_per_second": 4.958,
+      "eval_steps_per_second": 1.653,
+      "step": 75
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.989417989417989e-05,
+      "loss": 1.1112,
+      "step": 76
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.9937358498573303,
+      "eval_runtime": 11.2157,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 76
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.962962962962964e-05,
+      "loss": 1.0538,
+      "step": 77
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.992274284362793,
+      "eval_runtime": 11.3598,
+      "eval_samples_per_second": 5.018,
+      "eval_steps_per_second": 1.673,
+      "step": 77
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.936507936507937e-05,
+      "loss": 1.0636,
+      "step": 78
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.9908178448677063,
+      "eval_runtime": 11.0889,
+      "eval_samples_per_second": 5.14,
+      "eval_steps_per_second": 1.713,
+      "step": 78
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.910052910052911e-05,
+      "loss": 0.9079,
+      "step": 79
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.9905064105987549,
+      "eval_runtime": 11.1638,
+      "eval_samples_per_second": 5.106,
+      "eval_steps_per_second": 1.702,
+      "step": 79
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.883597883597884e-05,
+      "loss": 0.9104,
+      "step": 80
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.989557683467865,
+      "eval_runtime": 11.4363,
+      "eval_samples_per_second": 4.984,
+      "eval_steps_per_second": 1.661,
+      "step": 80
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.857142857142858e-05,
+      "loss": 0.9058,
+      "step": 81
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.9902251958847046,
+      "eval_runtime": 11.4861,
+      "eval_samples_per_second": 4.963,
+      "eval_steps_per_second": 1.654,
+      "step": 81
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.830687830687831e-05,
+      "loss": 0.9653,
+      "step": 82
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.9908925294876099,
+      "eval_runtime": 11.56,
+      "eval_samples_per_second": 4.931,
+      "eval_steps_per_second": 1.644,
+      "step": 82
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.804232804232805e-05,
+      "loss": 0.8773,
+      "step": 83
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.9919198751449585,
+      "eval_runtime": 11.6172,
+      "eval_samples_per_second": 4.906,
+      "eval_steps_per_second": 1.635,
+      "step": 83
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.7942,
+      "step": 84
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.9942153096199036,
+      "eval_runtime": 11.3584,
+      "eval_samples_per_second": 5.018,
+      "eval_steps_per_second": 1.673,
+      "step": 84
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.751322751322751e-05,
+      "loss": 0.894,
+      "step": 85
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.9940976500511169,
+      "eval_runtime": 11.4867,
+      "eval_samples_per_second": 4.962,
+      "eval_steps_per_second": 1.654,
+      "step": 85
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.724867724867725e-05,
+      "loss": 1.0039,
+      "step": 86
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.9931285977363586,
+      "eval_runtime": 11.171,
+      "eval_samples_per_second": 5.102,
+      "eval_steps_per_second": 1.701,
+      "step": 86
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.6984126984127e-05,
+      "loss": 1.0113,
+      "step": 87
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.9918901324272156,
+      "eval_runtime": 11.2632,
+      "eval_samples_per_second": 5.061,
+      "eval_steps_per_second": 1.687,
+      "step": 87
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 7.671957671957673e-05,
+      "loss": 1.1145,
+      "step": 88
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.9903886318206787,
+      "eval_runtime": 11.3871,
+      "eval_samples_per_second": 5.006,
+      "eval_steps_per_second": 1.669,
+      "step": 88
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 7.645502645502646e-05,
+      "loss": 0.8794,
+      "step": 89
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.9867441058158875,
+      "eval_runtime": 11.5122,
+      "eval_samples_per_second": 4.951,
+      "eval_steps_per_second": 1.65,
+      "step": 89
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 1.1085,
+      "step": 90
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.9836506843566895,
+      "eval_runtime": 11.2029,
+      "eval_samples_per_second": 5.088,
+      "eval_steps_per_second": 1.696,
+      "step": 90
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.592592592592593e-05,
+      "loss": 0.9683,
+      "step": 91
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.9819443225860596,
+      "eval_runtime": 11.4398,
+      "eval_samples_per_second": 4.983,
+      "eval_steps_per_second": 1.661,
+      "step": 91
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.566137566137567e-05,
+      "loss": 0.9378,
+      "step": 92
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.9801782369613647,
+      "eval_runtime": 11.2742,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 92
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 7.53968253968254e-05,
+      "loss": 1.0159,
+      "step": 93
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.9785143733024597,
+      "eval_runtime": 11.3247,
+      "eval_samples_per_second": 5.033,
+      "eval_steps_per_second": 1.678,
+      "step": 93
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.513227513227513e-05,
+      "loss": 0.8486,
+      "step": 94
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.9786610007286072,
+      "eval_runtime": 11.4473,
+      "eval_samples_per_second": 4.979,
+      "eval_steps_per_second": 1.66,
+      "step": 94
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 7.486772486772487e-05,
+      "loss": 1.0572,
+      "step": 95
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.9791378974914551,
+      "eval_runtime": 11.2435,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 95
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7.460317460317461e-05,
+      "loss": 0.9379,
+      "step": 96
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.9795575141906738,
+      "eval_runtime": 11.3264,
+      "eval_samples_per_second": 5.033,
+      "eval_steps_per_second": 1.678,
+      "step": 96
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 7.433862433862434e-05,
+      "loss": 0.8751,
+      "step": 97
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.9804602265357971,
+      "eval_runtime": 11.4754,
+      "eval_samples_per_second": 4.967,
+      "eval_steps_per_second": 1.656,
+      "step": 97
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 0.9252,
+      "step": 98
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.9810712933540344,
+      "eval_runtime": 11.2682,
+      "eval_samples_per_second": 5.058,
+      "eval_steps_per_second": 1.686,
+      "step": 98
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.380952380952382e-05,
+      "loss": 0.9204,
+      "step": 99
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.9811252951622009,
+      "eval_runtime": 11.2452,
+      "eval_samples_per_second": 5.069,
+      "eval_steps_per_second": 1.69,
+      "step": 99
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 7.354497354497355e-05,
+      "loss": 1.0881,
+      "step": 100
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.9820334315299988,
+      "eval_runtime": 11.2336,
+      "eval_samples_per_second": 5.074,
+      "eval_steps_per_second": 1.691,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 7.328042328042329e-05,
+      "loss": 1.0158,
+      "step": 101
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.9812192916870117,
+      "eval_runtime": 11.3196,
+      "eval_samples_per_second": 5.036,
+      "eval_steps_per_second": 1.679,
+      "step": 101
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 7.301587301587302e-05,
+      "loss": 0.9741,
+      "step": 102
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.9810476303100586,
+      "eval_runtime": 11.3145,
+      "eval_samples_per_second": 5.038,
+      "eval_steps_per_second": 1.679,
+      "step": 102
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.275132275132276e-05,
+      "loss": 0.7926,
+      "step": 103
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.9799998998641968,
+      "eval_runtime": 11.312,
+      "eval_samples_per_second": 5.039,
+      "eval_steps_per_second": 1.68,
+      "step": 103
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.248677248677249e-05,
+      "loss": 1.0417,
+      "step": 104
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.9790608882904053,
+      "eval_runtime": 11.3684,
+      "eval_samples_per_second": 5.014,
+      "eval_steps_per_second": 1.671,
+      "step": 104
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.222222222222222e-05,
+      "loss": 0.9738,
+      "step": 105
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.978194534778595,
+      "eval_runtime": 11.2366,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 105
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 7.195767195767196e-05,
+      "loss": 0.8303,
+      "step": 106
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.9778124690055847,
+      "eval_runtime": 11.4098,
+      "eval_samples_per_second": 4.996,
+      "eval_steps_per_second": 1.665,
+      "step": 106
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.16931216931217e-05,
+      "loss": 0.9291,
+      "step": 107
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.9771907925605774,
+      "eval_runtime": 11.2524,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 107
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9308,
+      "step": 108
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.9771039485931396,
+      "eval_runtime": 11.2551,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 108
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.116402116402116e-05,
+      "loss": 1.1245,
+      "step": 109
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.9759649038314819,
+      "eval_runtime": 11.2261,
+      "eval_samples_per_second": 5.077,
+      "eval_steps_per_second": 1.692,
+      "step": 109
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.089947089947089e-05,
+      "loss": 0.8853,
+      "step": 110
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.9750571250915527,
+      "eval_runtime": 11.2982,
+      "eval_samples_per_second": 5.045,
+      "eval_steps_per_second": 1.682,
+      "step": 110
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.063492063492065e-05,
+      "loss": 0.9007,
+      "step": 111
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.9742787480354309,
+      "eval_runtime": 11.564,
+      "eval_samples_per_second": 4.929,
+      "eval_steps_per_second": 1.643,
+      "step": 111
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 7.037037037037038e-05,
+      "loss": 0.8941,
+      "step": 112
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.9725750684738159,
+      "eval_runtime": 11.2109,
+      "eval_samples_per_second": 5.084,
+      "eval_steps_per_second": 1.695,
+      "step": 112
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.010582010582011e-05,
+      "loss": 1.0456,
+      "step": 113
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.970888614654541,
+      "eval_runtime": 11.2158,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 113
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 6.984126984126984e-05,
+      "loss": 1.0988,
+      "step": 114
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.9695295095443726,
+      "eval_runtime": 11.1973,
+      "eval_samples_per_second": 5.091,
+      "eval_steps_per_second": 1.697,
+      "step": 114
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.957671957671958e-05,
+      "loss": 0.9967,
+      "step": 115
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.9678014516830444,
+      "eval_runtime": 11.1632,
+      "eval_samples_per_second": 5.106,
+      "eval_steps_per_second": 1.702,
+      "step": 115
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.931216931216932e-05,
+      "loss": 0.9641,
+      "step": 116
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.9670316576957703,
+      "eval_runtime": 11.1753,
+      "eval_samples_per_second": 5.101,
+      "eval_steps_per_second": 1.7,
+      "step": 116
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 6.904761904761905e-05,
+      "loss": 0.8927,
+      "step": 117
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.965987503528595,
+      "eval_runtime": 11.4765,
+      "eval_samples_per_second": 4.967,
+      "eval_steps_per_second": 1.656,
+      "step": 117
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 6.878306878306878e-05,
+      "loss": 1.0487,
+      "step": 118
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.9654714465141296,
+      "eval_runtime": 11.3628,
+      "eval_samples_per_second": 5.016,
+      "eval_steps_per_second": 1.672,
+      "step": 118
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 6.851851851851852e-05,
+      "loss": 0.9565,
+      "step": 119
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.9648447036743164,
+      "eval_runtime": 11.2817,
+      "eval_samples_per_second": 5.052,
+      "eval_steps_per_second": 1.684,
+      "step": 119
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 6.825396825396825e-05,
+      "loss": 1.0075,
+      "step": 120
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.96305912733078,
+      "eval_runtime": 11.1896,
+      "eval_samples_per_second": 5.094,
+      "eval_steps_per_second": 1.698,
+      "step": 120
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.7989417989418e-05,
+      "loss": 0.8491,
+      "step": 121
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.9615995287895203,
+      "eval_runtime": 11.2744,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 121
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.772486772486773e-05,
+      "loss": 0.9697,
+      "step": 122
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.9624088406562805,
+      "eval_runtime": 11.2628,
+      "eval_samples_per_second": 5.061,
+      "eval_steps_per_second": 1.687,
+      "step": 122
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.746031746031747e-05,
+      "loss": 0.9846,
+      "step": 123
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.9619339108467102,
+      "eval_runtime": 11.1615,
+      "eval_samples_per_second": 5.107,
+      "eval_steps_per_second": 1.702,
+      "step": 123
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.71957671957672e-05,
+      "loss": 0.9166,
+      "step": 124
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.9619455933570862,
+      "eval_runtime": 11.1285,
+      "eval_samples_per_second": 5.122,
+      "eval_steps_per_second": 1.707,
+      "step": 124
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.693121693121693e-05,
+      "loss": 0.8547,
+      "step": 125
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.9620954990386963,
+      "eval_runtime": 11.2673,
+      "eval_samples_per_second": 5.059,
+      "eval_steps_per_second": 1.686,
+      "step": 125
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.9433,
+      "step": 126
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9631038308143616,
+      "eval_runtime": 11.2502,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 126
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6.640211640211641e-05,
+      "loss": 0.7734,
+      "step": 127
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.9639068841934204,
+      "eval_runtime": 11.3228,
+      "eval_samples_per_second": 5.034,
+      "eval_steps_per_second": 1.678,
+      "step": 127
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.613756613756614e-05,
+      "loss": 0.8896,
+      "step": 128
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 0.9645029306411743,
+      "eval_runtime": 11.5577,
+      "eval_samples_per_second": 4.932,
+      "eval_steps_per_second": 1.644,
+      "step": 128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 6.587301587301587e-05,
+      "loss": 0.8125,
+      "step": 129
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 0.9654971361160278,
+      "eval_runtime": 11.3486,
+      "eval_samples_per_second": 5.023,
+      "eval_steps_per_second": 1.674,
+      "step": 129
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 6.560846560846561e-05,
+      "loss": 0.9779,
+      "step": 130
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 0.9669473767280579,
+      "eval_runtime": 11.2418,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 130
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.534391534391536e-05,
+      "loss": 0.8023,
+      "step": 131
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 0.9675314426422119,
+      "eval_runtime": 11.3422,
+      "eval_samples_per_second": 5.025,
+      "eval_steps_per_second": 1.675,
+      "step": 131
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.507936507936509e-05,
+      "loss": 0.9203,
+      "step": 132
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 0.9673454165458679,
+      "eval_runtime": 11.2406,
+      "eval_samples_per_second": 5.071,
+      "eval_steps_per_second": 1.69,
+      "step": 132
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.481481481481482e-05,
+      "loss": 0.9531,
+      "step": 133
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 0.9672001004219055,
+      "eval_runtime": 11.38,
+      "eval_samples_per_second": 5.009,
+      "eval_steps_per_second": 1.67,
+      "step": 133
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.455026455026454e-05,
+      "loss": 0.7756,
+      "step": 134
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 0.9666074514389038,
+      "eval_runtime": 11.1727,
+      "eval_samples_per_second": 5.102,
+      "eval_steps_per_second": 1.701,
+      "step": 134
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.428571428571429e-05,
+      "loss": 0.7395,
+      "step": 135
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 0.9651405215263367,
+      "eval_runtime": 11.3323,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 135
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.402116402116403e-05,
+      "loss": 0.8366,
+      "step": 136
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 0.965033769607544,
+      "eval_runtime": 11.2585,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 136
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.375661375661376e-05,
+      "loss": 0.9231,
+      "step": 137
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 0.9626837372779846,
+      "eval_runtime": 11.1354,
+      "eval_samples_per_second": 5.119,
+      "eval_steps_per_second": 1.706,
+      "step": 137
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.349206349206349e-05,
+      "loss": 0.8717,
+      "step": 138
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 0.9614263772964478,
+      "eval_runtime": 11.2272,
+      "eval_samples_per_second": 5.077,
+      "eval_steps_per_second": 1.692,
+      "step": 138
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.322751322751323e-05,
+      "loss": 0.8157,
+      "step": 139
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 0.9611877202987671,
+      "eval_runtime": 11.3481,
+      "eval_samples_per_second": 5.023,
+      "eval_steps_per_second": 1.674,
+      "step": 139
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.296296296296296e-05,
+      "loss": 0.8617,
+      "step": 140
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 0.9607529044151306,
+      "eval_runtime": 11.2319,
+      "eval_samples_per_second": 5.075,
+      "eval_steps_per_second": 1.692,
+      "step": 140
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.26984126984127e-05,
+      "loss": 0.7694,
+      "step": 141
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 0.961409866809845,
+      "eval_runtime": 11.165,
+      "eval_samples_per_second": 5.105,
+      "eval_steps_per_second": 1.702,
+      "step": 141
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.243386243386243e-05,
+      "loss": 0.8626,
+      "step": 142
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 0.9628212451934814,
+      "eval_runtime": 11.2138,
+      "eval_samples_per_second": 5.083,
+      "eval_steps_per_second": 1.694,
+      "step": 142
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.216931216931218e-05,
+      "loss": 0.909,
+      "step": 143
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 0.9622757434844971,
+      "eval_runtime": 11.3331,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 143
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8004,
+      "step": 144
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 0.9607929587364197,
+      "eval_runtime": 11.2614,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 144
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.164021164021165e-05,
+      "loss": 0.8374,
+      "step": 145
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 0.9604395627975464,
+      "eval_runtime": 11.2681,
+      "eval_samples_per_second": 5.059,
+      "eval_steps_per_second": 1.686,
+      "step": 145
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.137566137566138e-05,
+      "loss": 0.7996,
+      "step": 146
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 0.9591620564460754,
+      "eval_runtime": 11.2545,
+      "eval_samples_per_second": 5.065,
+      "eval_steps_per_second": 1.688,
+      "step": 146
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.111111111111112e-05,
+      "loss": 0.8048,
+      "step": 147
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 0.9589068293571472,
+      "eval_runtime": 11.3293,
+      "eval_samples_per_second": 5.031,
+      "eval_steps_per_second": 1.677,
+      "step": 147
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.084656084656085e-05,
+      "loss": 0.8036,
+      "step": 148
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 0.9590967297554016,
+      "eval_runtime": 11.2762,
+      "eval_samples_per_second": 5.055,
+      "eval_steps_per_second": 1.685,
+      "step": 148
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.0582010582010586e-05,
+      "loss": 0.8703,
+      "step": 149
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 0.9593470096588135,
+      "eval_runtime": 11.2342,
+      "eval_samples_per_second": 5.074,
+      "eval_steps_per_second": 1.691,
+      "step": 149
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.0317460317460316e-05,
+      "loss": 0.7818,
+      "step": 150
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 0.9604654908180237,
+      "eval_runtime": 11.2,
+      "eval_samples_per_second": 5.089,
+      "eval_steps_per_second": 1.696,
+      "step": 150
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.005291005291006e-05,
+      "loss": 0.9607,
+      "step": 151
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 0.9621202349662781,
+      "eval_runtime": 11.1607,
+      "eval_samples_per_second": 5.107,
+      "eval_steps_per_second": 1.702,
+      "step": 151
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.9788359788359794e-05,
+      "loss": 0.7956,
+      "step": 152
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 0.9638999104499817,
+      "eval_runtime": 11.25,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 152
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.9523809523809524e-05,
+      "loss": 0.7905,
+      "step": 153
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 0.9654872417449951,
+      "eval_runtime": 11.2628,
+      "eval_samples_per_second": 5.061,
+      "eval_steps_per_second": 1.687,
+      "step": 153
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.8222,
+      "step": 154
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 0.9671238660812378,
+      "eval_runtime": 11.2352,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 154
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.8994708994709e-05,
+      "loss": 0.7141,
+      "step": 155
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 0.9681822657585144,
+      "eval_runtime": 11.4061,
+      "eval_samples_per_second": 4.997,
+      "eval_steps_per_second": 1.666,
+      "step": 155
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.873015873015873e-05,
+      "loss": 0.7559,
+      "step": 156
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 0.9698037505149841,
+      "eval_runtime": 11.2595,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 156
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.846560846560847e-05,
+      "loss": 0.7906,
+      "step": 157
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 0.9727188348770142,
+      "eval_runtime": 11.2154,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 157
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.82010582010582e-05,
+      "loss": 0.8808,
+      "step": 158
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 0.9737605452537537,
+      "eval_runtime": 11.2619,
+      "eval_samples_per_second": 5.061,
+      "eval_steps_per_second": 1.687,
+      "step": 158
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.793650793650795e-05,
+      "loss": 0.7659,
+      "step": 159
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.9749085307121277,
+      "eval_runtime": 11.2907,
+      "eval_samples_per_second": 5.048,
+      "eval_steps_per_second": 1.683,
+      "step": 159
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.7671957671957676e-05,
+      "loss": 0.746,
+      "step": 160
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 0.9759730100631714,
+      "eval_runtime": 11.2853,
+      "eval_samples_per_second": 5.051,
+      "eval_steps_per_second": 1.684,
+      "step": 160
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.740740740740741e-05,
+      "loss": 0.7761,
+      "step": 161
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 0.9752081632614136,
+      "eval_runtime": 11.2755,
+      "eval_samples_per_second": 5.055,
+      "eval_steps_per_second": 1.685,
+      "step": 161
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.8106,
+      "step": 162
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 0.9742076396942139,
+      "eval_runtime": 11.2173,
+      "eval_samples_per_second": 5.081,
+      "eval_steps_per_second": 1.694,
+      "step": 162
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.6878306878306885e-05,
+      "loss": 1.0345,
+      "step": 163
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 0.971428632736206,
+      "eval_runtime": 11.2646,
+      "eval_samples_per_second": 5.06,
+      "eval_steps_per_second": 1.687,
+      "step": 163
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.661375661375662e-05,
+      "loss": 0.8811,
+      "step": 164
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 0.9710219502449036,
+      "eval_runtime": 11.2704,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 164
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.634920634920635e-05,
+      "loss": 0.8288,
+      "step": 165
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 0.9695227742195129,
+      "eval_runtime": 11.498,
+      "eval_samples_per_second": 4.957,
+      "eval_steps_per_second": 1.652,
+      "step": 165
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.6084656084656086e-05,
+      "loss": 0.8506,
+      "step": 166
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 0.9673130512237549,
+      "eval_runtime": 11.2551,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 166
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.582010582010583e-05,
+      "loss": 0.8069,
+      "step": 167
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 0.9665858745574951,
+      "eval_runtime": 11.15,
+      "eval_samples_per_second": 5.112,
+      "eval_steps_per_second": 1.704,
+      "step": 167
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.9651,
+      "step": 168
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 0.96519535779953,
+      "eval_runtime": 11.3608,
+      "eval_samples_per_second": 5.017,
+      "eval_steps_per_second": 1.672,
+      "step": 168
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5291005291005294e-05,
+      "loss": 0.9814,
+      "step": 169
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 0.9634275436401367,
+      "eval_runtime": 11.1336,
+      "eval_samples_per_second": 5.12,
+      "eval_steps_per_second": 1.707,
+      "step": 169
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5026455026455024e-05,
+      "loss": 0.854,
+      "step": 170
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 0.9614959359169006,
+      "eval_runtime": 11.3353,
+      "eval_samples_per_second": 5.029,
+      "eval_steps_per_second": 1.676,
+      "step": 170
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.4761904761904766e-05,
+      "loss": 0.8542,
+      "step": 171
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 0.9592831134796143,
+      "eval_runtime": 11.3728,
+      "eval_samples_per_second": 5.012,
+      "eval_steps_per_second": 1.671,
+      "step": 171
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.44973544973545e-05,
+      "loss": 0.8012,
+      "step": 172
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.9572239518165588,
+      "eval_runtime": 11.1863,
+      "eval_samples_per_second": 5.096,
+      "eval_steps_per_second": 1.699,
+      "step": 172
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.423280423280423e-05,
+      "loss": 0.975,
+      "step": 173
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.9551581144332886,
+      "eval_runtime": 11.2719,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 173
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.396825396825397e-05,
+      "loss": 0.7755,
+      "step": 174
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 0.9545064568519592,
+      "eval_runtime": 11.5962,
+      "eval_samples_per_second": 4.915,
+      "eval_steps_per_second": 1.638,
+      "step": 174
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.370370370370371e-05,
+      "loss": 0.8363,
+      "step": 175
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 0.9535031914710999,
+      "eval_runtime": 11.4446,
+      "eval_samples_per_second": 4.98,
+      "eval_steps_per_second": 1.66,
+      "step": 175
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.343915343915345e-05,
+      "loss": 0.8024,
+      "step": 176
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 0.9528553485870361,
+      "eval_runtime": 11.1519,
+      "eval_samples_per_second": 5.111,
+      "eval_steps_per_second": 1.704,
+      "step": 176
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.3174603174603176e-05,
+      "loss": 0.783,
+      "step": 177
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 0.9524054527282715,
+      "eval_runtime": 11.2811,
+      "eval_samples_per_second": 5.053,
+      "eval_steps_per_second": 1.684,
+      "step": 177
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.291005291005291e-05,
+      "loss": 0.9196,
+      "step": 178
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 0.9519273042678833,
+      "eval_runtime": 11.2667,
+      "eval_samples_per_second": 5.059,
+      "eval_steps_per_second": 1.686,
+      "step": 178
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.2645502645502655e-05,
+      "loss": 0.7972,
+      "step": 179
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 0.9521083235740662,
+      "eval_runtime": 11.2333,
+      "eval_samples_per_second": 5.074,
+      "eval_steps_per_second": 1.691,
+      "step": 179
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.713,
+      "step": 180
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 0.9525256752967834,
+      "eval_runtime": 11.272,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 180
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.211640211640212e-05,
+      "loss": 0.8934,
+      "step": 181
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 0.9519696831703186,
+      "eval_runtime": 11.2602,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 181
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.185185185185185e-05,
+      "loss": 0.8072,
+      "step": 182
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 0.9514541625976562,
+      "eval_runtime": 11.3432,
+      "eval_samples_per_second": 5.025,
+      "eval_steps_per_second": 1.675,
+      "step": 182
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.158730158730159e-05,
+      "loss": 0.8216,
+      "step": 183
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 0.9507673382759094,
+      "eval_runtime": 11.1823,
+      "eval_samples_per_second": 5.097,
+      "eval_steps_per_second": 1.699,
+      "step": 183
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.132275132275133e-05,
+      "loss": 0.7242,
+      "step": 184
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 0.950447142124176,
+      "eval_runtime": 11.1002,
+      "eval_samples_per_second": 5.135,
+      "eval_steps_per_second": 1.712,
+      "step": 184
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.105820105820106e-05,
+      "loss": 0.8381,
+      "step": 185
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 0.9501126408576965,
+      "eval_runtime": 11.3125,
+      "eval_samples_per_second": 5.039,
+      "eval_steps_per_second": 1.68,
+      "step": 185
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 0.9721,
+      "step": 186
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.9495165348052979,
+      "eval_runtime": 11.2634,
+      "eval_samples_per_second": 5.061,
+      "eval_steps_per_second": 1.687,
+      "step": 186
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.052910052910054e-05,
+      "loss": 0.9755,
+      "step": 187
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.9505437016487122,
+      "eval_runtime": 11.1189,
+      "eval_samples_per_second": 5.126,
+      "eval_steps_per_second": 1.709,
+      "step": 187
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.026455026455027e-05,
+      "loss": 0.7596,
+      "step": 188
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 0.9504641890525818,
+      "eval_runtime": 11.2688,
+      "eval_samples_per_second": 5.058,
+      "eval_steps_per_second": 1.686,
+      "step": 188
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.8127,
+      "step": 189
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 0.9504896998405457,
+      "eval_runtime": 11.3596,
+      "eval_samples_per_second": 5.018,
+      "eval_steps_per_second": 1.673,
+      "step": 189
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.973544973544973e-05,
+      "loss": 0.7999,
+      "step": 190
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.9504294991493225,
+      "eval_runtime": 11.6748,
+      "eval_samples_per_second": 4.882,
+      "eval_steps_per_second": 1.627,
+      "step": 190
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9470899470899475e-05,
+      "loss": 0.8556,
+      "step": 191
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 0.950935423374176,
+      "eval_runtime": 11.4816,
+      "eval_samples_per_second": 4.964,
+      "eval_steps_per_second": 1.655,
+      "step": 191
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 0.7262,
+      "step": 192
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 0.9509056210517883,
+      "eval_runtime": 11.4591,
+      "eval_samples_per_second": 4.974,
+      "eval_steps_per_second": 1.658,
+      "step": 192
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.894179894179895e-05,
+      "loss": 0.7748,
+      "step": 193
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 0.9506821632385254,
+      "eval_runtime": 11.6655,
+      "eval_samples_per_second": 4.886,
+      "eval_steps_per_second": 1.629,
+      "step": 193
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.8677248677248676e-05,
+      "loss": 0.7988,
+      "step": 194
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 0.9508106112480164,
+      "eval_runtime": 11.2436,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 194
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.9548,
+      "step": 195
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 0.9502134323120117,
+      "eval_runtime": 11.2836,
+      "eval_samples_per_second": 5.052,
+      "eval_steps_per_second": 1.684,
+      "step": 195
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.8198,
+      "step": 196
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 0.9500845670700073,
+      "eval_runtime": 11.2561,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 196
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.7883597883597884e-05,
+      "loss": 0.8214,
+      "step": 197
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 0.9501683712005615,
+      "eval_runtime": 11.2431,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 197
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.8485,
+      "step": 198
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 0.9491482973098755,
+      "eval_runtime": 11.4349,
+      "eval_samples_per_second": 4.985,
+      "eval_steps_per_second": 1.662,
+      "step": 198
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7354497354497356e-05,
+      "loss": 0.8415,
+      "step": 199
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 0.9494752883911133,
+      "eval_runtime": 11.45,
+      "eval_samples_per_second": 4.978,
+      "eval_steps_per_second": 1.659,
+      "step": 199
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.708994708994709e-05,
+      "loss": 0.789,
+      "step": 200
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 0.9499990940093994,
+      "eval_runtime": 11.1948,
+      "eval_samples_per_second": 5.092,
+      "eval_steps_per_second": 1.697,
+      "step": 200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.7414,
+      "step": 201
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 0.950614869594574,
+      "eval_runtime": 11.2016,
+      "eval_samples_per_second": 5.089,
+      "eval_steps_per_second": 1.696,
+      "step": 201
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.656084656084656e-05,
+      "loss": 0.9238,
+      "step": 202
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 0.9510459303855896,
+      "eval_runtime": 11.2692,
+      "eval_samples_per_second": 5.058,
+      "eval_steps_per_second": 1.686,
+      "step": 202
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.7716,
+      "step": 203
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 0.9512655138969421,
+      "eval_runtime": 11.25,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 203
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.78,
+      "step": 204
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 0.9526160955429077,
+      "eval_runtime": 11.2309,
+      "eval_samples_per_second": 5.075,
+      "eval_steps_per_second": 1.692,
+      "step": 204
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.576719576719577e-05,
+      "loss": 0.8149,
+      "step": 205
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 0.9529213309288025,
+      "eval_runtime": 11.4627,
+      "eval_samples_per_second": 4.973,
+      "eval_steps_per_second": 1.658,
+      "step": 205
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.55026455026455e-05,
+      "loss": 0.9315,
+      "step": 206
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 0.9526896476745605,
+      "eval_runtime": 11.349,
+      "eval_samples_per_second": 5.022,
+      "eval_steps_per_second": 1.674,
+      "step": 206
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.8562,
+      "step": 207
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.9522102475166321,
+      "eval_runtime": 11.2689,
+      "eval_samples_per_second": 5.058,
+      "eval_steps_per_second": 1.686,
+      "step": 207
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.4973544973544974e-05,
+      "loss": 0.6728,
+      "step": 208
+    },
+    {
+      "epoch": 1.65,
+      "eval_loss": 0.9524990320205688,
+      "eval_runtime": 11.3464,
+      "eval_samples_per_second": 5.024,
+      "eval_steps_per_second": 1.675,
+      "step": 208
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.470899470899471e-05,
+      "loss": 0.8107,
+      "step": 209
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 0.9504749178886414,
+      "eval_runtime": 11.3852,
+      "eval_samples_per_second": 5.006,
+      "eval_steps_per_second": 1.669,
+      "step": 209
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.7363,
+      "step": 210
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 0.9504967927932739,
+      "eval_runtime": 11.2458,
+      "eval_samples_per_second": 5.069,
+      "eval_steps_per_second": 1.69,
+      "step": 210
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.417989417989418e-05,
+      "loss": 0.8278,
+      "step": 211
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 0.9501346349716187,
+      "eval_runtime": 11.3631,
+      "eval_samples_per_second": 5.016,
+      "eval_steps_per_second": 1.672,
+      "step": 211
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.391534391534391e-05,
+      "loss": 0.7827,
+      "step": 212
+    },
+    {
+      "epoch": 1.68,
+      "eval_loss": 0.9484946727752686,
+      "eval_runtime": 11.3608,
+      "eval_samples_per_second": 5.017,
+      "eval_steps_per_second": 1.672,
+      "step": 212
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.6884,
+      "step": 213
+    },
+    {
+      "epoch": 1.69,
+      "eval_loss": 0.9481523633003235,
+      "eval_runtime": 11.3269,
+      "eval_samples_per_second": 5.032,
+      "eval_steps_per_second": 1.677,
+      "step": 213
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.3386243386243384e-05,
+      "loss": 0.7418,
+      "step": 214
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.9482589960098267,
+      "eval_runtime": 11.3957,
+      "eval_samples_per_second": 5.002,
+      "eval_steps_per_second": 1.667,
+      "step": 214
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.312169312169313e-05,
+      "loss": 0.8012,
+      "step": 215
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 0.9473200440406799,
+      "eval_runtime": 11.4932,
+      "eval_samples_per_second": 4.959,
+      "eval_steps_per_second": 1.653,
+      "step": 215
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.7137,
+      "step": 216
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 0.9463203549385071,
+      "eval_runtime": 11.442,
+      "eval_samples_per_second": 4.982,
+      "eval_steps_per_second": 1.661,
+      "step": 216
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.8202,
+      "step": 217
+    },
+    {
+      "epoch": 1.72,
+      "eval_loss": 0.9461511969566345,
+      "eval_runtime": 11.4253,
+      "eval_samples_per_second": 4.989,
+      "eval_steps_per_second": 1.663,
+      "step": 217
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.232804232804233e-05,
+      "loss": 0.7196,
+      "step": 218
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 0.9457988142967224,
+      "eval_runtime": 11.4121,
+      "eval_samples_per_second": 4.995,
+      "eval_steps_per_second": 1.665,
+      "step": 218
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.8195,
+      "step": 219
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 0.9461942911148071,
+      "eval_runtime": 11.5171,
+      "eval_samples_per_second": 4.949,
+      "eval_steps_per_second": 1.65,
+      "step": 219
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.17989417989418e-05,
+      "loss": 0.8813,
+      "step": 220
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.9455254673957825,
+      "eval_runtime": 11.3423,
+      "eval_samples_per_second": 5.025,
+      "eval_steps_per_second": 1.675,
+      "step": 220
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.153439153439154e-05,
+      "loss": 0.7291,
+      "step": 221
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.9450626969337463,
+      "eval_runtime": 11.3241,
+      "eval_samples_per_second": 5.034,
+      "eval_steps_per_second": 1.678,
+      "step": 221
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 0.8033,
+      "step": 222
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.943817675113678,
+      "eval_runtime": 11.236,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 222
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.100529100529101e-05,
+      "loss": 0.7453,
+      "step": 223
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 0.9435564279556274,
+      "eval_runtime": 11.3459,
+      "eval_samples_per_second": 5.024,
+      "eval_steps_per_second": 1.675,
+      "step": 223
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.7112,
+      "step": 224
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 0.943976104259491,
+      "eval_runtime": 11.3955,
+      "eval_samples_per_second": 5.002,
+      "eval_steps_per_second": 1.667,
+      "step": 224
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.817,
+      "step": 225
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 0.9434075355529785,
+      "eval_runtime": 11.2349,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 225
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.021164021164021e-05,
+      "loss": 0.7895,
+      "step": 226
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 0.9428831338882446,
+      "eval_runtime": 11.3622,
+      "eval_samples_per_second": 5.017,
+      "eval_steps_per_second": 1.672,
+      "step": 226
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.9947089947089946e-05,
+      "loss": 0.7245,
+      "step": 227
+    },
+    {
+      "epoch": 1.8,
+      "eval_loss": 0.9428912401199341,
+      "eval_runtime": 11.2865,
+      "eval_samples_per_second": 5.05,
+      "eval_steps_per_second": 1.683,
+      "step": 227
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.7244,
+      "step": 228
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 0.9429775476455688,
+      "eval_runtime": 11.2166,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 228
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.941798941798942e-05,
+      "loss": 0.7241,
+      "step": 229
+    },
+    {
+      "epoch": 1.82,
+      "eval_loss": 0.943901002407074,
+      "eval_runtime": 11.2593,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 229
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.9153439153439155e-05,
+      "loss": 0.7879,
+      "step": 230
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 0.9442031383514404,
+      "eval_runtime": 11.3453,
+      "eval_samples_per_second": 5.024,
+      "eval_steps_per_second": 1.675,
+      "step": 230
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 1.0106,
+      "step": 231
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 0.944178581237793,
+      "eval_runtime": 11.2853,
+      "eval_samples_per_second": 5.051,
+      "eval_steps_per_second": 1.684,
+      "step": 231
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.862433862433863e-05,
+      "loss": 0.8942,
+      "step": 232
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 0.9436543583869934,
+      "eval_runtime": 11.5347,
+      "eval_samples_per_second": 4.942,
+      "eval_steps_per_second": 1.647,
+      "step": 232
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.835978835978836e-05,
+      "loss": 0.8406,
+      "step": 233
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 0.9440739154815674,
+      "eval_runtime": 11.1759,
+      "eval_samples_per_second": 5.1,
+      "eval_steps_per_second": 1.7,
+      "step": 233
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.6819,
+      "step": 234
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 0.9439970254898071,
+      "eval_runtime": 11.2746,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 234
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.7830687830687835e-05,
+      "loss": 0.7755,
+      "step": 235
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 0.9441021084785461,
+      "eval_runtime": 11.2612,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 235
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.7566137566137564e-05,
+      "loss": 0.827,
+      "step": 236
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 0.9449384808540344,
+      "eval_runtime": 11.2369,
+      "eval_samples_per_second": 5.073,
+      "eval_steps_per_second": 1.691,
+      "step": 236
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.8764,
+      "step": 237
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 0.9446438550949097,
+      "eval_runtime": 11.27,
+      "eval_samples_per_second": 5.058,
+      "eval_steps_per_second": 1.686,
+      "step": 237
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.7458,
+      "step": 238
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 0.9437587261199951,
+      "eval_runtime": 11.3557,
+      "eval_samples_per_second": 5.019,
+      "eval_steps_per_second": 1.673,
+      "step": 238
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.677248677248677e-05,
+      "loss": 0.9094,
+      "step": 239
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 0.9439199566841125,
+      "eval_runtime": 11.2235,
+      "eval_samples_per_second": 5.079,
+      "eval_steps_per_second": 1.693,
+      "step": 239
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.9898,
+      "step": 240
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 0.942205548286438,
+      "eval_runtime": 11.3626,
+      "eval_samples_per_second": 5.016,
+      "eval_steps_per_second": 1.672,
+      "step": 240
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.6243386243386245e-05,
+      "loss": 0.7391,
+      "step": 241
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 0.9424855709075928,
+      "eval_runtime": 11.4265,
+      "eval_samples_per_second": 4.988,
+      "eval_steps_per_second": 1.663,
+      "step": 241
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.597883597883598e-05,
+      "loss": 0.9334,
+      "step": 242
+    },
+    {
+      "epoch": 1.92,
+      "eval_loss": 0.9417003393173218,
+      "eval_runtime": 11.2556,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 242
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 1.0044,
+      "step": 243
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 0.9410524368286133,
+      "eval_runtime": 11.3604,
+      "eval_samples_per_second": 5.017,
+      "eval_steps_per_second": 1.672,
+      "step": 243
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.5449735449735446e-05,
+      "loss": 0.8364,
+      "step": 244
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 0.9403824806213379,
+      "eval_runtime": 11.2473,
+      "eval_samples_per_second": 5.068,
+      "eval_steps_per_second": 1.689,
+      "step": 244
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.8561,
+      "step": 245
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 0.9392634630203247,
+      "eval_runtime": 11.46,
+      "eval_samples_per_second": 4.974,
+      "eval_steps_per_second": 1.658,
+      "step": 245
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.7015,
+      "step": 246
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 0.9377325773239136,
+      "eval_runtime": 11.6413,
+      "eval_samples_per_second": 4.896,
+      "eval_steps_per_second": 1.632,
+      "step": 246
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.465608465608466e-05,
+      "loss": 0.7486,
+      "step": 247
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 0.9382759928703308,
+      "eval_runtime": 11.5771,
+      "eval_samples_per_second": 4.924,
+      "eval_steps_per_second": 1.641,
+      "step": 247
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.439153439153439e-05,
+      "loss": 0.8333,
+      "step": 248
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 0.9375688433647156,
+      "eval_runtime": 11.2607,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 248
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.8034,
+      "step": 249
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 0.9364966750144958,
+      "eval_runtime": 11.227,
+      "eval_samples_per_second": 5.077,
+      "eval_steps_per_second": 1.692,
+      "step": 249
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.386243386243386e-05,
+      "loss": 0.8792,
+      "step": 250
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 0.9363824129104614,
+      "eval_runtime": 11.3231,
+      "eval_samples_per_second": 5.034,
+      "eval_steps_per_second": 1.678,
+      "step": 250
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.35978835978836e-05,
+      "loss": 0.8516,
+      "step": 251
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 0.9360049962997437,
+      "eval_runtime": 11.3097,
+      "eval_samples_per_second": 5.04,
+      "eval_steps_per_second": 1.68,
+      "step": 251
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.8426,
+      "step": 252
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.9354138374328613,
+      "eval_runtime": 11.3548,
+      "eval_samples_per_second": 5.02,
+      "eval_steps_per_second": 1.673,
+      "step": 252
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.306878306878307e-05,
+      "loss": 0.7673,
+      "step": 253
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 0.935636043548584,
+      "eval_runtime": 11.2919,
+      "eval_samples_per_second": 5.048,
+      "eval_steps_per_second": 1.683,
+      "step": 253
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.280423280423281e-05,
+      "loss": 0.7638,
+      "step": 254
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 0.9350641369819641,
+      "eval_runtime": 11.586,
+      "eval_samples_per_second": 4.92,
+      "eval_steps_per_second": 1.64,
+      "step": 254
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.7744,
+      "step": 255
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 0.9358894228935242,
+      "eval_runtime": 11.2859,
+      "eval_samples_per_second": 5.051,
+      "eval_steps_per_second": 1.684,
+      "step": 255
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.227513227513227e-05,
+      "loss": 0.6928,
+      "step": 256
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 0.9359605312347412,
+      "eval_runtime": 11.3304,
+      "eval_samples_per_second": 5.031,
+      "eval_steps_per_second": 1.677,
+      "step": 256
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 3.2010582010582015e-05,
+      "loss": 0.634,
+      "step": 257
+    },
+    {
+      "epoch": 2.04,
+      "eval_loss": 0.9373098015785217,
+      "eval_runtime": 11.3748,
+      "eval_samples_per_second": 5.011,
+      "eval_steps_per_second": 1.67,
+      "step": 257
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.6593,
+      "step": 258
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 0.9381899833679199,
+      "eval_runtime": 11.2979,
+      "eval_samples_per_second": 5.045,
+      "eval_steps_per_second": 1.682,
+      "step": 258
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.6598,
+      "step": 259
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 0.9383821487426758,
+      "eval_runtime": 11.4676,
+      "eval_samples_per_second": 4.971,
+      "eval_steps_per_second": 1.657,
+      "step": 259
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 3.121693121693122e-05,
+      "loss": 0.7115,
+      "step": 260
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 0.9393886923789978,
+      "eval_runtime": 11.3514,
+      "eval_samples_per_second": 5.021,
+      "eval_steps_per_second": 1.674,
+      "step": 260
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.7168,
+      "step": 261
+    },
+    {
+      "epoch": 2.07,
+      "eval_loss": 0.9403014779090881,
+      "eval_runtime": 11.4331,
+      "eval_samples_per_second": 4.986,
+      "eval_steps_per_second": 1.662,
+      "step": 261
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.068783068783069e-05,
+      "loss": 0.6894,
+      "step": 262
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 0.9411486387252808,
+      "eval_runtime": 11.5658,
+      "eval_samples_per_second": 4.928,
+      "eval_steps_per_second": 1.643,
+      "step": 262
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 3.0423280423280425e-05,
+      "loss": 0.8357,
+      "step": 263
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 0.9419735670089722,
+      "eval_runtime": 11.4171,
+      "eval_samples_per_second": 4.993,
+      "eval_steps_per_second": 1.664,
+      "step": 263
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 0.7271,
+      "step": 264
+    },
+    {
+      "epoch": 2.1,
+      "eval_loss": 0.9432246088981628,
+      "eval_runtime": 11.3767,
+      "eval_samples_per_second": 5.01,
+      "eval_steps_per_second": 1.67,
+      "step": 264
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.9894179894179897e-05,
+      "loss": 0.7627,
+      "step": 265
+    },
+    {
+      "epoch": 2.1,
+      "eval_loss": 0.9447400569915771,
+      "eval_runtime": 11.4372,
+      "eval_samples_per_second": 4.984,
+      "eval_steps_per_second": 1.661,
+      "step": 265
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.6877,
+      "step": 266
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 0.9458262324333191,
+      "eval_runtime": 11.4818,
+      "eval_samples_per_second": 4.964,
+      "eval_steps_per_second": 1.655,
+      "step": 266
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.6246,
+      "step": 267
+    },
+    {
+      "epoch": 2.12,
+      "eval_loss": 0.9467290639877319,
+      "eval_runtime": 11.4153,
+      "eval_samples_per_second": 4.993,
+      "eval_steps_per_second": 1.664,
+      "step": 267
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.91005291005291e-05,
+      "loss": 0.6935,
+      "step": 268
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 0.9481924176216125,
+      "eval_runtime": 11.2889,
+      "eval_samples_per_second": 5.049,
+      "eval_steps_per_second": 1.683,
+      "step": 268
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.8835978835978838e-05,
+      "loss": 0.6503,
+      "step": 269
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 0.9486470222473145,
+      "eval_runtime": 11.1145,
+      "eval_samples_per_second": 5.128,
+      "eval_steps_per_second": 1.709,
+      "step": 269
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.7343,
+      "step": 270
+    },
+    {
+      "epoch": 2.14,
+      "eval_loss": 0.9499481320381165,
+      "eval_runtime": 11.3315,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 270
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.830687830687831e-05,
+      "loss": 0.6769,
+      "step": 271
+    },
+    {
+      "epoch": 2.15,
+      "eval_loss": 0.9516311883926392,
+      "eval_runtime": 11.2964,
+      "eval_samples_per_second": 5.046,
+      "eval_steps_per_second": 1.682,
+      "step": 271
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.8042328042328043e-05,
+      "loss": 0.775,
+      "step": 272
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 0.952580988407135,
+      "eval_runtime": 11.4456,
+      "eval_samples_per_second": 4.98,
+      "eval_steps_per_second": 1.66,
+      "step": 272
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.6589,
+      "step": 273
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 0.9532874226570129,
+      "eval_runtime": 11.499,
+      "eval_samples_per_second": 4.957,
+      "eval_steps_per_second": 1.652,
+      "step": 273
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.7513227513227512e-05,
+      "loss": 0.7719,
+      "step": 274
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 0.95458984375,
+      "eval_runtime": 11.2403,
+      "eval_samples_per_second": 5.071,
+      "eval_steps_per_second": 1.69,
+      "step": 274
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.724867724867725e-05,
+      "loss": 0.7433,
+      "step": 275
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 0.9541744589805603,
+      "eval_runtime": 11.6805,
+      "eval_samples_per_second": 4.88,
+      "eval_steps_per_second": 1.627,
+      "step": 275
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.6634,
+      "step": 276
+    },
+    {
+      "epoch": 2.19,
+      "eval_loss": 0.9544342160224915,
+      "eval_runtime": 11.6449,
+      "eval_samples_per_second": 4.895,
+      "eval_steps_per_second": 1.632,
+      "step": 276
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.6719576719576723e-05,
+      "loss": 0.7462,
+      "step": 277
+    },
+    {
+      "epoch": 2.2,
+      "eval_loss": 0.9551207423210144,
+      "eval_runtime": 11.4102,
+      "eval_samples_per_second": 4.996,
+      "eval_steps_per_second": 1.665,
+      "step": 277
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.6455026455026456e-05,
+      "loss": 0.7134,
+      "step": 278
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 0.9550294876098633,
+      "eval_runtime": 11.3305,
+      "eval_samples_per_second": 5.031,
+      "eval_steps_per_second": 1.677,
+      "step": 278
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.7388,
+      "step": 279
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 0.9552087783813477,
+      "eval_runtime": 11.2749,
+      "eval_samples_per_second": 5.055,
+      "eval_steps_per_second": 1.685,
+      "step": 279
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.6469,
+      "step": 280
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 0.9554876089096069,
+      "eval_runtime": 11.1407,
+      "eval_samples_per_second": 5.116,
+      "eval_steps_per_second": 1.705,
+      "step": 280
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.5661375661375664e-05,
+      "loss": 0.6742,
+      "step": 281
+    },
+    {
+      "epoch": 2.23,
+      "eval_loss": 0.9554134011268616,
+      "eval_runtime": 11.2215,
+      "eval_samples_per_second": 5.08,
+      "eval_steps_per_second": 1.693,
+      "step": 281
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.6205,
+      "step": 282
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 0.9555094242095947,
+      "eval_runtime": 11.265,
+      "eval_samples_per_second": 5.06,
+      "eval_steps_per_second": 1.687,
+      "step": 282
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.5132275132275137e-05,
+      "loss": 0.6558,
+      "step": 283
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 0.9550488591194153,
+      "eval_runtime": 11.4704,
+      "eval_samples_per_second": 4.969,
+      "eval_steps_per_second": 1.656,
+      "step": 283
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4867724867724866e-05,
+      "loss": 0.7589,
+      "step": 284
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 0.9551714062690735,
+      "eval_runtime": 11.2577,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 284
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.7287,
+      "step": 285
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 0.9543017148971558,
+      "eval_runtime": 11.4088,
+      "eval_samples_per_second": 4.996,
+      "eval_steps_per_second": 1.665,
+      "step": 285
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.4338624338624338e-05,
+      "loss": 0.6674,
+      "step": 286
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 0.9542912840843201,
+      "eval_runtime": 11.4616,
+      "eval_samples_per_second": 4.973,
+      "eval_steps_per_second": 1.658,
+      "step": 286
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.6885,
+      "step": 287
+    },
+    {
+      "epoch": 2.28,
+      "eval_loss": 0.9542593955993652,
+      "eval_runtime": 11.5507,
+      "eval_samples_per_second": 4.935,
+      "eval_steps_per_second": 1.645,
+      "step": 287
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.7712,
+      "step": 288
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 0.9528849124908447,
+      "eval_runtime": 11.2495,
+      "eval_samples_per_second": 5.067,
+      "eval_steps_per_second": 1.689,
+      "step": 288
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3544973544973546e-05,
+      "loss": 0.7018,
+      "step": 289
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 0.9520586729049683,
+      "eval_runtime": 11.3693,
+      "eval_samples_per_second": 5.014,
+      "eval_steps_per_second": 1.671,
+      "step": 289
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.328042328042328e-05,
+      "loss": 0.7959,
+      "step": 290
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 0.9512439370155334,
+      "eval_runtime": 11.2658,
+      "eval_samples_per_second": 5.06,
+      "eval_steps_per_second": 1.687,
+      "step": 290
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.6726,
+      "step": 291
+    },
+    {
+      "epoch": 2.31,
+      "eval_loss": 0.9499414563179016,
+      "eval_runtime": 11.4621,
+      "eval_samples_per_second": 4.973,
+      "eval_steps_per_second": 1.658,
+      "step": 291
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.275132275132275e-05,
+      "loss": 0.637,
+      "step": 292
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 0.9500595927238464,
+      "eval_runtime": 11.3492,
+      "eval_samples_per_second": 5.022,
+      "eval_steps_per_second": 1.674,
+      "step": 292
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2486772486772487e-05,
+      "loss": 0.6686,
+      "step": 293
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 0.9487565159797668,
+      "eval_runtime": 11.3746,
+      "eval_samples_per_second": 5.011,
+      "eval_steps_per_second": 1.67,
+      "step": 293
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.7371,
+      "step": 294
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 0.9490134119987488,
+      "eval_runtime": 11.4411,
+      "eval_samples_per_second": 4.982,
+      "eval_steps_per_second": 1.661,
+      "step": 294
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.1957671957671956e-05,
+      "loss": 0.7193,
+      "step": 295
+    },
+    {
+      "epoch": 2.34,
+      "eval_loss": 0.9486552476882935,
+      "eval_runtime": 11.4695,
+      "eval_samples_per_second": 4.97,
+      "eval_steps_per_second": 1.657,
+      "step": 295
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1693121693121692e-05,
+      "loss": 0.736,
+      "step": 296
+    },
+    {
+      "epoch": 2.35,
+      "eval_loss": 0.9484084844589233,
+      "eval_runtime": 11.3144,
+      "eval_samples_per_second": 5.038,
+      "eval_steps_per_second": 1.679,
+      "step": 296
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.7491,
+      "step": 297
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 0.9485853910446167,
+      "eval_runtime": 11.6178,
+      "eval_samples_per_second": 4.906,
+      "eval_steps_per_second": 1.635,
+      "step": 297
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1164021164021164e-05,
+      "loss": 0.5853,
+      "step": 298
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 0.9486856460571289,
+      "eval_runtime": 11.3554,
+      "eval_samples_per_second": 5.02,
+      "eval_steps_per_second": 1.673,
+      "step": 298
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.08994708994709e-05,
+      "loss": 0.6826,
+      "step": 299
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 0.9495511651039124,
+      "eval_runtime": 11.3494,
+      "eval_samples_per_second": 5.022,
+      "eval_steps_per_second": 1.674,
+      "step": 299
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.7434,
+      "step": 300
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 0.9495754837989807,
+      "eval_runtime": 11.8151,
+      "eval_samples_per_second": 4.824,
+      "eval_steps_per_second": 1.608,
+      "step": 300
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.6774,
+      "step": 301
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 0.949608564376831,
+      "eval_runtime": 11.5041,
+      "eval_samples_per_second": 4.955,
+      "eval_steps_per_second": 1.652,
+      "step": 301
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0105820105820105e-05,
+      "loss": 0.6531,
+      "step": 302
+    },
+    {
+      "epoch": 2.4,
+      "eval_loss": 0.9504867792129517,
+      "eval_runtime": 11.7595,
+      "eval_samples_per_second": 4.847,
+      "eval_steps_per_second": 1.616,
+      "step": 302
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.6669,
+      "step": 303
+    },
+    {
+      "epoch": 2.4,
+      "eval_loss": 0.9512569904327393,
+      "eval_runtime": 11.4775,
+      "eval_samples_per_second": 4.966,
+      "eval_steps_per_second": 1.655,
+      "step": 303
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9576719576719577e-05,
+      "loss": 0.7022,
+      "step": 304
+    },
+    {
+      "epoch": 2.41,
+      "eval_loss": 0.9509984254837036,
+      "eval_runtime": 11.5551,
+      "eval_samples_per_second": 4.933,
+      "eval_steps_per_second": 1.644,
+      "step": 304
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9312169312169313e-05,
+      "loss": 0.7916,
+      "step": 305
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 0.9520748257637024,
+      "eval_runtime": 11.445,
+      "eval_samples_per_second": 4.98,
+      "eval_steps_per_second": 1.66,
+      "step": 305
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.6472,
+      "step": 306
+    },
+    {
+      "epoch": 2.43,
+      "eval_loss": 0.9518752098083496,
+      "eval_runtime": 11.2707,
+      "eval_samples_per_second": 5.057,
+      "eval_steps_per_second": 1.686,
+      "step": 306
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8783068783068782e-05,
+      "loss": 0.753,
+      "step": 307
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 0.9526543021202087,
+      "eval_runtime": 11.3873,
+      "eval_samples_per_second": 5.006,
+      "eval_steps_per_second": 1.669,
+      "step": 307
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.7366,
+      "step": 308
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 0.9527204036712646,
+      "eval_runtime": 11.3551,
+      "eval_samples_per_second": 5.02,
+      "eval_steps_per_second": 1.673,
+      "step": 308
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.6592,
+      "step": 309
+    },
+    {
+      "epoch": 2.45,
+      "eval_loss": 0.9531780481338501,
+      "eval_runtime": 11.4348,
+      "eval_samples_per_second": 4.985,
+      "eval_steps_per_second": 1.662,
+      "step": 309
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.798941798941799e-05,
+      "loss": 0.764,
+      "step": 310
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 0.9540254473686218,
+      "eval_runtime": 11.5618,
+      "eval_samples_per_second": 4.93,
+      "eval_steps_per_second": 1.643,
+      "step": 310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.7724867724867723e-05,
+      "loss": 0.6437,
+      "step": 311
+    },
+    {
+      "epoch": 2.47,
+      "eval_loss": 0.9542732238769531,
+      "eval_runtime": 11.2729,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 311
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.7155,
+      "step": 312
+    },
+    {
+      "epoch": 2.48,
+      "eval_loss": 0.9537854194641113,
+      "eval_runtime": 11.3354,
+      "eval_samples_per_second": 5.028,
+      "eval_steps_per_second": 1.676,
+      "step": 312
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7195767195767195e-05,
+      "loss": 0.7399,
+      "step": 313
+    },
+    {
+      "epoch": 2.48,
+      "eval_loss": 0.9536993503570557,
+      "eval_runtime": 11.3021,
+      "eval_samples_per_second": 5.043,
+      "eval_steps_per_second": 1.681,
+      "step": 313
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.693121693121693e-05,
+      "loss": 0.7571,
+      "step": 314
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 0.9531766176223755,
+      "eval_runtime": 11.2165,
+      "eval_samples_per_second": 5.082,
+      "eval_steps_per_second": 1.694,
+      "step": 314
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.7344,
+      "step": 315
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 0.953641414642334,
+      "eval_runtime": 11.3868,
+      "eval_samples_per_second": 5.006,
+      "eval_steps_per_second": 1.669,
+      "step": 315
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6402116402116404e-05,
+      "loss": 0.6649,
+      "step": 316
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.9541411399841309,
+      "eval_runtime": 11.2237,
+      "eval_samples_per_second": 5.079,
+      "eval_steps_per_second": 1.693,
+      "step": 316
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.6137566137566136e-05,
+      "loss": 0.7167,
+      "step": 317
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 0.9530036449432373,
+      "eval_runtime": 11.4155,
+      "eval_samples_per_second": 4.993,
+      "eval_steps_per_second": 1.664,
+      "step": 317
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.6303,
+      "step": 318
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 0.9533805847167969,
+      "eval_runtime": 11.2507,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 318
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.560846560846561e-05,
+      "loss": 0.7081,
+      "step": 319
+    },
+    {
+      "epoch": 2.53,
+      "eval_loss": 0.9538733959197998,
+      "eval_runtime": 11.268,
+      "eval_samples_per_second": 5.059,
+      "eval_steps_per_second": 1.686,
+      "step": 319
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.5343915343915344e-05,
+      "loss": 0.6463,
+      "step": 320
+    },
+    {
+      "epoch": 2.54,
+      "eval_loss": 0.954064667224884,
+      "eval_runtime": 11.4326,
+      "eval_samples_per_second": 4.986,
+      "eval_steps_per_second": 1.662,
+      "step": 320
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.7071,
+      "step": 321
+    },
+    {
+      "epoch": 2.55,
+      "eval_loss": 0.9537148475646973,
+      "eval_runtime": 11.3668,
+      "eval_samples_per_second": 5.015,
+      "eval_steps_per_second": 1.672,
+      "step": 321
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.6748,
+      "step": 322
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 0.953828752040863,
+      "eval_runtime": 11.5259,
+      "eval_samples_per_second": 4.945,
+      "eval_steps_per_second": 1.648,
+      "step": 322
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.455026455026455e-05,
+      "loss": 0.823,
+      "step": 323
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 0.9535531401634216,
+      "eval_runtime": 11.3547,
+      "eval_samples_per_second": 5.02,
+      "eval_steps_per_second": 1.673,
+      "step": 323
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.7127,
+      "step": 324
+    },
+    {
+      "epoch": 2.57,
+      "eval_loss": 0.9530696272850037,
+      "eval_runtime": 11.2591,
+      "eval_samples_per_second": 5.063,
+      "eval_steps_per_second": 1.688,
+      "step": 324
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.4021164021164022e-05,
+      "loss": 0.7587,
+      "step": 325
+    },
+    {
+      "epoch": 2.58,
+      "eval_loss": 0.9529663324356079,
+      "eval_runtime": 11.2567,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 325
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.3756613756613756e-05,
+      "loss": 0.9753,
+      "step": 326
+    },
+    {
+      "epoch": 2.59,
+      "eval_loss": 0.9526656866073608,
+      "eval_runtime": 11.2425,
+      "eval_samples_per_second": 5.07,
+      "eval_steps_per_second": 1.69,
+      "step": 326
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.6198,
+      "step": 327
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 0.9523560404777527,
+      "eval_runtime": 11.4092,
+      "eval_samples_per_second": 4.996,
+      "eval_steps_per_second": 1.665,
+      "step": 327
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.3227513227513228e-05,
+      "loss": 0.7574,
+      "step": 328
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 0.9525935053825378,
+      "eval_runtime": 11.2788,
+      "eval_samples_per_second": 5.054,
+      "eval_steps_per_second": 1.685,
+      "step": 328
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.6888,
+      "step": 329
+    },
+    {
+      "epoch": 2.61,
+      "eval_loss": 0.9519758820533752,
+      "eval_runtime": 11.4679,
+      "eval_samples_per_second": 4.97,
+      "eval_steps_per_second": 1.657,
+      "step": 329
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.697,
+      "step": 330
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 0.9520052671432495,
+      "eval_runtime": 11.2561,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 330
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.2433862433862433e-05,
+      "loss": 0.6414,
+      "step": 331
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.9517341256141663,
+      "eval_runtime": 11.2602,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 331
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.2169312169312169e-05,
+      "loss": 0.6864,
+      "step": 332
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.9515615105628967,
+      "eval_runtime": 11.3829,
+      "eval_samples_per_second": 5.008,
+      "eval_steps_per_second": 1.669,
+      "step": 332
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.7473,
+      "step": 333
+    },
+    {
+      "epoch": 2.64,
+      "eval_loss": 0.9508950114250183,
+      "eval_runtime": 11.2347,
+      "eval_samples_per_second": 5.074,
+      "eval_steps_per_second": 1.691,
+      "step": 333
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.164021164021164e-05,
+      "loss": 0.7251,
+      "step": 334
+    },
+    {
+      "epoch": 2.65,
+      "eval_loss": 0.9501206278800964,
+      "eval_runtime": 11.4336,
+      "eval_samples_per_second": 4.985,
+      "eval_steps_per_second": 1.662,
+      "step": 334
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1375661375661376e-05,
+      "loss": 0.6182,
+      "step": 335
+    },
+    {
+      "epoch": 2.66,
+      "eval_loss": 0.9501546621322632,
+      "eval_runtime": 11.3541,
+      "eval_samples_per_second": 5.02,
+      "eval_steps_per_second": 1.673,
+      "step": 335
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.7664,
+      "step": 336
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 0.9501250386238098,
+      "eval_runtime": 11.2875,
+      "eval_samples_per_second": 5.05,
+      "eval_steps_per_second": 1.683,
+      "step": 336
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.0846560846560846e-05,
+      "loss": 0.725,
+      "step": 337
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 0.9499042630195618,
+      "eval_runtime": 11.3837,
+      "eval_samples_per_second": 5.007,
+      "eval_steps_per_second": 1.669,
+      "step": 337
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.0582010582010582e-05,
+      "loss": 0.7838,
+      "step": 338
+    },
+    {
+      "epoch": 2.68,
+      "eval_loss": 0.9498730897903442,
+      "eval_runtime": 11.3255,
+      "eval_samples_per_second": 5.033,
+      "eval_steps_per_second": 1.678,
+      "step": 338
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.6656,
+      "step": 339
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 0.9496493339538574,
+      "eval_runtime": 11.383,
+      "eval_samples_per_second": 5.007,
+      "eval_steps_per_second": 1.669,
+      "step": 339
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.0052910052910053e-05,
+      "loss": 0.6791,
+      "step": 340
+    },
+    {
+      "epoch": 2.7,
+      "eval_loss": 0.9489718079566956,
+      "eval_runtime": 11.3026,
+      "eval_samples_per_second": 5.043,
+      "eval_steps_per_second": 1.681,
+      "step": 340
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.788359788359789e-06,
+      "loss": 0.7617,
+      "step": 341
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 0.9494317173957825,
+      "eval_runtime": 11.252,
+      "eval_samples_per_second": 5.066,
+      "eval_steps_per_second": 1.689,
+      "step": 341
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.7795,
+      "step": 342
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 0.9486785531044006,
+      "eval_runtime": 11.3314,
+      "eval_samples_per_second": 5.03,
+      "eval_steps_per_second": 1.677,
+      "step": 342
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.6796,
+      "step": 343
+    },
+    {
+      "epoch": 2.72,
+      "eval_loss": 0.9485141634941101,
+      "eval_runtime": 11.2744,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 343
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.994708994708995e-06,
+      "loss": 0.7245,
+      "step": 344
+    },
+    {
+      "epoch": 2.73,
+      "eval_loss": 0.9485604763031006,
+      "eval_runtime": 11.433,
+      "eval_samples_per_second": 4.986,
+      "eval_steps_per_second": 1.662,
+      "step": 344
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.8149,
+      "step": 345
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 0.947691798210144,
+      "eval_runtime": 11.3515,
+      "eval_samples_per_second": 5.021,
+      "eval_steps_per_second": 1.674,
+      "step": 345
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.465608465608466e-06,
+      "loss": 0.8394,
+      "step": 346
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9479726552963257,
+      "eval_runtime": 11.4521,
+      "eval_samples_per_second": 4.977,
+      "eval_steps_per_second": 1.659,
+      "step": 346
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.201058201058202e-06,
+      "loss": 0.7003,
+      "step": 347
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9474308490753174,
+      "eval_runtime": 11.2744,
+      "eval_samples_per_second": 5.056,
+      "eval_steps_per_second": 1.685,
+      "step": 347
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.6286,
+      "step": 348
+    },
+    {
+      "epoch": 2.76,
+      "eval_loss": 0.9480258226394653,
+      "eval_runtime": 11.2479,
+      "eval_samples_per_second": 5.068,
+      "eval_steps_per_second": 1.689,
+      "step": 348
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 7.671957671957672e-06,
+      "loss": 0.6665,
+      "step": 349
+    },
+    {
+      "epoch": 2.77,
+      "eval_loss": 0.9478098750114441,
+      "eval_runtime": 11.161,
+      "eval_samples_per_second": 5.107,
+      "eval_steps_per_second": 1.702,
+      "step": 349
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6232,
+      "step": 350
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 0.9475322365760803,
+      "eval_runtime": 11.3592,
+      "eval_samples_per_second": 5.018,
+      "eval_steps_per_second": 1.673,
+      "step": 350
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.6761,
+      "step": 351
+    },
+    {
+      "epoch": 2.79,
+      "eval_loss": 0.9477640390396118,
+      "eval_runtime": 11.304,
+      "eval_samples_per_second": 5.042,
+      "eval_steps_per_second": 1.681,
+      "step": 351
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 6.878306878306878e-06,
+      "loss": 0.6568,
+      "step": 352
+    },
+    {
+      "epoch": 2.79,
+      "eval_loss": 0.9471892714500427,
+      "eval_runtime": 11.3513,
+      "eval_samples_per_second": 5.021,
+      "eval_steps_per_second": 1.674,
+      "step": 352
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 6.613756613756614e-06,
+      "loss": 0.7018,
+      "step": 353
+    },
+    {
+      "epoch": 2.8,
+      "eval_loss": 0.9472895264625549,
+      "eval_runtime": 11.3699,
+      "eval_samples_per_second": 5.013,
+      "eval_steps_per_second": 1.671,
+      "step": 353
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.8351,
+      "step": 354
+    },
+    {
+      "epoch": 2.81,
+      "eval_loss": 0.9476719498634338,
+      "eval_runtime": 11.2602,
+      "eval_samples_per_second": 5.062,
+      "eval_steps_per_second": 1.687,
+      "step": 354
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 6.0846560846560845e-06,
+      "loss": 0.7903,
+      "step": 355
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 0.9467577934265137,
+      "eval_runtime": 11.7206,
+      "eval_samples_per_second": 4.863,
+      "eval_steps_per_second": 1.621,
+      "step": 355
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.82010582010582e-06,
+      "loss": 0.6964,
+      "step": 356
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 0.9467446804046631,
+      "eval_runtime": 11.3687,
+      "eval_samples_per_second": 5.014,
+      "eval_steps_per_second": 1.671,
+      "step": 356
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.6799,
+      "step": 357
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 0.9476259350776672,
+      "eval_runtime": 11.3754,
+      "eval_samples_per_second": 5.011,
+      "eval_steps_per_second": 1.67,
+      "step": 357
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.291005291005291e-06,
+      "loss": 0.7541,
+      "step": 358
+    },
+    {
+      "epoch": 2.84,
+      "eval_loss": 0.9473299980163574,
+      "eval_runtime": 11.3419,
+      "eval_samples_per_second": 5.026,
+      "eval_steps_per_second": 1.675,
+      "step": 358
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.026455026455026e-06,
+      "loss": 0.6696,
+      "step": 359
+    },
+    {
+      "epoch": 2.85,
+      "eval_loss": 0.9470127820968628,
+      "eval_runtime": 11.4992,
+      "eval_samples_per_second": 4.957,
+      "eval_steps_per_second": 1.652,
+      "step": 359
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.7816,
+      "step": 360
+    },
+    {
+      "epoch": 2.86,
+      "eval_loss": 0.947172999382019,
+      "eval_runtime": 11.4371,
+      "eval_samples_per_second": 4.984,
+      "eval_steps_per_second": 1.661,
+      "step": 360
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.497354497354498e-06,
+      "loss": 0.6759,
+      "step": 361
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 0.9469743967056274,
+      "eval_runtime": 11.9344,
+      "eval_samples_per_second": 4.776,
+      "eval_steps_per_second": 1.592,
+      "step": 361
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.232804232804233e-06,
+      "loss": 0.7542,
+      "step": 362
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 0.9472029805183411,
+      "eval_runtime": 11.3702,
+      "eval_samples_per_second": 5.013,
+      "eval_steps_per_second": 1.671,
+      "step": 362
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.7483,
+      "step": 363
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 0.9470475316047668,
+      "eval_runtime": 11.5479,
+      "eval_samples_per_second": 4.936,
+      "eval_steps_per_second": 1.645,
+      "step": 363
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.6999,
+      "step": 364
+    },
+    {
+      "epoch": 2.89,
+      "eval_loss": 0.947037398815155,
+      "eval_runtime": 11.5187,
+      "eval_samples_per_second": 4.948,
+      "eval_steps_per_second": 1.649,
+      "step": 364
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.439153439153439e-06,
+      "loss": 0.6414,
+      "step": 365
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 0.9470526576042175,
+      "eval_runtime": 11.4702,
+      "eval_samples_per_second": 4.969,
+      "eval_steps_per_second": 1.656,
+      "step": 365
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.671,
+      "step": 366
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 0.9468225240707397,
+      "eval_runtime": 11.5492,
+      "eval_samples_per_second": 4.935,
+      "eval_steps_per_second": 1.645,
+      "step": 366
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.91005291005291e-06,
+      "loss": 0.8444,
+      "step": 367
+    },
+    {
+      "epoch": 2.91,
+      "eval_loss": 0.9471501111984253,
+      "eval_runtime": 11.3616,
+      "eval_samples_per_second": 5.017,
+      "eval_steps_per_second": 1.672,
+      "step": 367
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.6455026455026455e-06,
+      "loss": 0.7024,
+      "step": 368
+    },
+    {
+      "epoch": 2.92,
+      "eval_loss": 0.9467151761054993,
+      "eval_runtime": 11.4772,
+      "eval_samples_per_second": 4.966,
+      "eval_steps_per_second": 1.655,
+      "step": 368
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.6453,
+      "step": 369
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 0.9473956823348999,
+      "eval_runtime": 11.2462,
+      "eval_samples_per_second": 5.068,
+      "eval_steps_per_second": 1.689,
+      "step": 369
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1164021164021164e-06,
+      "loss": 0.7084,
+      "step": 370
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 0.9468894004821777,
+      "eval_runtime": 11.5365,
+      "eval_samples_per_second": 4.941,
+      "eval_steps_per_second": 1.647,
+      "step": 370
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.6694,
+      "step": 371
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 0.9477434158325195,
+      "eval_runtime": 11.3819,
+      "eval_samples_per_second": 5.008,
+      "eval_steps_per_second": 1.669,
+      "step": 371
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.6431,
+      "step": 372
+    },
+    {
+      "epoch": 2.95,
+      "eval_loss": 0.9472144246101379,
+      "eval_runtime": 11.5563,
+      "eval_samples_per_second": 4.932,
+      "eval_steps_per_second": 1.644,
+      "step": 372
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3227513227513228e-06,
+      "loss": 0.6392,
+      "step": 373
+    },
+    {
+      "epoch": 2.96,
+      "eval_loss": 0.9473467469215393,
+      "eval_runtime": 11.3189,
+      "eval_samples_per_second": 5.036,
+      "eval_steps_per_second": 1.679,
+      "step": 373
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0582010582010582e-06,
+      "loss": 0.6886,
+      "step": 374
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 0.9472782611846924,
+      "eval_runtime": 11.4763,
+      "eval_samples_per_second": 4.967,
+      "eval_steps_per_second": 1.656,
+      "step": 374
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.6486,
+      "step": 375
+    },
+    {
+      "epoch": 2.98,
+      "eval_loss": 0.9476314187049866,
+      "eval_runtime": 11.3663,
+      "eval_samples_per_second": 5.015,
+      "eval_steps_per_second": 1.672,
+      "step": 375
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.291005291005291e-07,
+      "loss": 0.7884,
+      "step": 376
+    },
+    {
+      "epoch": 2.98,
+      "eval_loss": 0.9473698735237122,
+      "eval_runtime": 11.5399,
+      "eval_samples_per_second": 4.939,
+      "eval_steps_per_second": 1.646,
+      "step": 376
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.6455026455026455e-07,
+      "loss": 0.6699,
+      "step": 377
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 0.9473439455032349,
+      "eval_runtime": 11.2567,
+      "eval_samples_per_second": 5.064,
+      "eval_steps_per_second": 1.688,
+      "step": 377
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.8267,
+      "step": 378
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.9474151730537415,
+      "eval_runtime": 11.4289,
+      "eval_samples_per_second": 4.987,
+      "eval_steps_per_second": 1.662,
+      "step": 378
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 378,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1,
+  "total_flos": 9987994596962304.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}