diff --git "a/checkpoint-1040/trainer_state.json" "b/checkpoint-1040/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1040/trainer_state.json"
@@ -0,0 +1,7450 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 65,
+  "global_step": 1040,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019230769230769232,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.0,
+      "loss": 1.1154,
+      "step": 1
+    },
+    {
+      "epoch": 0.0019230769230769232,
+      "eval_loss": 1.137184739112854,
+      "eval_runtime": 34.305,
+      "eval_samples_per_second": 68.299,
+      "eval_steps_per_second": 17.082,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038461538461538464,
+      "grad_norm": 3.5,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 1.1559,
+      "step": 2
+    },
+    {
+      "epoch": 0.0057692307692307696,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.1031,
+      "step": 3
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 1.1168,
+      "step": 4
+    },
+    {
+      "epoch": 0.009615384615384616,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.0839,
+      "step": 5
+    },
+    {
+      "epoch": 0.011538461538461539,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 1.0938,
+      "step": 6
+    },
+    {
+      "epoch": 0.013461538461538462,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.1459,
+      "step": 7
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 1.0944,
+      "step": 8
+    },
+    {
+      "epoch": 0.01730769230769231,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.1185,
+      "step": 9
+    },
+    {
+      "epoch": 0.019230769230769232,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 1.0774,
+      "step": 10
+    },
+    {
+      "epoch": 0.021153846153846155,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.1568,
+      "step": 11
+    },
+    {
+      "epoch": 0.023076923076923078,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 1.086,
+      "step": 12
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.0978,
+      "step": 13
+    },
+    {
+      "epoch": 0.026923076923076925,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.5e-06,
+      "loss": 1.109,
+      "step": 14
+    },
+    {
+      "epoch": 0.028846153846153848,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.0631,
+      "step": 15
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 1.1241,
+      "step": 16
+    },
+    {
+      "epoch": 0.032692307692307694,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0644,
+      "step": 17
+    },
+    {
+      "epoch": 0.03461538461538462,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 1.0823,
+      "step": 18
+    },
+    {
+      "epoch": 0.03653846153846154,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.0494,
+      "step": 19
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 2.375,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 1.0778,
+      "step": 20
+    },
+    {
+      "epoch": 0.04038461538461539,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.9911,
+      "step": 21
+    },
+    {
+      "epoch": 0.04230769230769231,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 1.1214,
+      "step": 22
+    },
+    {
+      "epoch": 0.04423076923076923,
+      "grad_norm": 1.859375,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.0294,
+      "step": 23
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 1.0473,
+      "step": 24
+    },
+    {
+      "epoch": 0.04807692307692308,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.0562,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 1.0273,
+      "step": 26
+    },
+    {
+      "epoch": 0.051923076923076926,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5e-06,
+      "loss": 1.0713,
+      "step": 27
+    },
+    {
+      "epoch": 0.05384615384615385,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 1.0218,
+      "step": 28
+    },
+    {
+      "epoch": 0.05576923076923077,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.0322,
+      "step": 29
+    },
+    {
+      "epoch": 0.057692307692307696,
+      "grad_norm": 1.125,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.9993,
+      "step": 30
+    },
+    {
+      "epoch": 0.05961538461538462,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.9909,
+      "step": 31
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 1.0497,
+      "step": 32
+    },
+    {
+      "epoch": 0.06346153846153846,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.0699,
+      "step": 33
+    },
+    {
+      "epoch": 0.06538461538461539,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 1.0065,
+      "step": 34
+    },
+    {
+      "epoch": 0.0673076923076923,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.9839,
+      "step": 35
+    },
+    {
+      "epoch": 0.06923076923076923,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 1.0069,
+      "step": 36
+    },
+    {
+      "epoch": 0.07115384615384615,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.0441,
+      "step": 37
+    },
+    {
+      "epoch": 0.07307692307692308,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.9343,
+      "step": 38
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.9853,
+      "step": 39
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9502,
+      "step": 40
+    },
+    {
+      "epoch": 0.07884615384615384,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.987,
+      "step": 41
+    },
+    {
+      "epoch": 0.08076923076923077,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 1.009,
+      "step": 42
+    },
+    {
+      "epoch": 0.08269230769230769,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.0174,
+      "step": 43
+    },
+    {
+      "epoch": 0.08461538461538462,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 1.0515,
+      "step": 44
+    },
+    {
+      "epoch": 0.08653846153846154,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.0196,
+      "step": 45
+    },
+    {
+      "epoch": 0.08846153846153847,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.9957,
+      "step": 46
+    },
+    {
+      "epoch": 0.09038461538461538,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.9422,
+      "step": 47
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 0.9527,
+      "step": 48
+    },
+    {
+      "epoch": 0.09423076923076923,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9944,
+      "step": 49
+    },
+    {
+      "epoch": 0.09615384615384616,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 0.9794,
+      "step": 50
+    },
+    {
+      "epoch": 0.09807692307692308,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.0226,
+      "step": 51
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 1.0325,
+      "step": 52
+    },
+    {
+      "epoch": 0.10192307692307692,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1e-05,
+      "loss": 0.9893,
+      "step": 53
+    },
+    {
+      "epoch": 0.10384615384615385,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.999974723001716e-06,
+      "loss": 1.0543,
+      "step": 54
+    },
+    {
+      "epoch": 0.10576923076923077,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.999898892262433e-06,
+      "loss": 1.0227,
+      "step": 55
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.999772508548863e-06,
+      "loss": 1.067,
+      "step": 56
+    },
+    {
+      "epoch": 0.10961538461538461,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.999595573138845e-06,
+      "loss": 0.9794,
+      "step": 57
+    },
+    {
+      "epoch": 0.11153846153846154,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.999368087821337e-06,
+      "loss": 1.0166,
+      "step": 58
+    },
+    {
+      "epoch": 0.11346153846153846,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.999090054896397e-06,
+      "loss": 1.0092,
+      "step": 59
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.99876147717516e-06,
+      "loss": 0.9518,
+      "step": 60
+    },
+    {
+      "epoch": 0.11730769230769231,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.99838235797981e-06,
+      "loss": 0.9558,
+      "step": 61
+    },
+    {
+      "epoch": 0.11923076923076924,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.997952701143547e-06,
+      "loss": 1.0134,
+      "step": 62
+    },
+    {
+      "epoch": 0.12115384615384615,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.997472511010543e-06,
+      "loss": 0.8941,
+      "step": 63
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.996941792435903e-06,
+      "loss": 0.9207,
+      "step": 64
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.996360550785619e-06,
+      "loss": 0.9351,
+      "step": 65
+    },
+    {
+      "epoch": 0.125,
+      "eval_loss": 1.0073611736297607,
+      "eval_runtime": 34.5277,
+      "eval_samples_per_second": 67.859,
+      "eval_steps_per_second": 16.972,
+      "step": 65
+    },
+    {
+      "epoch": 0.12692307692307692,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.995728791936505e-06,
+      "loss": 1.0024,
+      "step": 66
+    },
+    {
+      "epoch": 0.12884615384615383,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.995046522276152e-06,
+      "loss": 1.0015,
+      "step": 67
+    },
+    {
+      "epoch": 0.13076923076923078,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.994313748702848e-06,
+      "loss": 0.9595,
+      "step": 68
+    },
+    {
+      "epoch": 0.1326923076923077,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.993530478625524e-06,
+      "loss": 0.9331,
+      "step": 69
+    },
+    {
+      "epoch": 0.1346153846153846,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.992696719963662e-06,
+      "loss": 0.9174,
+      "step": 70
+    },
+    {
+      "epoch": 0.13653846153846153,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.99181248114723e-06,
+      "loss": 1.009,
+      "step": 71
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.9661,
+      "step": 72
+    },
+    {
+      "epoch": 0.14038461538461539,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.989892599322404e-06,
+      "loss": 0.9398,
+      "step": 73
+    },
+    {
+      "epoch": 0.1423076923076923,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.988856975725551e-06,
+      "loss": 0.9973,
+      "step": 74
+    },
+    {
+      "epoch": 0.14423076923076922,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.987770910797014e-06,
+      "loss": 0.8935,
+      "step": 75
+    },
+    {
+      "epoch": 0.14615384615384616,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.986634415517774e-06,
+      "loss": 0.958,
+      "step": 76
+    },
+    {
+      "epoch": 0.14807692307692308,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.985447501378706e-06,
+      "loss": 0.9349,
+      "step": 77
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.984210180380464e-06,
+      "loss": 0.9474,
+      "step": 78
+    },
+    {
+      "epoch": 0.1519230769230769,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.9704,
+      "step": 79
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.981584368357198e-06,
+      "loss": 0.9745,
+      "step": 80
+    },
+    {
+      "epoch": 0.15576923076923077,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.980195903881231e-06,
+      "loss": 0.9527,
+      "step": 81
+    },
+    {
+      "epoch": 0.1576923076923077,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.978757085643937e-06,
+      "loss": 0.9732,
+      "step": 82
+    },
+    {
+      "epoch": 0.1596153846153846,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.97726792819292e-06,
+      "loss": 0.9304,
+      "step": 83
+    },
+    {
+      "epoch": 0.16153846153846155,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.975728446584748e-06,
+      "loss": 0.9999,
+      "step": 84
+    },
+    {
+      "epoch": 0.16346153846153846,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.974138656384815e-06,
+      "loss": 0.9477,
+      "step": 85
+    },
+    {
+      "epoch": 0.16538461538461538,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.97249857366717e-06,
+      "loss": 0.9661,
+      "step": 86
+    },
+    {
+      "epoch": 0.1673076923076923,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.970808215014357e-06,
+      "loss": 0.9763,
+      "step": 87
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.969067597517255e-06,
+      "loss": 0.9292,
+      "step": 88
+    },
+    {
+      "epoch": 0.17115384615384616,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.967276738774897e-06,
+      "loss": 0.9083,
+      "step": 89
+    },
+    {
+      "epoch": 0.17307692307692307,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.9654356568943e-06,
+      "loss": 0.9764,
+      "step": 90
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.9924,
+      "step": 91
+    },
+    {
+      "epoch": 0.17692307692307693,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.961602898685225e-06,
+      "loss": 0.9551,
+      "step": 92
+    },
+    {
+      "epoch": 0.17884615384615385,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.959611261108999e-06,
+      "loss": 1.0074,
+      "step": 93
+    },
+    {
+      "epoch": 0.18076923076923077,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.957569477898636e-06,
+      "loss": 1.0348,
+      "step": 94
+    },
+    {
+      "epoch": 0.18269230769230768,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.955477569698197e-06,
+      "loss": 0.9745,
+      "step": 95
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.95333555765855e-06,
+      "loss": 0.9278,
+      "step": 96
+    },
+    {
+      "epoch": 0.18653846153846154,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.951143463437145e-06,
+      "loss": 0.9497,
+      "step": 97
+    },
+    {
+      "epoch": 0.18846153846153846,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.948901309197807e-06,
+      "loss": 1.0283,
+      "step": 98
+    },
+    {
+      "epoch": 0.19038461538461537,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.946609117610508e-06,
+      "loss": 0.9384,
+      "step": 99
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.94426691185114e-06,
+      "loss": 1.0316,
+      "step": 100
+    },
+    {
+      "epoch": 0.19423076923076923,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.94187471560127e-06,
+      "loss": 0.9401,
+      "step": 101
+    },
+    {
+      "epoch": 0.19615384615384615,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.939432553047919e-06,
+      "loss": 0.9112,
+      "step": 102
+    },
+    {
+      "epoch": 0.19807692307692307,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.936940448883299e-06,
+      "loss": 0.9085,
+      "step": 103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.934398428304577e-06,
+      "loss": 0.9583,
+      "step": 104
+    },
+    {
+      "epoch": 0.20192307692307693,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.9923,
+      "step": 105
+    },
+    {
+      "epoch": 0.20384615384615384,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.929164741216702e-06,
+      "loss": 0.9281,
+      "step": 106
+    },
+    {
+      "epoch": 0.20576923076923076,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.926473127624306e-06,
+      "loss": 0.9767,
+      "step": 107
+    },
+    {
+      "epoch": 0.2076923076923077,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.923731703450794e-06,
+      "loss": 0.9255,
+      "step": 108
+    },
+    {
+      "epoch": 0.20961538461538462,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.920940496414153e-06,
+      "loss": 0.9394,
+      "step": 109
+    },
+    {
+      "epoch": 0.21153846153846154,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.8881,
+      "step": 110
+    },
+    {
+      "epoch": 0.21346153846153845,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.915208847139883e-06,
+      "loss": 0.9327,
+      "step": 111
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.912268462853811e-06,
+      "loss": 0.9637,
+      "step": 112
+    },
+    {
+      "epoch": 0.2173076923076923,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.909278411607134e-06,
+      "loss": 0.904,
+      "step": 113
+    },
+    {
+      "epoch": 0.21923076923076923,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.906238723631662e-06,
+      "loss": 1.0109,
+      "step": 114
+    },
+    {
+      "epoch": 0.22115384615384615,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.903149429661072e-06,
+      "loss": 0.9612,
+      "step": 115
+    },
+    {
+      "epoch": 0.2230769230769231,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.90001056093059e-06,
+      "loss": 0.966,
+      "step": 116
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.896822149176695e-06,
+      "loss": 0.9076,
+      "step": 117
+    },
+    {
+      "epoch": 0.22692307692307692,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.893584226636773e-06,
+      "loss": 0.9435,
+      "step": 118
+    },
+    {
+      "epoch": 0.22884615384615384,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.89029682604881e-06,
+      "loss": 0.9738,
+      "step": 119
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.886959980651056e-06,
+      "loss": 1.0009,
+      "step": 120
+    },
+    {
+      "epoch": 0.2326923076923077,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.883573724181683e-06,
+      "loss": 0.9864,
+      "step": 121
+    },
+    {
+      "epoch": 0.23461538461538461,
+      "grad_norm": 1.125,
+      "learning_rate": 9.880138090878452e-06,
+      "loss": 0.9537,
+      "step": 122
+    },
+    {
+      "epoch": 0.23653846153846153,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.87665311547836e-06,
+      "loss": 1.0204,
+      "step": 123
+    },
+    {
+      "epoch": 0.23846153846153847,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.873118833217294e-06,
+      "loss": 0.9623,
+      "step": 124
+    },
+    {
+      "epoch": 0.2403846153846154,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.869535279829674e-06,
+      "loss": 0.9458,
+      "step": 125
+    },
+    {
+      "epoch": 0.2423076923076923,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.86590249154809e-06,
+      "loss": 0.9223,
+      "step": 126
+    },
+    {
+      "epoch": 0.24423076923076922,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.862220505102933e-06,
+      "loss": 0.9847,
+      "step": 127
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.858489357722028e-06,
+      "loss": 0.9633,
+      "step": 128
+    },
+    {
+      "epoch": 0.24807692307692308,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.896,
+      "step": 129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.850879731549188e-06,
+      "loss": 0.8884,
+      "step": 130
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.9758404493331909,
+      "eval_runtime": 34.4304,
+      "eval_samples_per_second": 68.05,
+      "eval_steps_per_second": 17.02,
+      "step": 130
+    },
+    {
+      "epoch": 0.2519230769230769,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 0.9047,
+      "step": 131
+    },
+    {
+      "epoch": 0.25384615384615383,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.843073920786402e-06,
+      "loss": 0.979,
+      "step": 132
+    },
+    {
+      "epoch": 0.25576923076923075,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.839097544527674e-06,
+      "loss": 0.9224,
+      "step": 133
+    },
+    {
+      "epoch": 0.25769230769230766,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.835072241124815e-06,
+      "loss": 0.8739,
+      "step": 134
+    },
+    {
+      "epoch": 0.25961538461538464,
+      "grad_norm": 1.125,
+      "learning_rate": 9.830998051276858e-06,
+      "loss": 0.9326,
+      "step": 135
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.82687501617712e-06,
+      "loss": 0.9303,
+      "step": 136
+    },
+    {
+      "epoch": 0.26346153846153847,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.822703177512783e-06,
+      "loss": 1.0002,
+      "step": 137
+    },
+    {
+      "epoch": 0.2653846153846154,
+      "grad_norm": 1.125,
+      "learning_rate": 9.818482577464466e-06,
+      "loss": 0.9562,
+      "step": 138
+    },
+    {
+      "epoch": 0.2673076923076923,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.814213258705813e-06,
+      "loss": 0.9212,
+      "step": 139
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 1.125,
+      "learning_rate": 9.809895264403046e-06,
+      "loss": 0.9679,
+      "step": 140
+    },
+    {
+      "epoch": 0.27115384615384613,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.805528638214543e-06,
+      "loss": 0.9903,
+      "step": 141
+    },
+    {
+      "epoch": 0.27307692307692305,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.801113424290381e-06,
+      "loss": 0.9754,
+      "step": 142
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 1.125,
+      "learning_rate": 9.796649667271905e-06,
+      "loss": 0.8977,
+      "step": 143
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.792137412291265e-06,
+      "loss": 0.8798,
+      "step": 144
+    },
+    {
+      "epoch": 0.27884615384615385,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.787576704970965e-06,
+      "loss": 0.9382,
+      "step": 145
+    },
+    {
+      "epoch": 0.28076923076923077,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.7829675914234e-06,
+      "loss": 0.9692,
+      "step": 146
+    },
+    {
+      "epoch": 0.2826923076923077,
+      "grad_norm": 1.125,
+      "learning_rate": 9.778310118250397e-06,
+      "loss": 0.8939,
+      "step": 147
+    },
+    {
+      "epoch": 0.2846153846153846,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 1.0047,
+      "step": 148
+    },
+    {
+      "epoch": 0.2865384615384615,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.768850281879651e-06,
+      "loss": 0.9484,
+      "step": 149
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.764048014328417e-06,
+      "loss": 0.9004,
+      "step": 150
+    },
+    {
+      "epoch": 0.2903846153846154,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.759197578443787e-06,
+      "loss": 0.9281,
+      "step": 151
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.754299023267548e-06,
+      "loss": 0.9638,
+      "step": 152
+    },
+    {
+      "epoch": 0.29423076923076924,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.74935239832801e-06,
+      "loss": 0.9366,
+      "step": 153
+    },
+    {
+      "epoch": 0.29615384615384616,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.7443577536395e-06,
+      "loss": 0.9125,
+      "step": 154
+    },
+    {
+      "epoch": 0.2980769230769231,
+      "grad_norm": 1.125,
+      "learning_rate": 9.739315139701868e-06,
+      "loss": 0.9226,
+      "step": 155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.734224607499978e-06,
+      "loss": 0.9384,
+      "step": 156
+    },
+    {
+      "epoch": 0.3019230769230769,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.8788,
+      "step": 157
+    },
+    {
+      "epoch": 0.3038461538461538,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.723899994664779e-06,
+      "loss": 0.9672,
+      "step": 158
+    },
+    {
+      "epoch": 0.3057692307692308,
+      "grad_norm": 1.125,
+      "learning_rate": 9.71866601842156e-06,
+      "loss": 0.8889,
+      "step": 159
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.713384332693199e-06,
+      "loss": 0.8975,
+      "step": 160
+    },
+    {
+      "epoch": 0.3096153846153846,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.708054990881763e-06,
+      "loss": 0.9614,
+      "step": 161
+    },
+    {
+      "epoch": 0.31153846153846154,
+      "grad_norm": 1.125,
+      "learning_rate": 9.702678046871157e-06,
+      "loss": 0.9061,
+      "step": 162
+    },
+    {
+      "epoch": 0.31346153846153846,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.69725355502658e-06,
+      "loss": 0.9322,
+      "step": 163
+    },
+    {
+      "epoch": 0.3153846153846154,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.691781570193983e-06,
+      "loss": 0.9512,
+      "step": 164
+    },
+    {
+      "epoch": 0.3173076923076923,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.686262147699507e-06,
+      "loss": 1.0271,
+      "step": 165
+    },
+    {
+      "epoch": 0.3192307692307692,
+      "grad_norm": 1.125,
+      "learning_rate": 9.680695343348923e-06,
+      "loss": 0.9529,
+      "step": 166
+    },
+    {
+      "epoch": 0.3211538461538462,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.9636,
+      "step": 167
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.669419814697303e-06,
+      "loss": 0.8879,
+      "step": 168
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 1.125,
+      "learning_rate": 9.663711204400872e-06,
+      "loss": 0.8889,
+      "step": 169
+    },
+    {
+      "epoch": 0.3269230769230769,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.657955440256396e-06,
+      "loss": 0.9735,
+      "step": 170
+    },
+    {
+      "epoch": 0.32884615384615384,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.65215258045925e-06,
+      "loss": 1.0088,
+      "step": 171
+    },
+    {
+      "epoch": 0.33076923076923076,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.64630268368099e-06,
+      "loss": 0.9318,
+      "step": 172
+    },
+    {
+      "epoch": 0.3326923076923077,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.640405809068743e-06,
+      "loss": 0.9902,
+      "step": 173
+    },
+    {
+      "epoch": 0.3346153846153846,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.634462016244625e-06,
+      "loss": 0.9315,
+      "step": 174
+    },
+    {
+      "epoch": 0.33653846153846156,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.628471365305134e-06,
+      "loss": 0.931,
+      "step": 175
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.622433916820539e-06,
+      "loss": 0.9167,
+      "step": 176
+    },
+    {
+      "epoch": 0.3403846153846154,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.616349731834271e-06,
+      "loss": 0.9214,
+      "step": 177
+    },
+    {
+      "epoch": 0.3423076923076923,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.610218871862303e-06,
+      "loss": 0.945,
+      "step": 178
+    },
+    {
+      "epoch": 0.34423076923076923,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.604041398892528e-06,
+      "loss": 0.9445,
+      "step": 179
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.597817375384138e-06,
+      "loss": 1.0135,
+      "step": 180
+    },
+    {
+      "epoch": 0.34807692307692306,
+      "grad_norm": 1.3203125,
+      "learning_rate": 9.591546864266983e-06,
+      "loss": 0.9393,
+      "step": 181
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.585229928940944e-06,
+      "loss": 0.9273,
+      "step": 182
+    },
+    {
+      "epoch": 0.35192307692307695,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 1.0091,
+      "step": 183
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.572457041608018e-06,
+      "loss": 0.9301,
+      "step": 184
+    },
+    {
+      "epoch": 0.3557692307692308,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.56600121874523e-06,
+      "loss": 0.991,
+      "step": 185
+    },
+    {
+      "epoch": 0.3576923076923077,
+      "grad_norm": 1.125,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.8897,
+      "step": 186
+    },
+    {
+      "epoch": 0.3596153846153846,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.55295114099399e-06,
+      "loss": 0.8844,
+      "step": 187
+    },
+    {
+      "epoch": 0.36153846153846153,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.546357018052254e-06,
+      "loss": 0.9226,
+      "step": 188
+    },
+    {
+      "epoch": 0.36346153846153845,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.539716927807102e-06,
+      "loss": 0.9273,
+      "step": 189
+    },
+    {
+      "epoch": 0.36538461538461536,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.533030937395151e-06,
+      "loss": 0.9116,
+      "step": 190
+    },
+    {
+      "epoch": 0.36730769230769234,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.526299114417108e-06,
+      "loss": 0.9732,
+      "step": 191
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.519521526937087e-06,
+      "loss": 0.918,
+      "step": 192
+    },
+    {
+      "epoch": 0.37115384615384617,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.512698243481914e-06,
+      "loss": 0.9045,
+      "step": 193
+    },
+    {
+      "epoch": 0.3730769230769231,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.505829333040437e-06,
+      "loss": 0.9189,
+      "step": 194
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.498914865062831e-06,
+      "loss": 0.9853,
+      "step": 195
+    },
+    {
+      "epoch": 0.375,
+      "eval_loss": 0.9608431458473206,
+      "eval_runtime": 34.5143,
+      "eval_samples_per_second": 67.885,
+      "eval_steps_per_second": 16.978,
+      "step": 195
+    },
+    {
+      "epoch": 0.3769230769230769,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.491954909459895e-06,
+      "loss": 0.9078,
+      "step": 196
+    },
+    {
+      "epoch": 0.37884615384615383,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.484949536602343e-06,
+      "loss": 0.8859,
+      "step": 197
+    },
+    {
+      "epoch": 0.38076923076923075,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.477898817320094e-06,
+      "loss": 0.9183,
+      "step": 198
+    },
+    {
+      "epoch": 0.38269230769230766,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.470802822901558e-06,
+      "loss": 0.918,
+      "step": 199
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.463661625092907e-06,
+      "loss": 0.9444,
+      "step": 200
+    },
+    {
+      "epoch": 0.38653846153846155,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.45647529609736e-06,
+      "loss": 0.8958,
+      "step": 201
+    },
+    {
+      "epoch": 0.38846153846153847,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.44924390857445e-06,
+      "loss": 0.9071,
+      "step": 202
+    },
+    {
+      "epoch": 0.3903846153846154,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.44196753563928e-06,
+      "loss": 0.962,
+      "step": 203
+    },
+    {
+      "epoch": 0.3923076923076923,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.434646250861801e-06,
+      "loss": 0.9814,
+      "step": 204
+    },
+    {
+      "epoch": 0.3942307692307692,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.9873,
+      "step": 205
+    },
+    {
+      "epoch": 0.39615384615384613,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.419869242329417e-06,
+      "loss": 0.9034,
+      "step": 206
+    },
+    {
+      "epoch": 0.39807692307692305,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.412413667981884e-06,
+      "loss": 0.8953,
+      "step": 207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.404913480605264e-06,
+      "loss": 1.0005,
+      "step": 208
+    },
+    {
+      "epoch": 0.40192307692307694,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.9455,
+      "step": 209
+    },
+    {
+      "epoch": 0.40384615384615385,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.389779570546628e-06,
+      "loss": 0.9336,
+      "step": 210
+    },
+    {
+      "epoch": 0.40576923076923077,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.38214600088054e-06,
+      "loss": 0.8907,
+      "step": 211
+    },
+    {
+      "epoch": 0.4076923076923077,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.374468124215676e-06,
+      "loss": 0.8735,
+      "step": 212
+    },
+    {
+      "epoch": 0.4096153846153846,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.366746018181503e-06,
+      "loss": 0.9682,
+      "step": 213
+    },
+    {
+      "epoch": 0.4115384615384615,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.358979760854686e-06,
+      "loss": 0.9069,
+      "step": 214
+    },
+    {
+      "epoch": 0.41346153846153844,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.351169430758293e-06,
+      "loss": 0.9371,
+      "step": 215
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.343315106861008e-06,
+      "loss": 0.9691,
+      "step": 216
+    },
+    {
+      "epoch": 0.4173076923076923,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.33541686857632e-06,
+      "loss": 0.9521,
+      "step": 217
+    },
+    {
+      "epoch": 0.41923076923076924,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.327474795761734e-06,
+      "loss": 0.9503,
+      "step": 218
+    },
+    {
+      "epoch": 0.42115384615384616,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.31948896871795e-06,
+      "loss": 0.9311,
+      "step": 219
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.311459468188066e-06,
+      "loss": 0.8998,
+      "step": 220
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.303386375356752e-06,
+      "loss": 0.8776,
+      "step": 221
+    },
+    {
+      "epoch": 0.4269230769230769,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.295269771849426e-06,
+      "loss": 0.9003,
+      "step": 222
+    },
+    {
+      "epoch": 0.4288461538461538,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.28710973973144e-06,
+      "loss": 0.9597,
+      "step": 223
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.8842,
+      "step": 224
+    },
+    {
+      "epoch": 0.4326923076923077,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.270659720119533e-06,
+      "loss": 0.9524,
+      "step": 225
+    },
+    {
+      "epoch": 0.4346153846153846,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.262369898948462e-06,
+      "loss": 0.9271,
+      "step": 226
+    },
+    {
+      "epoch": 0.43653846153846154,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.254036981810741e-06,
+      "loss": 0.9627,
+      "step": 227
+    },
+    {
+      "epoch": 0.43846153846153846,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.245661052958823e-06,
+      "loss": 0.9305,
+      "step": 228
+    },
+    {
+      "epoch": 0.4403846153846154,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.237242197080045e-06,
+      "loss": 0.9855,
+      "step": 229
+    },
+    {
+      "epoch": 0.4423076923076923,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.22878049929577e-06,
+      "loss": 0.9237,
+      "step": 230
+    },
+    {
+      "epoch": 0.4442307692307692,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.220276045160524e-06,
+      "loss": 0.9733,
+      "step": 231
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.211728920661136e-06,
+      "loss": 0.9613,
+      "step": 232
+    },
+    {
+      "epoch": 0.4480769230769231,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.203139212215868e-06,
+      "loss": 0.9317,
+      "step": 233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.19450700667354e-06,
+      "loss": 0.9067,
+      "step": 234
+    },
+    {
+      "epoch": 0.4519230769230769,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.9088,
+      "step": 235
+    },
+    {
+      "epoch": 0.45384615384615384,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.17711545384048e-06,
+      "loss": 0.9359,
+      "step": 236
+    },
+    {
+      "epoch": 0.45576923076923076,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.168356282392253e-06,
+      "loss": 0.96,
+      "step": 237
+    },
+    {
+      "epoch": 0.4576923076923077,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.159554965530184e-06,
+      "loss": 0.9193,
+      "step": 238
+    },
+    {
+      "epoch": 0.4596153846153846,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.150711592242627e-06,
+      "loss": 0.9439,
+      "step": 239
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.14182625194315e-06,
+      "loss": 0.8804,
+      "step": 240
+    },
+    {
+      "epoch": 0.4634615384615385,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.132899034469648e-06,
+      "loss": 0.916,
+      "step": 241
+    },
+    {
+      "epoch": 0.4653846153846154,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.123930030083425e-06,
+      "loss": 0.8793,
+      "step": 242
+    },
+    {
+      "epoch": 0.4673076923076923,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.9162,
+      "step": 243
+    },
+    {
+      "epoch": 0.46923076923076923,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.10586702372961e-06,
+      "loss": 0.9098,
+      "step": 244
+    },
+    {
+      "epoch": 0.47115384615384615,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.09677320439345e-06,
+      "loss": 0.9219,
+      "step": 245
+    },
+    {
+      "epoch": 0.47307692307692306,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.087637963405586e-06,
+      "loss": 0.9554,
+      "step": 246
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.07846139313061e-06,
+      "loss": 0.9994,
+      "step": 247
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.069243586350976e-06,
+      "loss": 0.9052,
+      "step": 248
+    },
+    {
+      "epoch": 0.47884615384615387,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.059984636266082e-06,
+      "loss": 0.9232,
+      "step": 249
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.050684636491317e-06,
+      "loss": 0.8964,
+      "step": 250
+    },
+    {
+      "epoch": 0.4826923076923077,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.041343681057106e-06,
+      "loss": 0.9386,
+      "step": 251
+    },
+    {
+      "epoch": 0.4846153846153846,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.03196186440798e-06,
+      "loss": 0.9078,
+      "step": 252
+    },
+    {
+      "epoch": 0.48653846153846153,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.022539281401601e-06,
+      "loss": 0.9056,
+      "step": 253
+    },
+    {
+      "epoch": 0.48846153846153845,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.013076027307817e-06,
+      "loss": 0.8973,
+      "step": 254
+    },
+    {
+      "epoch": 0.49038461538461536,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.00357219780769e-06,
+      "loss": 0.9663,
+      "step": 255
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.994027888992533e-06,
+      "loss": 0.8857,
+      "step": 256
+    },
+    {
+      "epoch": 0.49423076923076925,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.984443197362938e-06,
+      "loss": 0.9815,
+      "step": 257
+    },
+    {
+      "epoch": 0.49615384615384617,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.974818219827796e-06,
+      "loss": 0.9693,
+      "step": 258
+    },
+    {
+      "epoch": 0.4980769230769231,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.965153053703325e-06,
+      "loss": 0.8971,
+      "step": 259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.955447796712083e-06,
+      "loss": 0.8998,
+      "step": 260
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.9489682912826538,
+      "eval_runtime": 34.8214,
+      "eval_samples_per_second": 67.286,
+      "eval_steps_per_second": 16.829,
+      "step": 260
+    },
+    {
+      "epoch": 0.5019230769230769,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.9153,
+      "step": 261
+    },
+    {
+      "epoch": 0.5038461538461538,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.9449,
+      "step": 262
+    },
+    {
+      "epoch": 0.5057692307692307,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.926092463837557e-06,
+      "loss": 0.9087,
+      "step": 263
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.916227828696873e-06,
+      "loss": 0.8946,
+      "step": 264
+    },
+    {
+      "epoch": 0.5096153846153846,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.906323597362547e-06,
+      "loss": 0.9261,
+      "step": 265
+    },
+    {
+      "epoch": 0.5115384615384615,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.896379869974273e-06,
+      "loss": 0.8826,
+      "step": 266
+    },
+    {
+      "epoch": 0.5134615384615384,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.886396747071085e-06,
+      "loss": 0.9224,
+      "step": 267
+    },
+    {
+      "epoch": 0.5153846153846153,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.876374329590331e-06,
+      "loss": 0.8849,
+      "step": 268
+    },
+    {
+      "epoch": 0.5173076923076924,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.866312718866669e-06,
+      "loss": 0.9516,
+      "step": 269
+    },
+    {
+      "epoch": 0.5192307692307693,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.85621201663102e-06,
+      "loss": 0.9049,
+      "step": 270
+    },
+    {
+      "epoch": 0.5211538461538462,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.846072325009562e-06,
+      "loss": 0.8953,
+      "step": 271
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.83589374652268e-06,
+      "loss": 0.8507,
+      "step": 272
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.825676384083936e-06,
+      "loss": 0.8904,
+      "step": 273
+    },
+    {
+      "epoch": 0.5269230769230769,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.815420340999034e-06,
+      "loss": 0.9469,
+      "step": 274
+    },
+    {
+      "epoch": 0.5288461538461539,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.805125720964766e-06,
+      "loss": 0.9144,
+      "step": 275
+    },
+    {
+      "epoch": 0.5307692307692308,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.79479262806797e-06,
+      "loss": 0.8757,
+      "step": 276
+    },
+    {
+      "epoch": 0.5326923076923077,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.784421166784476e-06,
+      "loss": 0.8781,
+      "step": 277
+    },
+    {
+      "epoch": 0.5346153846153846,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.774011441978046e-06,
+      "loss": 0.9348,
+      "step": 278
+    },
+    {
+      "epoch": 0.5365384615384615,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.763563558899317e-06,
+      "loss": 0.9949,
+      "step": 279
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.75307762318474e-06,
+      "loss": 0.9171,
+      "step": 280
+    },
+    {
+      "epoch": 0.5403846153846154,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 1.0024,
+      "step": 281
+    },
+    {
+      "epoch": 0.5423076923076923,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.731992018316478e-06,
+      "loss": 0.8619,
+      "step": 282
+    },
+    {
+      "epoch": 0.5442307692307692,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.721392562355113e-06,
+      "loss": 0.955,
+      "step": 283
+    },
+    {
+      "epoch": 0.5461538461538461,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.71075548014038e-06,
+      "loss": 0.9189,
+      "step": 284
+    },
+    {
+      "epoch": 0.5480769230769231,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.700080879221689e-06,
+      "loss": 0.9118,
+      "step": 285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.689368867527781e-06,
+      "loss": 0.8916,
+      "step": 286
+    },
+    {
+      "epoch": 0.551923076923077,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 0.8934,
+      "step": 287
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.875,
+      "learning_rate": 8.667833045419483e-06,
+      "loss": 0.8921,
+      "step": 288
+    },
+    {
+      "epoch": 0.5557692307692308,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.657009452749466e-06,
+      "loss": 1.0005,
+      "step": 289
+    },
+    {
+      "epoch": 0.5576923076923077,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.646148884790786e-06,
+      "loss": 0.8828,
+      "step": 290
+    },
+    {
+      "epoch": 0.5596153846153846,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.635251451352463e-06,
+      "loss": 0.8704,
+      "step": 291
+    },
+    {
+      "epoch": 0.5615384615384615,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.624317262616261e-06,
+      "loss": 0.9182,
+      "step": 292
+    },
+    {
+      "epoch": 0.5634615384615385,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.613346429135567e-06,
+      "loss": 0.9128,
+      "step": 293
+    },
+    {
+      "epoch": 0.5653846153846154,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.602339061834278e-06,
+      "loss": 0.9893,
+      "step": 294
+    },
+    {
+      "epoch": 0.5673076923076923,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.591295272005674e-06,
+      "loss": 0.9299,
+      "step": 295
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.5802151713113e-06,
+      "loss": 0.894,
+      "step": 296
+    },
+    {
+      "epoch": 0.5711538461538461,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.569098871779828e-06,
+      "loss": 0.9472,
+      "step": 297
+    },
+    {
+      "epoch": 0.573076923076923,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.557946485805932e-06,
+      "loss": 0.919,
+      "step": 298
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.546758126149148e-06,
+      "loss": 0.882,
+      "step": 299
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.8685,
+      "step": 300
+    },
+    {
+      "epoch": 0.5788461538461539,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.524273938642539e-06,
+      "loss": 0.8966,
+      "step": 301
+    },
+    {
+      "epoch": 0.5807692307692308,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.512978338125818e-06,
+      "loss": 0.9205,
+      "step": 302
+    },
+    {
+      "epoch": 0.5826923076923077,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.501647218590127e-06,
+      "loss": 0.9815,
+      "step": 303
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.490280694602142e-06,
+      "loss": 0.9317,
+      "step": 304
+    },
+    {
+      "epoch": 0.5865384615384616,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.478878881086505e-06,
+      "loss": 0.9498,
+      "step": 305
+    },
+    {
+      "epoch": 0.5884615384615385,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.467441893324667e-06,
+      "loss": 0.9088,
+      "step": 306
+    },
+    {
+      "epoch": 0.5903846153846154,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.455969846953711e-06,
+      "loss": 0.8728,
+      "step": 307
+    },
+    {
+      "epoch": 0.5923076923076923,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.444462857965198e-06,
+      "loss": 0.9345,
+      "step": 308
+    },
+    {
+      "epoch": 0.5942307692307692,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.432921042703985e-06,
+      "loss": 0.8951,
+      "step": 309
+    },
+    {
+      "epoch": 0.5961538461538461,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.42134451786705e-06,
+      "loss": 1.005,
+      "step": 310
+    },
+    {
+      "epoch": 0.5980769230769231,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.409733400502311e-06,
+      "loss": 0.919,
+      "step": 311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.398087808007447e-06,
+      "loss": 0.882,
+      "step": 312
+    },
+    {
+      "epoch": 0.6019230769230769,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.9254,
+      "step": 313
+    },
+    {
+      "epoch": 0.6038461538461538,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.374693668959717e-06,
+      "loss": 0.9312,
+      "step": 314
+    },
+    {
+      "epoch": 0.6057692307692307,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.362945358940295e-06,
+      "loss": 0.9124,
+      "step": 315
+    },
+    {
+      "epoch": 0.6076923076923076,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.351163046855246e-06,
+      "loss": 0.9181,
+      "step": 316
+    },
+    {
+      "epoch": 0.6096153846153847,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.339346851833163e-06,
+      "loss": 0.9124,
+      "step": 317
+    },
+    {
+      "epoch": 0.6115384615384616,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.327496893345223e-06,
+      "loss": 0.9282,
+      "step": 318
+    },
+    {
+      "epoch": 0.6134615384615385,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.9125,
+      "step": 319
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.303696165562141e-06,
+      "loss": 0.9366,
+      "step": 320
+    },
+    {
+      "epoch": 0.6173076923076923,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.291745636911382e-06,
+      "loss": 0.9808,
+      "step": 321
+    },
+    {
+      "epoch": 0.6192307692307693,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.279761826081096e-06,
+      "loss": 0.9105,
+      "step": 322
+    },
+    {
+      "epoch": 0.6211538461538462,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.26774485423719e-06,
+      "loss": 0.9444,
+      "step": 323
+    },
+    {
+      "epoch": 0.6230769230769231,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.255694842880854e-06,
+      "loss": 0.8981,
+      "step": 324
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.243611913847337e-06,
+      "loss": 0.8919,
+      "step": 325
+    },
+    {
+      "epoch": 0.625,
+      "eval_loss": 0.9419646263122559,
+      "eval_runtime": 34.6043,
+      "eval_samples_per_second": 67.708,
+      "eval_steps_per_second": 16.934,
+      "step": 325
+    },
+    {
+      "epoch": 0.6269230769230769,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.231496189304704e-06,
+      "loss": 0.8434,
+      "step": 326
+    },
+    {
+      "epoch": 0.6288461538461538,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.21934779175262e-06,
+      "loss": 0.912,
+      "step": 327
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.207166844021093e-06,
+      "loss": 0.9085,
+      "step": 328
+    },
+    {
+      "epoch": 0.6326923076923077,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.19495346926924e-06,
+      "loss": 0.9301,
+      "step": 329
+    },
+    {
+      "epoch": 0.6346153846153846,
+      "grad_norm": 0.75,
+      "learning_rate": 8.182707790984043e-06,
+      "loss": 0.9023,
+      "step": 330
+    },
+    {
+      "epoch": 0.6365384615384615,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.170429932979097e-06,
+      "loss": 0.9363,
+      "step": 331
+    },
+    {
+      "epoch": 0.6384615384615384,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.15812001939336e-06,
+      "loss": 0.927,
+      "step": 332
+    },
+    {
+      "epoch": 0.6403846153846153,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.145778174689897e-06,
+      "loss": 0.9826,
+      "step": 333
+    },
+    {
+      "epoch": 0.6423076923076924,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.133404523654626e-06,
+      "loss": 0.922,
+      "step": 334
+    },
+    {
+      "epoch": 0.6442307692307693,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.120999191395048e-06,
+      "loss": 0.9405,
+      "step": 335
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.108562303338987e-06,
+      "loss": 0.8947,
+      "step": 336
+    },
+    {
+      "epoch": 0.6480769230769231,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.096093985233323e-06,
+      "loss": 0.9109,
+      "step": 337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.9111,
+      "step": 338
+    },
+    {
+      "epoch": 0.6519230769230769,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 0.8957,
+      "step": 339
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.058501712846594e-06,
+      "loss": 0.9003,
+      "step": 340
+    },
+    {
+      "epoch": 0.6557692307692308,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.045908938347828e-06,
+      "loss": 0.9372,
+      "step": 341
+    },
+    {
+      "epoch": 0.6576923076923077,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.03328536727506e-06,
+      "loss": 0.8854,
+      "step": 342
+    },
+    {
+      "epoch": 0.6596153846153846,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.020631127262681e-06,
+      "loss": 0.9505,
+      "step": 343
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.007946346255176e-06,
+      "loss": 0.9581,
+      "step": 344
+    },
+    {
+      "epoch": 0.6634615384615384,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.995231152505815e-06,
+      "loss": 0.9068,
+      "step": 345
+    },
+    {
+      "epoch": 0.6653846153846154,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.982485674575373e-06,
+      "loss": 0.9159,
+      "step": 346
+    },
+    {
+      "epoch": 0.6673076923076923,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.96971004133082e-06,
+      "loss": 0.9015,
+      "step": 347
+    },
+    {
+      "epoch": 0.6692307692307692,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.95690438194402e-06,
+      "loss": 0.9531,
+      "step": 348
+    },
+    {
+      "epoch": 0.6711538461538461,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.944068825890424e-06,
+      "loss": 0.8971,
+      "step": 349
+    },
+    {
+      "epoch": 0.6730769230769231,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.931203502947762e-06,
+      "loss": 0.868,
+      "step": 350
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.918308543194735e-06,
+      "loss": 0.9151,
+      "step": 351
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.905384077009693e-06,
+      "loss": 0.9949,
+      "step": 352
+    },
+    {
+      "epoch": 0.6788461538461539,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.892430235069317e-06,
+      "loss": 0.9025,
+      "step": 353
+    },
+    {
+      "epoch": 0.6807692307692308,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.879447148347307e-06,
+      "loss": 0.8969,
+      "step": 354
+    },
+    {
+      "epoch": 0.6826923076923077,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.866434948113046e-06,
+      "loss": 0.9086,
+      "step": 355
+    },
+    {
+      "epoch": 0.6846153846153846,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.853393765930279e-06,
+      "loss": 0.865,
+      "step": 356
+    },
+    {
+      "epoch": 0.6865384615384615,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.9514,
+      "step": 357
+    },
+    {
+      "epoch": 0.6884615384615385,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.827224983438024e-06,
+      "loss": 0.8866,
+      "step": 358
+    },
+    {
+      "epoch": 0.6903846153846154,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.814097647715848e-06,
+      "loss": 0.8856,
+      "step": 359
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.800941859217103e-06,
+      "loss": 0.8864,
+      "step": 360
+    },
+    {
+      "epoch": 0.6942307692307692,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.787757750957335e-06,
+      "loss": 0.9212,
+      "step": 361
+    },
+    {
+      "epoch": 0.6961538461538461,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.77454545623841e-06,
+      "loss": 0.9006,
+      "step": 362
+    },
+    {
+      "epoch": 0.698076923076923,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.761305108647188e-06,
+      "loss": 0.9427,
+      "step": 363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.74803684205417e-06,
+      "loss": 0.9583,
+      "step": 364
+    },
+    {
+      "epoch": 0.7019230769230769,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.9303,
+      "step": 365
+    },
+    {
+      "epoch": 0.7038461538461539,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.72141708875479e-06,
+      "loss": 0.9017,
+      "step": 366
+    },
+    {
+      "epoch": 0.7057692307692308,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.708065871195413e-06,
+      "loss": 0.9247,
+      "step": 367
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.694687272925487e-06,
+      "loss": 0.8598,
+      "step": 368
+    },
+    {
+      "epoch": 0.7096153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.681281429213328e-06,
+      "loss": 0.9719,
+      "step": 369
+    },
+    {
+      "epoch": 0.7115384615384616,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.667848475602735e-06,
+      "loss": 0.9588,
+      "step": 370
+    },
+    {
+      "epoch": 0.7134615384615385,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.654388547911605e-06,
+      "loss": 0.8185,
+      "step": 371
+    },
+    {
+      "epoch": 0.7153846153846154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.640901782230567e-06,
+      "loss": 0.93,
+      "step": 372
+    },
+    {
+      "epoch": 0.7173076923076923,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.627388314921602e-06,
+      "loss": 0.9846,
+      "step": 373
+    },
+    {
+      "epoch": 0.7192307692307692,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.613848282616665e-06,
+      "loss": 0.9807,
+      "step": 374
+    },
+    {
+      "epoch": 0.7211538461538461,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.600281822216307e-06,
+      "loss": 0.9011,
+      "step": 375
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.8961,
+      "step": 376
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.5730701660661795e-06,
+      "loss": 0.9279,
+      "step": 377
+    },
+    {
+      "epoch": 0.7269230769230769,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.559425245448006e-06,
+      "loss": 0.9177,
+      "step": 378
+    },
+    {
+      "epoch": 0.7288461538461538,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.5457544469948164e-06,
+      "loss": 0.9309,
+      "step": 379
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.75,
+      "learning_rate": 7.532057908929311e-06,
+      "loss": 0.8937,
+      "step": 380
+    },
+    {
+      "epoch": 0.7326923076923076,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.5183357697344395e-06,
+      "loss": 0.895,
+      "step": 381
+    },
+    {
+      "epoch": 0.7346153846153847,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.504588168151994e-06,
+      "loss": 0.9167,
+      "step": 382
+    },
+    {
+      "epoch": 0.7365384615384616,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.4908152431812175e-06,
+      "loss": 0.921,
+      "step": 383
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.477017134077389e-06,
+      "loss": 0.8987,
+      "step": 384
+    },
+    {
+      "epoch": 0.7403846153846154,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.4631939803504215e-06,
+      "loss": 0.8866,
+      "step": 385
+    },
+    {
+      "epoch": 0.7423076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.449345921763449e-06,
+      "loss": 0.8745,
+      "step": 386
+    },
+    {
+      "epoch": 0.7442307692307693,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.435473098331411e-06,
+      "loss": 0.865,
+      "step": 387
+    },
+    {
+      "epoch": 0.7461538461538462,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.421575650319641e-06,
+      "loss": 0.8841,
+      "step": 388
+    },
+    {
+      "epoch": 0.7480769230769231,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.407653718242449e-06,
+      "loss": 0.9637,
+      "step": 389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.393707442861693e-06,
+      "loss": 0.914,
+      "step": 390
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.9376137256622314,
+      "eval_runtime": 34.5412,
+      "eval_samples_per_second": 67.832,
+      "eval_steps_per_second": 16.965,
+      "step": 390
+    },
+    {
+      "epoch": 0.7519230769230769,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.9394,
+      "step": 391
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.365742426466169e-06,
+      "loss": 0.9122,
+      "step": 392
+    },
+    {
+      "epoch": 0.7557692307692307,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.3517239682000675e-06,
+      "loss": 0.9033,
+      "step": 393
+    },
+    {
+      "epoch": 0.7576923076923077,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.337681732124882e-06,
+      "loss": 0.8908,
+      "step": 394
+    },
+    {
+      "epoch": 0.7596153846153846,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.8938,
+      "step": 395
+    },
+    {
+      "epoch": 0.7615384615384615,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.30952649469916e-06,
+      "loss": 0.9013,
+      "step": 396
+    },
+    {
+      "epoch": 0.7634615384615384,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.295413778020579e-06,
+      "loss": 0.9203,
+      "step": 397
+    },
+    {
+      "epoch": 0.7653846153846153,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.281277852873947e-06,
+      "loss": 0.9713,
+      "step": 398
+    },
+    {
+      "epoch": 0.7673076923076924,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.267118862184767e-06,
+      "loss": 0.9376,
+      "step": 399
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.252936949111749e-06,
+      "loss": 0.9329,
+      "step": 400
+    },
+    {
+      "epoch": 0.7711538461538462,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.2387322570453724e-06,
+      "loss": 0.8421,
+      "step": 401
+    },
+    {
+      "epoch": 0.7730769230769231,
+      "grad_norm": 0.75,
+      "learning_rate": 7.224504929606429e-06,
+      "loss": 0.8929,
+      "step": 402
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.210255110644569e-06,
+      "loss": 0.9063,
+      "step": 403
+    },
+    {
+      "epoch": 0.7769230769230769,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.195982944236853e-06,
+      "loss": 0.9735,
+      "step": 404
+    },
+    {
+      "epoch": 0.7788461538461539,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.181688574686292e-06,
+      "loss": 0.8794,
+      "step": 405
+    },
+    {
+      "epoch": 0.7807692307692308,
+      "grad_norm": 0.75,
+      "learning_rate": 7.167372146520386e-06,
+      "loss": 0.8891,
+      "step": 406
+    },
+    {
+      "epoch": 0.7826923076923077,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.15303380448967e-06,
+      "loss": 0.8951,
+      "step": 407
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.138673693566241e-06,
+      "loss": 0.897,
+      "step": 408
+    },
+    {
+      "epoch": 0.7865384615384615,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.1242919589422974e-06,
+      "loss": 0.9431,
+      "step": 409
+    },
+    {
+      "epoch": 0.7884615384615384,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.1098887460286745e-06,
+      "loss": 0.8704,
+      "step": 410
+    },
+    {
+      "epoch": 0.7903846153846154,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.095464200453366e-06,
+      "loss": 0.9813,
+      "step": 411
+    },
+    {
+      "epoch": 0.7923076923076923,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.081018468060057e-06,
+      "loss": 0.8657,
+      "step": 412
+    },
+    {
+      "epoch": 0.7942307692307692,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.066551694906651e-06,
+      "loss": 0.9216,
+      "step": 413
+    },
+    {
+      "epoch": 0.7961538461538461,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.9203,
+      "step": 414
+    },
+    {
+      "epoch": 0.7980769230769231,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.0375556116133605e-06,
+      "loss": 0.9002,
+      "step": 415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.023026594647057e-06,
+      "loss": 0.9279,
+      "step": 416
+    },
+    {
+      "epoch": 0.801923076923077,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.8851,
+      "step": 417
+    },
+    {
+      "epoch": 0.8038461538461539,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.9939073445735205e-06,
+      "loss": 0.8718,
+      "step": 418
+    },
+    {
+      "epoch": 0.8057692307692308,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.9793174058851805e-06,
+      "loss": 0.8874,
+      "step": 419
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.964707454715772e-06,
+      "loss": 0.8747,
+      "step": 420
+    },
+    {
+      "epoch": 0.8096153846153846,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.9500776387835785e-06,
+      "loss": 0.9146,
+      "step": 421
+    },
+    {
+      "epoch": 0.8115384615384615,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.935428106007734e-06,
+      "loss": 0.9598,
+      "step": 422
+    },
+    {
+      "epoch": 0.8134615384615385,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.920759004506723e-06,
+      "loss": 0.873,
+      "step": 423
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.906070482596887e-06,
+      "loss": 0.9395,
+      "step": 424
+    },
+    {
+      "epoch": 0.8173076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.891362688790925e-06,
+      "loss": 0.8713,
+      "step": 425
+    },
+    {
+      "epoch": 0.8192307692307692,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.876635771796386e-06,
+      "loss": 0.8427,
+      "step": 426
+    },
+    {
+      "epoch": 0.8211538461538461,
+      "grad_norm": 0.75,
+      "learning_rate": 6.8618898805141744e-06,
+      "loss": 0.9148,
+      "step": 427
+    },
+    {
+      "epoch": 0.823076923076923,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.847125164037036e-06,
+      "loss": 0.8788,
+      "step": 428
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.832341771648057e-06,
+      "loss": 0.8523,
+      "step": 429
+    },
+    {
+      "epoch": 0.8269230769230769,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.817539852819149e-06,
+      "loss": 0.869,
+      "step": 430
+    },
+    {
+      "epoch": 0.8288461538461539,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.802719557209547e-06,
+      "loss": 0.8934,
+      "step": 431
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.787881034664283e-06,
+      "loss": 0.9127,
+      "step": 432
+    },
+    {
+      "epoch": 0.8326923076923077,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.9617,
+      "step": 433
+    },
+    {
+      "epoch": 0.8346153846153846,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.758149909066832e-06,
+      "loss": 0.8918,
+      "step": 434
+    },
+    {
+      "epoch": 0.8365384615384616,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.743257606620094e-06,
+      "loss": 0.9721,
+      "step": 435
+    },
+    {
+      "epoch": 0.8384615384615385,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.728347678445539e-06,
+      "loss": 0.9183,
+      "step": 436
+    },
+    {
+      "epoch": 0.8403846153846154,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.713420275294467e-06,
+      "loss": 0.8995,
+      "step": 437
+    },
+    {
+      "epoch": 0.8423076923076923,
+      "grad_norm": 0.75,
+      "learning_rate": 6.69847554809485e-06,
+      "loss": 0.879,
+      "step": 438
+    },
+    {
+      "epoch": 0.8442307692307692,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.683513647949826e-06,
+      "loss": 0.927,
+      "step": 439
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.668534726136166e-06,
+      "loss": 0.9,
+      "step": 440
+    },
+    {
+      "epoch": 0.8480769230769231,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.653538934102743e-06,
+      "loss": 0.8526,
+      "step": 441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.75,
+      "learning_rate": 6.638526423468999e-06,
+      "loss": 0.8354,
+      "step": 442
+    },
+    {
+      "epoch": 0.8519230769230769,
+      "grad_norm": 0.75,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.8852,
+      "step": 443
+    },
+    {
+      "epoch": 0.8538461538461538,
+      "grad_norm": 0.75,
+      "learning_rate": 6.608451853721985e-06,
+      "loss": 0.9275,
+      "step": 444
+    },
+    {
+      "epoch": 0.8557692307692307,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.593390098686653e-06,
+      "loss": 0.9023,
+      "step": 445
+    },
+    {
+      "epoch": 0.8576923076923076,
+      "grad_norm": 0.75,
+      "learning_rate": 6.578312233203804e-06,
+      "loss": 0.8804,
+      "step": 446
+    },
+    {
+      "epoch": 0.8596153846153847,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.563218409722712e-06,
+      "loss": 0.9229,
+      "step": 447
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.548108780853995e-06,
+      "loss": 0.8995,
+      "step": 448
+    },
+    {
+      "epoch": 0.8634615384615385,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.532983499368078e-06,
+      "loss": 0.8847,
+      "step": 449
+    },
+    {
+      "epoch": 0.8653846153846154,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.5178427181936485e-06,
+      "loss": 0.923,
+      "step": 450
+    },
+    {
+      "epoch": 0.8673076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.502686590416105e-06,
+      "loss": 0.8987,
+      "step": 451
+    },
+    {
+      "epoch": 0.8692307692307693,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.9345,
+      "step": 452
+    },
+    {
+      "epoch": 0.8711538461538462,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.472328908167562e-06,
+      "loss": 0.8575,
+      "step": 453
+    },
+    {
+      "epoch": 0.8730769230769231,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.457127660636994e-06,
+      "loss": 0.9209,
+      "step": 454
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.441911680381074e-06,
+      "loss": 0.8873,
+      "step": 455
+    },
+    {
+      "epoch": 0.875,
+      "eval_loss": 0.9346491098403931,
+      "eval_runtime": 34.5947,
+      "eval_samples_per_second": 67.727,
+      "eval_steps_per_second": 16.939,
+      "step": 455
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.426681121245527e-06,
+      "loss": 0.9187,
+      "step": 456
+    },
+    {
+      "epoch": 0.8788461538461538,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.411436137223479e-06,
+      "loss": 0.9509,
+      "step": 457
+    },
+    {
+      "epoch": 0.8807692307692307,
+      "grad_norm": 0.75,
+      "learning_rate": 6.396176882453902e-06,
+      "loss": 0.9401,
+      "step": 458
+    },
+    {
+      "epoch": 0.8826923076923077,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.38090351122006e-06,
+      "loss": 0.8767,
+      "step": 459
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.365616177947945e-06,
+      "loss": 0.8637,
+      "step": 460
+    },
+    {
+      "epoch": 0.8865384615384615,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.350315037204714e-06,
+      "loss": 0.9081,
+      "step": 461
+    },
+    {
+      "epoch": 0.8884615384615384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.335000243697134e-06,
+      "loss": 0.9054,
+      "step": 462
+    },
+    {
+      "epoch": 0.8903846153846153,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.319671952270004e-06,
+      "loss": 0.9045,
+      "step": 463
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.304330317904605e-06,
+      "loss": 0.9227,
+      "step": 464
+    },
+    {
+      "epoch": 0.8942307692307693,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.288975495717124e-06,
+      "loss": 0.8882,
+      "step": 465
+    },
+    {
+      "epoch": 0.8961538461538462,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.273607640957085e-06,
+      "loss": 0.9967,
+      "step": 466
+    },
+    {
+      "epoch": 0.8980769230769231,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.258226909005783e-06,
+      "loss": 0.9474,
+      "step": 467
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.2428334553747135e-06,
+      "loss": 0.912,
+      "step": 468
+    },
+    {
+      "epoch": 0.9019230769230769,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.9355,
+      "step": 469
+    },
+    {
+      "epoch": 0.9038461538461539,
+      "grad_norm": 0.75,
+      "learning_rate": 6.212009005760805e-06,
+      "loss": 0.9328,
+      "step": 470
+    },
+    {
+      "epoch": 0.9057692307692308,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.9323,
+      "step": 471
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.181135538751504e-06,
+      "loss": 0.8865,
+      "step": 472
+    },
+    {
+      "epoch": 0.9096153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.165680813840822e-06,
+      "loss": 0.9123,
+      "step": 473
+    },
+    {
+      "epoch": 0.9115384615384615,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.150214302965368e-06,
+      "loss": 0.9209,
+      "step": 474
+    },
+    {
+      "epoch": 0.9134615384615384,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.134736162503929e-06,
+      "loss": 0.9377,
+      "step": 475
+    },
+    {
+      "epoch": 0.9153846153846154,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.119246548952877e-06,
+      "loss": 0.9317,
+      "step": 476
+    },
+    {
+      "epoch": 0.9173076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.103745618924587e-06,
+      "loss": 0.8839,
+      "step": 477
+    },
+    {
+      "epoch": 0.9192307692307692,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.088233529145849e-06,
+      "loss": 0.8823,
+      "step": 478
+    },
+    {
+      "epoch": 0.9211538461538461,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.072710436456293e-06,
+      "loss": 0.9031,
+      "step": 479
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.057176497806791e-06,
+      "loss": 0.9132,
+      "step": 480
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.041631870257882e-06,
+      "loss": 0.8772,
+      "step": 481
+    },
+    {
+      "epoch": 0.926923076923077,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.026076710978172e-06,
+      "loss": 0.901,
+      "step": 482
+    },
+    {
+      "epoch": 0.9288461538461539,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.010511177242757e-06,
+      "loss": 0.9196,
+      "step": 483
+    },
+    {
+      "epoch": 0.9307692307692308,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.994935426431627e-06,
+      "loss": 0.9718,
+      "step": 484
+    },
+    {
+      "epoch": 0.9326923076923077,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.979349616028067e-06,
+      "loss": 0.963,
+      "step": 485
+    },
+    {
+      "epoch": 0.9346153846153846,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.963753903617084e-06,
+      "loss": 0.9048,
+      "step": 486
+    },
+    {
+      "epoch": 0.9365384615384615,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.948148446883794e-06,
+      "loss": 0.9705,
+      "step": 487
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.932533403611835e-06,
+      "loss": 0.8878,
+      "step": 488
+    },
+    {
+      "epoch": 0.9403846153846154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.916908931681781e-06,
+      "loss": 0.9049,
+      "step": 489
+    },
+    {
+      "epoch": 0.9423076923076923,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.8733,
+      "step": 490
+    },
+    {
+      "epoch": 0.9442307692307692,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.885632333844714e-06,
+      "loss": 0.8896,
+      "step": 491
+    },
+    {
+      "epoch": 0.9461538461538461,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.8699805241691065e-06,
+      "loss": 0.9191,
+      "step": 492
+    },
+    {
+      "epoch": 0.948076923076923,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.854319918295012e-06,
+      "loss": 0.8721,
+      "step": 493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.838650674563674e-06,
+      "loss": 0.8746,
+      "step": 494
+    },
+    {
+      "epoch": 0.9519230769230769,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.9072,
+      "step": 495
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.807286907329315e-06,
+      "loss": 0.8981,
+      "step": 496
+    },
+    {
+      "epoch": 0.9557692307692308,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.79159270093905e-06,
+      "loss": 0.9358,
+      "step": 497
+    },
+    {
+      "epoch": 0.9576923076923077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.7758904909138495e-06,
+      "loss": 0.915,
+      "step": 498
+    },
+    {
+      "epoch": 0.9596153846153846,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.760180436015604e-06,
+      "loss": 0.8652,
+      "step": 499
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.74446269508553e-06,
+      "loss": 0.8926,
+      "step": 500
+    },
+    {
+      "epoch": 0.9634615384615385,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.7287374270425475e-06,
+      "loss": 0.8889,
+      "step": 501
+    },
+    {
+      "epoch": 0.9653846153846154,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.7130047908816884e-06,
+      "loss": 0.9027,
+      "step": 502
+    },
+    {
+      "epoch": 0.9673076923076923,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.69726494567248e-06,
+      "loss": 0.9419,
+      "step": 503
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.681518050557336e-06,
+      "loss": 0.9396,
+      "step": 504
+    },
+    {
+      "epoch": 0.9711538461538461,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.6657642647499545e-06,
+      "loss": 0.8828,
+      "step": 505
+    },
+    {
+      "epoch": 0.9730769230769231,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.650003747533701e-06,
+      "loss": 0.944,
+      "step": 506
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.6342366582600035e-06,
+      "loss": 0.9072,
+      "step": 507
+    },
+    {
+      "epoch": 0.9769230769230769,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.61846315634674e-06,
+      "loss": 0.9015,
+      "step": 508
+    },
+    {
+      "epoch": 0.9788461538461538,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.9117,
+      "step": 509
+    },
+    {
+      "epoch": 0.9807692307692307,
+      "grad_norm": 0.75,
+      "learning_rate": 5.586897552595573e-06,
+      "loss": 0.971,
+      "step": 510
+    },
+    {
+      "epoch": 0.9826923076923076,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.571105769911159e-06,
+      "loss": 0.8729,
+      "step": 511
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.555308212890917e-06,
+      "loss": 0.9132,
+      "step": 512
+    },
+    {
+      "epoch": 0.9865384615384616,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.539505041260779e-06,
+      "loss": 0.8606,
+      "step": 513
+    },
+    {
+      "epoch": 0.9884615384615385,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.523696414803438e-06,
+      "loss": 0.8937,
+      "step": 514
+    },
+    {
+      "epoch": 0.9903846153846154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.507882493356745e-06,
+      "loss": 0.913,
+      "step": 515
+    },
+    {
+      "epoch": 0.9923076923076923,
+      "grad_norm": 4.5625,
+      "learning_rate": 5.49206343681209e-06,
+      "loss": 0.862,
+      "step": 516
+    },
+    {
+      "epoch": 0.9942307692307693,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.476239405112775e-06,
+      "loss": 0.8662,
+      "step": 517
+    },
+    {
+      "epoch": 0.9961538461538462,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.460410558252408e-06,
+      "loss": 0.8443,
+      "step": 518
+    },
+    {
+      "epoch": 0.9980769230769231,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.444577056273284e-06,
+      "loss": 0.9569,
+      "step": 519
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.428739059264767e-06,
+      "loss": 0.8854,
+      "step": 520
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9325999617576599,
+      "eval_runtime": 34.5297,
+      "eval_samples_per_second": 67.855,
+      "eval_steps_per_second": 16.971,
+      "step": 520
+    },
+    {
+      "epoch": 1.001923076923077,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.9419,
+      "step": 521
+    },
+    {
+      "epoch": 1.0038461538461538,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.39705022074261e-06,
+      "loss": 0.843,
+      "step": 522
+    },
+    {
+      "epoch": 1.0057692307692307,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.381199699628459e-06,
+      "loss": 0.8494,
+      "step": 523
+    },
+    {
+      "epoch": 1.0076923076923077,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.365345324280646e-06,
+      "loss": 0.8623,
+      "step": 524
+    },
+    {
+      "epoch": 1.0096153846153846,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.349487254999579e-06,
+      "loss": 0.8716,
+      "step": 525
+    },
+    {
+      "epoch": 1.0115384615384615,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.333625652123014e-06,
+      "loss": 0.9326,
+      "step": 526
+    },
+    {
+      "epoch": 1.0134615384615384,
+      "grad_norm": 0.75,
+      "learning_rate": 5.317760676024436e-06,
+      "loss": 0.8741,
+      "step": 527
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.8816,
+      "step": 528
+    },
+    {
+      "epoch": 1.0173076923076922,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.286021245824075e-06,
+      "loss": 0.8976,
+      "step": 529
+    },
+    {
+      "epoch": 1.0192307692307692,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.270147112633304e-06,
+      "loss": 0.8997,
+      "step": 530
+    },
+    {
+      "epoch": 1.021153846153846,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.254270248039291e-06,
+      "loss": 0.9766,
+      "step": 531
+    },
+    {
+      "epoch": 1.023076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.238390812569828e-06,
+      "loss": 0.8946,
+      "step": 532
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.222508966778702e-06,
+      "loss": 0.909,
+      "step": 533
+    },
+    {
+      "epoch": 1.0269230769230768,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.206624871244066e-06,
+      "loss": 0.8777,
+      "step": 534
+    },
+    {
+      "epoch": 1.0288461538461537,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.190738686566826e-06,
+      "loss": 0.8775,
+      "step": 535
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.1748505733690035e-06,
+      "loss": 0.8591,
+      "step": 536
+    },
+    {
+      "epoch": 1.0326923076923078,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.158960692292122e-06,
+      "loss": 0.942,
+      "step": 537
+    },
+    {
+      "epoch": 1.0346153846153847,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.143069203995586e-06,
+      "loss": 0.8728,
+      "step": 538
+    },
+    {
+      "epoch": 1.0365384615384616,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.1271762691550375e-06,
+      "loss": 0.8437,
+      "step": 539
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.111282048460753e-06,
+      "loss": 0.8885,
+      "step": 540
+    },
+    {
+      "epoch": 1.0403846153846155,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.095386702616012e-06,
+      "loss": 0.9171,
+      "step": 541
+    },
+    {
+      "epoch": 1.0423076923076924,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.079490392335463e-06,
+      "loss": 0.8742,
+      "step": 542
+    },
+    {
+      "epoch": 1.0442307692307693,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.06359327834351e-06,
+      "loss": 0.9521,
+      "step": 543
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.047695521372681e-06,
+      "loss": 0.913,
+      "step": 544
+    },
+    {
+      "epoch": 1.0480769230769231,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.031797282162007e-06,
+      "loss": 0.9162,
+      "step": 545
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.015898721455394e-06,
+      "loss": 0.9098,
+      "step": 546
+    },
+    {
+      "epoch": 1.051923076923077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5e-06,
+      "loss": 0.8851,
+      "step": 547
+    },
+    {
+      "epoch": 1.0538461538461539,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.984101278544607e-06,
+      "loss": 0.9895,
+      "step": 548
+    },
+    {
+      "epoch": 1.0557692307692308,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.968202717837996e-06,
+      "loss": 0.8406,
+      "step": 549
+    },
+    {
+      "epoch": 1.0576923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.9523044786273214e-06,
+      "loss": 0.9372,
+      "step": 550
+    },
+    {
+      "epoch": 1.0596153846153846,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.936406721656492e-06,
+      "loss": 0.8837,
+      "step": 551
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.92050960766454e-06,
+      "loss": 0.9099,
+      "step": 552
+    },
+    {
+      "epoch": 1.0634615384615385,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.9046132973839895e-06,
+      "loss": 0.8802,
+      "step": 553
+    },
+    {
+      "epoch": 1.0653846153846154,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.8887179515392465e-06,
+      "loss": 0.8892,
+      "step": 554
+    },
+    {
+      "epoch": 1.0673076923076923,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.872823730844966e-06,
+      "loss": 0.9367,
+      "step": 555
+    },
+    {
+      "epoch": 1.0692307692307692,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.856930796004417e-06,
+      "loss": 0.999,
+      "step": 556
+    },
+    {
+      "epoch": 1.0711538461538461,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.841039307707878e-06,
+      "loss": 0.8129,
+      "step": 557
+    },
+    {
+      "epoch": 1.073076923076923,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.825149426630999e-06,
+      "loss": 0.8832,
+      "step": 558
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.809261313433176e-06,
+      "loss": 0.9148,
+      "step": 559
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.793375128755934e-06,
+      "loss": 0.9256,
+      "step": 560
+    },
+    {
+      "epoch": 1.0788461538461538,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.7774910332213005e-06,
+      "loss": 0.9012,
+      "step": 561
+    },
+    {
+      "epoch": 1.0807692307692307,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.761609187430174e-06,
+      "loss": 0.85,
+      "step": 562
+    },
+    {
+      "epoch": 1.0826923076923076,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.74572975196071e-06,
+      "loss": 0.9216,
+      "step": 563
+    },
+    {
+      "epoch": 1.0846153846153845,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.7298528873666985e-06,
+      "loss": 0.8746,
+      "step": 564
+    },
+    {
+      "epoch": 1.0865384615384615,
+      "grad_norm": 0.75,
+      "learning_rate": 4.713978754175926e-06,
+      "loss": 0.8979,
+      "step": 565
+    },
+    {
+      "epoch": 1.0884615384615384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.888,
+      "step": 566
+    },
+    {
+      "epoch": 1.0903846153846153,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.682239323975566e-06,
+      "loss": 0.9075,
+      "step": 567
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.666374347876987e-06,
+      "loss": 0.8873,
+      "step": 568
+    },
+    {
+      "epoch": 1.0942307692307693,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.6505127450004216e-06,
+      "loss": 0.8379,
+      "step": 569
+    },
+    {
+      "epoch": 1.0961538461538463,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.634654675719355e-06,
+      "loss": 0.9033,
+      "step": 570
+    },
+    {
+      "epoch": 1.0980769230769232,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.618800300371543e-06,
+      "loss": 0.896,
+      "step": 571
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.60294977925739e-06,
+      "loss": 0.8901,
+      "step": 572
+    },
+    {
+      "epoch": 1.101923076923077,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.9841,
+      "step": 573
+    },
+    {
+      "epoch": 1.103846153846154,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.571260940735235e-06,
+      "loss": 0.8862,
+      "step": 574
+    },
+    {
+      "epoch": 1.1057692307692308,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.555422943726715e-06,
+      "loss": 0.9193,
+      "step": 575
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.539589441747595e-06,
+      "loss": 0.9534,
+      "step": 576
+    },
+    {
+      "epoch": 1.1096153846153847,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.523760594887228e-06,
+      "loss": 0.9352,
+      "step": 577
+    },
+    {
+      "epoch": 1.1115384615384616,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.507936563187911e-06,
+      "loss": 0.8709,
+      "step": 578
+    },
+    {
+      "epoch": 1.1134615384615385,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.492117506643256e-06,
+      "loss": 0.9662,
+      "step": 579
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.476303585196563e-06,
+      "loss": 0.8556,
+      "step": 580
+    },
+    {
+      "epoch": 1.1173076923076923,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.460494958739223e-06,
+      "loss": 0.9715,
+      "step": 581
+    },
+    {
+      "epoch": 1.1192307692307693,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.444691787109085e-06,
+      "loss": 0.8285,
+      "step": 582
+    },
+    {
+      "epoch": 1.1211538461538462,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.428894230088842e-06,
+      "loss": 0.908,
+      "step": 583
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.413102447404428e-06,
+      "loss": 0.946,
+      "step": 584
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.9365,
+      "step": 585
+    },
+    {
+      "epoch": 1.125,
+      "eval_loss": 0.9315983653068542,
+      "eval_runtime": 34.5056,
+      "eval_samples_per_second": 67.902,
+      "eval_steps_per_second": 16.983,
+      "step": 585
+    },
+    {
+      "epoch": 1.126923076923077,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.381536843653262e-06,
+      "loss": 0.9313,
+      "step": 586
+    },
+    {
+      "epoch": 1.1288461538461538,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.365763341739996e-06,
+      "loss": 0.882,
+      "step": 587
+    },
+    {
+      "epoch": 1.1307692307692307,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.3499962524662995e-06,
+      "loss": 0.8806,
+      "step": 588
+    },
+    {
+      "epoch": 1.1326923076923077,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.334235735250047e-06,
+      "loss": 0.8719,
+      "step": 589
+    },
+    {
+      "epoch": 1.1346153846153846,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.318481949442665e-06,
+      "loss": 0.9148,
+      "step": 590
+    },
+    {
+      "epoch": 1.1365384615384615,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.302735054327523e-06,
+      "loss": 0.8417,
+      "step": 591
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.286995209118313e-06,
+      "loss": 0.8802,
+      "step": 592
+    },
+    {
+      "epoch": 1.1403846153846153,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.271262572957453e-06,
+      "loss": 0.9018,
+      "step": 593
+    },
+    {
+      "epoch": 1.1423076923076922,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.255537304914472e-06,
+      "loss": 0.8499,
+      "step": 594
+    },
+    {
+      "epoch": 1.1442307692307692,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.239819563984397e-06,
+      "loss": 0.948,
+      "step": 595
+    },
+    {
+      "epoch": 1.146153846153846,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.224109509086151e-06,
+      "loss": 0.9229,
+      "step": 596
+    },
+    {
+      "epoch": 1.148076923076923,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.2084072990609505e-06,
+      "loss": 0.8326,
+      "step": 597
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.192713092670687e-06,
+      "loss": 0.899,
+      "step": 598
+    },
+    {
+      "epoch": 1.1519230769230768,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.8957,
+      "step": 599
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.161349325436328e-06,
+      "loss": 0.9516,
+      "step": 600
+    },
+    {
+      "epoch": 1.1557692307692307,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.145680081704989e-06,
+      "loss": 0.9012,
+      "step": 601
+    },
+    {
+      "epoch": 1.1576923076923076,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.1300194758308935e-06,
+      "loss": 0.8685,
+      "step": 602
+    },
+    {
+      "epoch": 1.1596153846153845,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.1143676661552876e-06,
+      "loss": 0.9122,
+      "step": 603
+    },
+    {
+      "epoch": 1.1615384615384616,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.9732,
+      "step": 604
+    },
+    {
+      "epoch": 1.1634615384615385,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.08309106831822e-06,
+      "loss": 0.8518,
+      "step": 605
+    },
+    {
+      "epoch": 1.1653846153846155,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.067466596388166e-06,
+      "loss": 0.8326,
+      "step": 606
+    },
+    {
+      "epoch": 1.1673076923076924,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.051851553116208e-06,
+      "loss": 0.8629,
+      "step": 607
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.036246096382916e-06,
+      "loss": 0.9053,
+      "step": 608
+    },
+    {
+      "epoch": 1.1711538461538462,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.0206503839719335e-06,
+      "loss": 0.9191,
+      "step": 609
+    },
+    {
+      "epoch": 1.1730769230769231,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.0050645735683745e-06,
+      "loss": 0.9079,
+      "step": 610
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.989488822757244e-06,
+      "loss": 0.9048,
+      "step": 611
+    },
+    {
+      "epoch": 1.176923076923077,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.973923289021829e-06,
+      "loss": 0.8945,
+      "step": 612
+    },
+    {
+      "epoch": 1.1788461538461539,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.9583681297421194e-06,
+      "loss": 0.8879,
+      "step": 613
+    },
+    {
+      "epoch": 1.1807692307692308,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.9428235021932104e-06,
+      "loss": 0.982,
+      "step": 614
+    },
+    {
+      "epoch": 1.1826923076923077,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.927289563543709e-06,
+      "loss": 0.8807,
+      "step": 615
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.911766470854152e-06,
+      "loss": 0.9508,
+      "step": 616
+    },
+    {
+      "epoch": 1.1865384615384615,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.896254381075416e-06,
+      "loss": 0.8701,
+      "step": 617
+    },
+    {
+      "epoch": 1.1884615384615385,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.880753451047124e-06,
+      "loss": 0.9148,
+      "step": 618
+    },
+    {
+      "epoch": 1.1903846153846154,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.865263837496072e-06,
+      "loss": 0.8937,
+      "step": 619
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.849785697034634e-06,
+      "loss": 0.855,
+      "step": 620
+    },
+    {
+      "epoch": 1.1942307692307692,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.834319186159179e-06,
+      "loss": 0.8574,
+      "step": 621
+    },
+    {
+      "epoch": 1.1961538461538461,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.818864461248498e-06,
+      "loss": 0.9341,
+      "step": 622
+    },
+    {
+      "epoch": 1.198076923076923,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.9029,
+      "step": 623
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.7879909942391963e-06,
+      "loss": 0.9188,
+      "step": 624
+    },
+    {
+      "epoch": 1.2019230769230769,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.8459,
+      "step": 625
+    },
+    {
+      "epoch": 1.2038461538461538,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.7571665446252886e-06,
+      "loss": 0.8931,
+      "step": 626
+    },
+    {
+      "epoch": 1.2057692307692307,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.7417730909942184e-06,
+      "loss": 0.8719,
+      "step": 627
+    },
+    {
+      "epoch": 1.2076923076923076,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.726392359042917e-06,
+      "loss": 0.922,
+      "step": 628
+    },
+    {
+      "epoch": 1.2096153846153845,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.7110245042828786e-06,
+      "loss": 0.8538,
+      "step": 629
+    },
+    {
+      "epoch": 1.2115384615384615,
+      "grad_norm": 0.75,
+      "learning_rate": 3.695669682095397e-06,
+      "loss": 0.9149,
+      "step": 630
+    },
+    {
+      "epoch": 1.2134615384615384,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.6803280477299975e-06,
+      "loss": 0.8769,
+      "step": 631
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.664999756302869e-06,
+      "loss": 0.8532,
+      "step": 632
+    },
+    {
+      "epoch": 1.2173076923076924,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.6496849627952875e-06,
+      "loss": 0.9334,
+      "step": 633
+    },
+    {
+      "epoch": 1.2192307692307693,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.634383822052057e-06,
+      "loss": 0.9112,
+      "step": 634
+    },
+    {
+      "epoch": 1.2211538461538463,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.6190964887799418e-06,
+      "loss": 0.8462,
+      "step": 635
+    },
+    {
+      "epoch": 1.2230769230769232,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.6038231175461004e-06,
+      "loss": 0.915,
+      "step": 636
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.5885638627765228e-06,
+      "loss": 0.9869,
+      "step": 637
+    },
+    {
+      "epoch": 1.226923076923077,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.573318878754475e-06,
+      "loss": 0.8105,
+      "step": 638
+    },
+    {
+      "epoch": 1.228846153846154,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.5580883196189265e-06,
+      "loss": 0.929,
+      "step": 639
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.5428723393630067e-06,
+      "loss": 0.9044,
+      "step": 640
+    },
+    {
+      "epoch": 1.2326923076923078,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.52767109183244e-06,
+      "loss": 0.8723,
+      "step": 641
+    },
+    {
+      "epoch": 1.2346153846153847,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.9235,
+      "step": 642
+    },
+    {
+      "epoch": 1.2365384615384616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.4973134095838943e-06,
+      "loss": 0.9271,
+      "step": 643
+    },
+    {
+      "epoch": 1.2384615384615385,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.4821572818063544e-06,
+      "loss": 0.8456,
+      "step": 644
+    },
+    {
+      "epoch": 1.2403846153846154,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.4670165006319236e-06,
+      "loss": 0.9204,
+      "step": 645
+    },
+    {
+      "epoch": 1.2423076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.4518912191460073e-06,
+      "loss": 0.908,
+      "step": 646
+    },
+    {
+      "epoch": 1.2442307692307693,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.4367815902772917e-06,
+      "loss": 0.8835,
+      "step": 647
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.4216877667961975e-06,
+      "loss": 0.8741,
+      "step": 648
+    },
+    {
+      "epoch": 1.248076923076923,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.406609901313349e-06,
+      "loss": 0.9994,
+      "step": 649
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.3915481462780174e-06,
+      "loss": 0.8865,
+      "step": 650
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 0.9308112859725952,
+      "eval_runtime": 34.3336,
+      "eval_samples_per_second": 68.242,
+      "eval_steps_per_second": 17.068,
+      "step": 650
+    },
+    {
+      "epoch": 1.251923076923077,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 0.8766,
+      "step": 651
+    },
+    {
+      "epoch": 1.2538461538461538,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.3614735765310013e-06,
+      "loss": 0.8693,
+      "step": 652
+    },
+    {
+      "epoch": 1.2557692307692307,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.3464610658972584e-06,
+      "loss": 0.858,
+      "step": 653
+    },
+    {
+      "epoch": 1.2576923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.331465273863834e-06,
+      "loss": 0.8857,
+      "step": 654
+    },
+    {
+      "epoch": 1.2596153846153846,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.3164863520501744e-06,
+      "loss": 0.8839,
+      "step": 655
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.3015244519051525e-06,
+      "loss": 0.8891,
+      "step": 656
+    },
+    {
+      "epoch": 1.2634615384615384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.2865797247055354e-06,
+      "loss": 0.8891,
+      "step": 657
+    },
+    {
+      "epoch": 1.2653846153846153,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.2716523215544602e-06,
+      "loss": 0.8977,
+      "step": 658
+    },
+    {
+      "epoch": 1.2673076923076922,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.256742393379909e-06,
+      "loss": 0.9019,
+      "step": 659
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.2418500909331684e-06,
+      "loss": 0.8878,
+      "step": 660
+    },
+    {
+      "epoch": 1.271153846153846,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.8501,
+      "step": 661
+    },
+    {
+      "epoch": 1.273076923076923,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.21211896533572e-06,
+      "loss": 0.8878,
+      "step": 662
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.197280442790455e-06,
+      "loss": 0.9186,
+      "step": 663
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.1824601471808504e-06,
+      "loss": 0.8288,
+      "step": 664
+    },
+    {
+      "epoch": 1.2788461538461537,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.1676582283519454e-06,
+      "loss": 0.9064,
+      "step": 665
+    },
+    {
+      "epoch": 1.2807692307692307,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.1528748359629657e-06,
+      "loss": 0.8728,
+      "step": 666
+    },
+    {
+      "epoch": 1.2826923076923076,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.1381101194858264e-06,
+      "loss": 0.8885,
+      "step": 667
+    },
+    {
+      "epoch": 1.2846153846153845,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.1233642282036147e-06,
+      "loss": 0.9146,
+      "step": 668
+    },
+    {
+      "epoch": 1.2865384615384614,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.1086373112090762e-06,
+      "loss": 0.897,
+      "step": 669
+    },
+    {
+      "epoch": 1.2884615384615383,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.0939295174031127e-06,
+      "loss": 0.8649,
+      "step": 670
+    },
+    {
+      "epoch": 1.2903846153846155,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.079240995493279e-06,
+      "loss": 0.8316,
+      "step": 671
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.0645718939922668e-06,
+      "loss": 0.8968,
+      "step": 672
+    },
+    {
+      "epoch": 1.2942307692307693,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.049922361216422e-06,
+      "loss": 0.9268,
+      "step": 673
+    },
+    {
+      "epoch": 1.2961538461538462,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.03529254528423e-06,
+      "loss": 0.8527,
+      "step": 674
+    },
+    {
+      "epoch": 1.2980769230769231,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.0206825941148203e-06,
+      "loss": 0.8926,
+      "step": 675
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.006092655426481e-06,
+      "loss": 0.8785,
+      "step": 676
+    },
+    {
+      "epoch": 1.301923076923077,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.9372,
+      "step": 677
+    },
+    {
+      "epoch": 1.3038461538461539,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9769734053529443e-06,
+      "loss": 0.876,
+      "step": 678
+    },
+    {
+      "epoch": 1.3057692307692308,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9624443883866403e-06,
+      "loss": 0.9531,
+      "step": 679
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.881,
+      "step": 680
+    },
+    {
+      "epoch": 1.3096153846153846,
+      "grad_norm": 0.75,
+      "learning_rate": 2.9334483050933506e-06,
+      "loss": 0.9362,
+      "step": 681
+    },
+    {
+      "epoch": 1.3115384615384615,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9189815319399422e-06,
+      "loss": 0.9015,
+      "step": 682
+    },
+    {
+      "epoch": 1.3134615384615385,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.904535799546636e-06,
+      "loss": 0.9278,
+      "step": 683
+    },
+    {
+      "epoch": 1.3153846153846154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.890111253971327e-06,
+      "loss": 0.8761,
+      "step": 684
+    },
+    {
+      "epoch": 1.3173076923076923,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.8757080410577042e-06,
+      "loss": 0.8785,
+      "step": 685
+    },
+    {
+      "epoch": 1.3192307692307692,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.8613263064337617e-06,
+      "loss": 0.8794,
+      "step": 686
+    },
+    {
+      "epoch": 1.3211538461538461,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.846966195510332e-06,
+      "loss": 0.891,
+      "step": 687
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.8326278534796154e-06,
+      "loss": 0.903,
+      "step": 688
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.81831142531371e-06,
+      "loss": 0.9126,
+      "step": 689
+    },
+    {
+      "epoch": 1.3269230769230769,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.804017055763149e-06,
+      "loss": 0.8889,
+      "step": 690
+    },
+    {
+      "epoch": 1.3288461538461538,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.7897448893554335e-06,
+      "loss": 0.8606,
+      "step": 691
+    },
+    {
+      "epoch": 1.3307692307692307,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7754950703935735e-06,
+      "loss": 0.8373,
+      "step": 692
+    },
+    {
+      "epoch": 1.3326923076923076,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.761267742954629e-06,
+      "loss": 0.9246,
+      "step": 693
+    },
+    {
+      "epoch": 1.3346153846153845,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7470630508882525e-06,
+      "loss": 0.9381,
+      "step": 694
+    },
+    {
+      "epoch": 1.3365384615384617,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.7328811378152355e-06,
+      "loss": 0.8936,
+      "step": 695
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.718722147126054e-06,
+      "loss": 0.9314,
+      "step": 696
+    },
+    {
+      "epoch": 1.3403846153846155,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.704586221979422e-06,
+      "loss": 0.8957,
+      "step": 697
+    },
+    {
+      "epoch": 1.3423076923076924,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.6904735053008405e-06,
+      "loss": 0.8703,
+      "step": 698
+    },
+    {
+      "epoch": 1.3442307692307693,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.9244,
+      "step": 699
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.662318267875119e-06,
+      "loss": 0.8768,
+      "step": 700
+    },
+    {
+      "epoch": 1.3480769230769232,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6482760317999338e-06,
+      "loss": 0.9445,
+      "step": 701
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.634257573533833e-06,
+      "loss": 0.9024,
+      "step": 702
+    },
+    {
+      "epoch": 1.351923076923077,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.8766,
+      "step": 703
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.606292557138307e-06,
+      "loss": 0.889,
+      "step": 704
+    },
+    {
+      "epoch": 1.3557692307692308,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.592346281757552e-06,
+      "loss": 0.9081,
+      "step": 705
+    },
+    {
+      "epoch": 1.3576923076923078,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.5784243496803596e-06,
+      "loss": 0.8632,
+      "step": 706
+    },
+    {
+      "epoch": 1.3596153846153847,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5645269016685905e-06,
+      "loss": 0.8161,
+      "step": 707
+    },
+    {
+      "epoch": 1.3615384615384616,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.550654078236552e-06,
+      "loss": 0.8689,
+      "step": 708
+    },
+    {
+      "epoch": 1.3634615384615385,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.5368060196495785e-06,
+      "loss": 0.9514,
+      "step": 709
+    },
+    {
+      "epoch": 1.3653846153846154,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.5229828659226114e-06,
+      "loss": 0.8675,
+      "step": 710
+    },
+    {
+      "epoch": 1.3673076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.5091847568187834e-06,
+      "loss": 0.9018,
+      "step": 711
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.4954118318480063e-06,
+      "loss": 0.9049,
+      "step": 712
+    },
+    {
+      "epoch": 1.3711538461538462,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.4816642302655634e-06,
+      "loss": 0.8864,
+      "step": 713
+    },
+    {
+      "epoch": 1.373076923076923,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.4679420910706887e-06,
+      "loss": 0.9208,
+      "step": 714
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.454245553005184e-06,
+      "loss": 0.9696,
+      "step": 715
+    },
+    {
+      "epoch": 1.375,
+      "eval_loss": 0.9303730130195618,
+      "eval_runtime": 34.4007,
+      "eval_samples_per_second": 68.109,
+      "eval_steps_per_second": 17.035,
+      "step": 715
+    },
+    {
+      "epoch": 1.376923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.4405747545519966e-06,
+      "loss": 0.8986,
+      "step": 716
+    },
+    {
+      "epoch": 1.3788461538461538,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4269298339338205e-06,
+      "loss": 0.9069,
+      "step": 717
+    },
+    {
+      "epoch": 1.3807692307692307,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.8608,
+      "step": 718
+    },
+    {
+      "epoch": 1.3826923076923077,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3997181777836955e-06,
+      "loss": 0.9137,
+      "step": 719
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3861517173833347e-06,
+      "loss": 0.8828,
+      "step": 720
+    },
+    {
+      "epoch": 1.3865384615384615,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.3726116850783987e-06,
+      "loss": 0.9207,
+      "step": 721
+    },
+    {
+      "epoch": 1.3884615384615384,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.3590982177694348e-06,
+      "loss": 0.8221,
+      "step": 722
+    },
+    {
+      "epoch": 1.3903846153846153,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.3456114520883956e-06,
+      "loss": 0.8922,
+      "step": 723
+    },
+    {
+      "epoch": 1.3923076923076922,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.3321515243972663e-06,
+      "loss": 0.8462,
+      "step": 724
+    },
+    {
+      "epoch": 1.3942307692307692,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.318718570786675e-06,
+      "loss": 0.8804,
+      "step": 725
+    },
+    {
+      "epoch": 1.396153846153846,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3053127270745163e-06,
+      "loss": 0.8969,
+      "step": 726
+    },
+    {
+      "epoch": 1.398076923076923,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.2919341288045853e-06,
+      "loss": 0.9326,
+      "step": 727
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.27858291124521e-06,
+      "loss": 0.9407,
+      "step": 728
+    },
+    {
+      "epoch": 1.4019230769230768,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.8988,
+      "step": 729
+    },
+    {
+      "epoch": 1.4038461538461537,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.25196315794583e-06,
+      "loss": 0.8821,
+      "step": 730
+    },
+    {
+      "epoch": 1.4057692307692307,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.238694891352814e-06,
+      "loss": 0.8893,
+      "step": 731
+    },
+    {
+      "epoch": 1.4076923076923076,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2254545437615932e-06,
+      "loss": 0.9305,
+      "step": 732
+    },
+    {
+      "epoch": 1.4096153846153845,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.2122422490426676e-06,
+      "loss": 0.9039,
+      "step": 733
+    },
+    {
+      "epoch": 1.4115384615384614,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.199058140782897e-06,
+      "loss": 0.8692,
+      "step": 734
+    },
+    {
+      "epoch": 1.4134615384615383,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.1859023522841543e-06,
+      "loss": 0.9083,
+      "step": 735
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.172775016561977e-06,
+      "loss": 0.8858,
+      "step": 736
+    },
+    {
+      "epoch": 1.4173076923076924,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.8998,
+      "step": 737
+    },
+    {
+      "epoch": 1.4192307692307693,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.1466062340697234e-06,
+      "loss": 0.8965,
+      "step": 738
+    },
+    {
+      "epoch": 1.4211538461538462,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.1335650518869555e-06,
+      "loss": 0.8807,
+      "step": 739
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.120552851652694e-06,
+      "loss": 0.9295,
+      "step": 740
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.1075697649306838e-06,
+      "loss": 0.8947,
+      "step": 741
+    },
+    {
+      "epoch": 1.426923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.094615922990309e-06,
+      "loss": 0.9056,
+      "step": 742
+    },
+    {
+      "epoch": 1.4288461538461539,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0816914568052664e-06,
+      "loss": 0.9257,
+      "step": 743
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.0687964970522394e-06,
+      "loss": 0.9331,
+      "step": 744
+    },
+    {
+      "epoch": 1.4326923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.055931174109579e-06,
+      "loss": 0.9563,
+      "step": 745
+    },
+    {
+      "epoch": 1.4346153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.043095618055982e-06,
+      "loss": 0.8613,
+      "step": 746
+    },
+    {
+      "epoch": 1.4365384615384615,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.030289958669181e-06,
+      "loss": 0.8563,
+      "step": 747
+    },
+    {
+      "epoch": 1.4384615384615385,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.0175143254246277e-06,
+      "loss": 0.8817,
+      "step": 748
+    },
+    {
+      "epoch": 1.4403846153846154,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.004768847494186e-06,
+      "loss": 0.8463,
+      "step": 749
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.992053653744826e-06,
+      "loss": 0.9228,
+      "step": 750
+    },
+    {
+      "epoch": 1.4442307692307692,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.979368872737319e-06,
+      "loss": 0.8673,
+      "step": 751
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.966714632724941e-06,
+      "loss": 0.8408,
+      "step": 752
+    },
+    {
+      "epoch": 1.448076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.954091061652172e-06,
+      "loss": 0.9419,
+      "step": 753
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.941498287153409e-06,
+      "loss": 0.9032,
+      "step": 754
+    },
+    {
+      "epoch": 1.4519230769230769,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.9398,
+      "step": 755
+    },
+    {
+      "epoch": 1.4538461538461538,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.8877,
+      "step": 756
+    },
+    {
+      "epoch": 1.4557692307692307,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.903906014766681e-06,
+      "loss": 0.9077,
+      "step": 757
+    },
+    {
+      "epoch": 1.4576923076923076,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.891437696661015e-06,
+      "loss": 0.9012,
+      "step": 758
+    },
+    {
+      "epoch": 1.4596153846153845,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8790008086049534e-06,
+      "loss": 0.8972,
+      "step": 759
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8665954763453764e-06,
+      "loss": 0.885,
+      "step": 760
+    },
+    {
+      "epoch": 1.4634615384615386,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.854221825310103e-06,
+      "loss": 0.9224,
+      "step": 761
+    },
+    {
+      "epoch": 1.4653846153846155,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8418799806066413e-06,
+      "loss": 0.8654,
+      "step": 762
+    },
+    {
+      "epoch": 1.4673076923076924,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.829570067020906e-06,
+      "loss": 0.9112,
+      "step": 763
+    },
+    {
+      "epoch": 1.4692307692307693,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8172922090159578e-06,
+      "loss": 0.8584,
+      "step": 764
+    },
+    {
+      "epoch": 1.4711538461538463,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8050465307307602e-06,
+      "loss": 0.8674,
+      "step": 765
+    },
+    {
+      "epoch": 1.4730769230769232,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7928331559789087e-06,
+      "loss": 0.8901,
+      "step": 766
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7806522082473809e-06,
+      "loss": 0.9509,
+      "step": 767
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7685038106952952e-06,
+      "loss": 0.8837,
+      "step": 768
+    },
+    {
+      "epoch": 1.478846153846154,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7563880861526656e-06,
+      "loss": 0.9192,
+      "step": 769
+    },
+    {
+      "epoch": 1.4807692307692308,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7443051571191472e-06,
+      "loss": 0.8741,
+      "step": 770
+    },
+    {
+      "epoch": 1.4826923076923078,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.73225514576281e-06,
+      "loss": 0.9072,
+      "step": 771
+    },
+    {
+      "epoch": 1.4846153846153847,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7202381739189055e-06,
+      "loss": 0.851,
+      "step": 772
+    },
+    {
+      "epoch": 1.4865384615384616,
+      "grad_norm": 0.75,
+      "learning_rate": 1.70825436308862e-06,
+      "loss": 0.9552,
+      "step": 773
+    },
+    {
+      "epoch": 1.4884615384615385,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.696303834437859e-06,
+      "loss": 0.8915,
+      "step": 774
+    },
+    {
+      "epoch": 1.4903846153846154,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.9092,
+      "step": 775
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6725031066547786e-06,
+      "loss": 0.8589,
+      "step": 776
+    },
+    {
+      "epoch": 1.4942307692307693,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6606531481668364e-06,
+      "loss": 0.8642,
+      "step": 777
+    },
+    {
+      "epoch": 1.4961538461538462,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.648836953144755e-06,
+      "loss": 0.9987,
+      "step": 778
+    },
+    {
+      "epoch": 1.498076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6370546410597066e-06,
+      "loss": 0.8802,
+      "step": 779
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6253063310402833e-06,
+      "loss": 0.9119,
+      "step": 780
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 0.9302220344543457,
+      "eval_runtime": 34.2902,
+      "eval_samples_per_second": 68.329,
+      "eval_steps_per_second": 17.089,
+      "step": 780
+    },
+    {
+      "epoch": 1.501923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.8816,
+      "step": 781
+    },
+    {
+      "epoch": 1.5038461538461538,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.601912191992554e-06,
+      "loss": 0.9097,
+      "step": 782
+    },
+    {
+      "epoch": 1.5057692307692307,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.5902665994976896e-06,
+      "loss": 0.9058,
+      "step": 783
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5786554821329515e-06,
+      "loss": 0.8729,
+      "step": 784
+    },
+    {
+      "epoch": 1.5096153846153846,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.567078957296016e-06,
+      "loss": 0.8892,
+      "step": 785
+    },
+    {
+      "epoch": 1.5115384615384615,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5555371420348031e-06,
+      "loss": 0.8654,
+      "step": 786
+    },
+    {
+      "epoch": 1.5134615384615384,
+      "grad_norm": 0.75,
+      "learning_rate": 1.544030153046291e-06,
+      "loss": 0.8445,
+      "step": 787
+    },
+    {
+      "epoch": 1.5153846153846153,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5325581066753354e-06,
+      "loss": 0.8402,
+      "step": 788
+    },
+    {
+      "epoch": 1.5173076923076922,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5211211189134955e-06,
+      "loss": 0.8546,
+      "step": 789
+    },
+    {
+      "epoch": 1.5192307692307692,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5097193053978587e-06,
+      "loss": 0.9212,
+      "step": 790
+    },
+    {
+      "epoch": 1.521153846153846,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4983527814098736e-06,
+      "loss": 0.8601,
+      "step": 791
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4870216618741833e-06,
+      "loss": 0.8572,
+      "step": 792
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.475726061357463e-06,
+      "loss": 0.9179,
+      "step": 793
+    },
+    {
+      "epoch": 1.5269230769230768,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.9296,
+      "step": 794
+    },
+    {
+      "epoch": 1.5288461538461537,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4532418738508525e-06,
+      "loss": 0.8652,
+      "step": 795
+    },
+    {
+      "epoch": 1.5307692307692307,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.44205351419407e-06,
+      "loss": 0.8567,
+      "step": 796
+    },
+    {
+      "epoch": 1.5326923076923076,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.430901128220174e-06,
+      "loss": 0.9234,
+      "step": 797
+    },
+    {
+      "epoch": 1.5346153846153845,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4197848286887017e-06,
+      "loss": 0.8839,
+      "step": 798
+    },
+    {
+      "epoch": 1.5365384615384614,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4087047279943267e-06,
+      "loss": 0.9021,
+      "step": 799
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.397660938165723e-06,
+      "loss": 0.9017,
+      "step": 800
+    },
+    {
+      "epoch": 1.5403846153846152,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3866535708644335e-06,
+      "loss": 0.9315,
+      "step": 801
+    },
+    {
+      "epoch": 1.5423076923076922,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3756827373837396e-06,
+      "loss": 0.8562,
+      "step": 802
+    },
+    {
+      "epoch": 1.544230769230769,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3647485486475376e-06,
+      "loss": 0.9092,
+      "step": 803
+    },
+    {
+      "epoch": 1.546153846153846,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.353851115209215e-06,
+      "loss": 0.9142,
+      "step": 804
+    },
+    {
+      "epoch": 1.5480769230769231,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3429905472505344e-06,
+      "loss": 0.8867,
+      "step": 805
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3321669545805188e-06,
+      "loss": 0.9228,
+      "step": 806
+    },
+    {
+      "epoch": 1.551923076923077,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.8635,
+      "step": 807
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.310631132472222e-06,
+      "loss": 0.8625,
+      "step": 808
+    },
+    {
+      "epoch": 1.5557692307692308,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2999191207783129e-06,
+      "loss": 0.8367,
+      "step": 809
+    },
+    {
+      "epoch": 1.5576923076923077,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2892445198596198e-06,
+      "loss": 0.9199,
+      "step": 810
+    },
+    {
+      "epoch": 1.5596153846153846,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.27860743764489e-06,
+      "loss": 0.8648,
+      "step": 811
+    },
+    {
+      "epoch": 1.5615384615384615,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2680079816835228e-06,
+      "loss": 0.9395,
+      "step": 812
+    },
+    {
+      "epoch": 1.5634615384615385,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.8859,
+      "step": 813
+    },
+    {
+      "epoch": 1.5653846153846154,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2469223768152622e-06,
+      "loss": 0.884,
+      "step": 814
+    },
+    {
+      "epoch": 1.5673076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2364364411006841e-06,
+      "loss": 0.8908,
+      "step": 815
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2259885580219555e-06,
+      "loss": 0.871,
+      "step": 816
+    },
+    {
+      "epoch": 1.5711538461538461,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.215578833215526e-06,
+      "loss": 0.9022,
+      "step": 817
+    },
+    {
+      "epoch": 1.573076923076923,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2052073719320296e-06,
+      "loss": 0.9199,
+      "step": 818
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1948742790352342e-06,
+      "loss": 0.9211,
+      "step": 819
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1845796590009684e-06,
+      "loss": 0.9067,
+      "step": 820
+    },
+    {
+      "epoch": 1.578846153846154,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1743236159160654e-06,
+      "loss": 0.9342,
+      "step": 821
+    },
+    {
+      "epoch": 1.580769230769231,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1641062534773218e-06,
+      "loss": 0.915,
+      "step": 822
+    },
+    {
+      "epoch": 1.5826923076923078,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.15392767499044e-06,
+      "loss": 0.8886,
+      "step": 823
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1437879833689808e-06,
+      "loss": 0.9112,
+      "step": 824
+    },
+    {
+      "epoch": 1.5865384615384617,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.133687281133331e-06,
+      "loss": 0.9056,
+      "step": 825
+    },
+    {
+      "epoch": 1.5884615384615386,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1236256704096693e-06,
+      "loss": 0.91,
+      "step": 826
+    },
+    {
+      "epoch": 1.5903846153846155,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.113603252928917e-06,
+      "loss": 0.8725,
+      "step": 827
+    },
+    {
+      "epoch": 1.5923076923076924,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1036201300257266e-06,
+      "loss": 0.9173,
+      "step": 828
+    },
+    {
+      "epoch": 1.5942307692307693,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0936764026374547e-06,
+      "loss": 0.8928,
+      "step": 829
+    },
+    {
+      "epoch": 1.5961538461538463,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.083772171303128e-06,
+      "loss": 0.8897,
+      "step": 830
+    },
+    {
+      "epoch": 1.5980769230769232,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.073907536162443e-06,
+      "loss": 0.9053,
+      "step": 831
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.8655,
+      "step": 832
+    },
+    {
+      "epoch": 1.601923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.8698,
+      "step": 833
+    },
+    {
+      "epoch": 1.603846153846154,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0445522032879184e-06,
+      "loss": 0.8369,
+      "step": 834
+    },
+    {
+      "epoch": 1.6057692307692308,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0348469462966753e-06,
+      "loss": 0.8818,
+      "step": 835
+    },
+    {
+      "epoch": 1.6076923076923078,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0251817801722047e-06,
+      "loss": 0.837,
+      "step": 836
+    },
+    {
+      "epoch": 1.6096153846153847,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.0155568026370637e-06,
+      "loss": 0.8907,
+      "step": 837
+    },
+    {
+      "epoch": 1.6115384615384616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0059721110074678e-06,
+      "loss": 0.9129,
+      "step": 838
+    },
+    {
+      "epoch": 1.6134615384615385,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.964278021923107e-07,
+      "loss": 0.8743,
+      "step": 839
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.869239726921843e-07,
+      "loss": 0.8883,
+      "step": 840
+    },
+    {
+      "epoch": 1.6173076923076923,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.774607185984004e-07,
+      "loss": 0.904,
+      "step": 841
+    },
+    {
+      "epoch": 1.6192307692307693,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.68038135592022e-07,
+      "loss": 0.9311,
+      "step": 842
+    },
+    {
+      "epoch": 1.6211538461538462,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.586563189428954e-07,
+      "loss": 0.8621,
+      "step": 843
+    },
+    {
+      "epoch": 1.623076923076923,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.493153635086855e-07,
+      "loss": 0.8783,
+      "step": 844
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.400153637339182e-07,
+      "loss": 0.8793,
+      "step": 845
+    },
+    {
+      "epoch": 1.625,
+      "eval_loss": 0.9300703406333923,
+      "eval_runtime": 34.3126,
+      "eval_samples_per_second": 68.284,
+      "eval_steps_per_second": 17.078,
+      "step": 845
+    },
+    {
+      "epoch": 1.626923076923077,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.307564136490255e-07,
+      "loss": 0.8341,
+      "step": 846
+    },
+    {
+      "epoch": 1.6288461538461538,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.215386068693927e-07,
+      "loss": 0.8605,
+      "step": 847
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.123620365944147e-07,
+      "loss": 0.8513,
+      "step": 848
+    },
+    {
+      "epoch": 1.6326923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.032267956065516e-07,
+      "loss": 0.9004,
+      "step": 849
+    },
+    {
+      "epoch": 1.6346153846153846,
+      "grad_norm": 0.75,
+      "learning_rate": 8.941329762703921e-07,
+      "loss": 0.9814,
+      "step": 850
+    },
+    {
+      "epoch": 1.6365384615384615,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.9074,
+      "step": 851
+    },
+    {
+      "epoch": 1.6384615384615384,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.76069969916577e-07,
+      "loss": 0.9301,
+      "step": 852
+    },
+    {
+      "epoch": 1.6403846153846153,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.671009655303531e-07,
+      "loss": 0.9227,
+      "step": 853
+    },
+    {
+      "epoch": 1.6423076923076922,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.581737480568514e-07,
+      "loss": 0.8816,
+      "step": 854
+    },
+    {
+      "epoch": 1.6442307692307692,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.492884077573749e-07,
+      "loss": 0.8869,
+      "step": 855
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.404450344698167e-07,
+      "loss": 0.9363,
+      "step": 856
+    },
+    {
+      "epoch": 1.648076923076923,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.316437176077491e-07,
+      "loss": 0.8935,
+      "step": 857
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.228845461595225e-07,
+      "loss": 0.8805,
+      "step": 858
+    },
+    {
+      "epoch": 1.6519230769230768,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.8784,
+      "step": 859
+    },
+    {
+      "epoch": 1.6538461538461537,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.054929933264626e-07,
+      "loss": 0.9377,
+      "step": 860
+    },
+    {
+      "epoch": 1.6557692307692307,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.968607877841333e-07,
+      "loss": 0.8825,
+      "step": 861
+    },
+    {
+      "epoch": 1.6576923076923076,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.882710793388643e-07,
+      "loss": 0.915,
+      "step": 862
+    },
+    {
+      "epoch": 1.6596153846153845,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.79723954839477e-07,
+      "loss": 0.9108,
+      "step": 863
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.712195007042322e-07,
+      "loss": 0.8689,
+      "step": 864
+    },
+    {
+      "epoch": 1.6634615384615383,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.627578029199562e-07,
+      "loss": 0.8841,
+      "step": 865
+    },
+    {
+      "epoch": 1.6653846153846152,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.543389470411772e-07,
+      "loss": 0.9736,
+      "step": 866
+    },
+    {
+      "epoch": 1.6673076923076922,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.459630181892608e-07,
+      "loss": 0.9249,
+      "step": 867
+    },
+    {
+      "epoch": 1.669230769230769,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.376301010515397e-07,
+      "loss": 0.9258,
+      "step": 868
+    },
+    {
+      "epoch": 1.671153846153846,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.293402798804667e-07,
+      "loss": 0.909,
+      "step": 869
+    },
+    {
+      "epoch": 1.6730769230769231,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.888,
+      "step": 870
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.128902602685617e-07,
+      "loss": 0.9196,
+      "step": 871
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.047302281505735e-07,
+      "loss": 0.943,
+      "step": 872
+    },
+    {
+      "epoch": 1.6788461538461539,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.966136246432492e-07,
+      "loss": 0.9123,
+      "step": 873
+    },
+    {
+      "epoch": 1.6807692307692308,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.885405318119342e-07,
+      "loss": 0.8793,
+      "step": 874
+    },
+    {
+      "epoch": 1.6826923076923077,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.805110312820501e-07,
+      "loss": 0.9478,
+      "step": 875
+    },
+    {
+      "epoch": 1.6846153846153846,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.725252042382691e-07,
+      "loss": 0.8718,
+      "step": 876
+    },
+    {
+      "epoch": 1.6865384615384615,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.645831314236817e-07,
+      "loss": 0.8476,
+      "step": 877
+    },
+    {
+      "epoch": 1.6884615384615385,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.566848931389935e-07,
+      "loss": 0.9334,
+      "step": 878
+    },
+    {
+      "epoch": 1.6903846153846154,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.488305692417074e-07,
+      "loss": 0.8902,
+      "step": 879
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.410202391453157e-07,
+      "loss": 0.8728,
+      "step": 880
+    },
+    {
+      "epoch": 1.6942307692307692,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.332539818184985e-07,
+      "loss": 0.8788,
+      "step": 881
+    },
+    {
+      "epoch": 1.6961538461538461,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.255318757843249e-07,
+      "loss": 0.8917,
+      "step": 882
+    },
+    {
+      "epoch": 1.698076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.178539991194599e-07,
+      "loss": 0.8898,
+      "step": 883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.102204294533731e-07,
+      "loss": 0.894,
+      "step": 884
+    },
+    {
+      "epoch": 1.7019230769230769,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.882,
+      "step": 885
+    },
+    {
+      "epoch": 1.703846153846154,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.95086519394738e-07,
+      "loss": 0.917,
+      "step": 886
+    },
+    {
+      "epoch": 1.705769230769231,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.875863320181175e-07,
+      "loss": 0.8931,
+      "step": 887
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.801307576705833e-07,
+      "loss": 0.8623,
+      "step": 888
+    },
+    {
+      "epoch": 1.7096153846153848,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.8815,
+      "step": 889
+    },
+    {
+      "epoch": 1.7115384615384617,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.653537491382011e-07,
+      "loss": 0.9418,
+      "step": 890
+    },
+    {
+      "epoch": 1.7134615384615386,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.58032464360721e-07,
+      "loss": 0.8381,
+      "step": 891
+    },
+    {
+      "epoch": 1.7153846153846155,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.507560914255516e-07,
+      "loss": 0.9239,
+      "step": 892
+    },
+    {
+      "epoch": 1.7173076923076924,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.435247039026398e-07,
+      "loss": 0.9262,
+      "step": 893
+    },
+    {
+      "epoch": 1.7192307692307693,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.363383749070939e-07,
+      "loss": 0.844,
+      "step": 894
+    },
+    {
+      "epoch": 1.7211538461538463,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.291971770984428e-07,
+      "loss": 0.9041,
+      "step": 895
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.221011826799055e-07,
+      "loss": 0.8342,
+      "step": 896
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.150504633976572e-07,
+      "loss": 0.8119,
+      "step": 897
+    },
+    {
+      "epoch": 1.726923076923077,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.080450905401057e-07,
+      "loss": 0.9174,
+      "step": 898
+    },
+    {
+      "epoch": 1.728846153846154,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.010851349371704e-07,
+      "loss": 0.8972,
+      "step": 899
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.941706669595647e-07,
+      "loss": 0.9052,
+      "step": 900
+    },
+    {
+      "epoch": 1.7326923076923078,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.873017565180871e-07,
+      "loss": 0.8823,
+      "step": 901
+    },
+    {
+      "epoch": 1.7346153846153847,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.804784730629131e-07,
+      "loss": 0.8687,
+      "step": 902
+    },
+    {
+      "epoch": 1.7365384615384616,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.7370088558289175e-07,
+      "loss": 0.8865,
+      "step": 903
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.6696906260485007e-07,
+      "loss": 0.8775,
+      "step": 904
+    },
+    {
+      "epoch": 1.7403846153846154,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.602830721928997e-07,
+      "loss": 0.8814,
+      "step": 905
+    },
+    {
+      "epoch": 1.7423076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.536429819477478e-07,
+      "loss": 0.9094,
+      "step": 906
+    },
+    {
+      "epoch": 1.7442307692307693,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.4704885900601236e-07,
+      "loss": 0.9433,
+      "step": 907
+    },
+    {
+      "epoch": 1.7461538461538462,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.8625,
+      "step": 908
+    },
+    {
+      "epoch": 1.748076923076923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.33998781254773e-07,
+      "loss": 0.8546,
+      "step": 909
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.2754295839198325e-07,
+      "loss": 0.9265,
+      "step": 910
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.9300666451454163,
+      "eval_runtime": 34.3132,
+      "eval_samples_per_second": 68.283,
+      "eval_steps_per_second": 17.078,
+      "step": 910
+    },
+    {
+      "epoch": 1.751923076923077,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.8753,
+      "step": 911
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.147700710590563e-07,
+      "loss": 0.8862,
+      "step": 912
+    },
+    {
+      "epoch": 1.7557692307692307,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.0845313573301736e-07,
+      "loss": 0.9027,
+      "step": 913
+    },
+    {
+      "epoch": 1.7576923076923077,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.021826246158628e-07,
+      "loss": 0.9113,
+      "step": 914
+    },
+    {
+      "epoch": 1.7596153846153846,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.959586011074729e-07,
+      "loss": 0.8939,
+      "step": 915
+    },
+    {
+      "epoch": 1.7615384615384615,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.8978112813769786e-07,
+      "loss": 0.9124,
+      "step": 916
+    },
+    {
+      "epoch": 1.7634615384615384,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.836502681657289e-07,
+      "loss": 0.9074,
+      "step": 917
+    },
+    {
+      "epoch": 1.7653846153846153,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.7756608317946144e-07,
+      "loss": 0.8721,
+      "step": 918
+    },
+    {
+      "epoch": 1.7673076923076922,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.715286346948671e-07,
+      "loss": 0.9188,
+      "step": 919
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.6553798375537574e-07,
+      "loss": 0.8737,
+      "step": 920
+    },
+    {
+      "epoch": 1.771153846153846,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.595941909312595e-07,
+      "loss": 0.9252,
+      "step": 921
+    },
+    {
+      "epoch": 1.773076923076923,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.5369731631901214e-07,
+      "loss": 0.921,
+      "step": 922
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.4784741954074884e-07,
+      "loss": 0.8851,
+      "step": 923
+    },
+    {
+      "epoch": 1.7769230769230768,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.420445597436056e-07,
+      "loss": 0.8651,
+      "step": 924
+    },
+    {
+      "epoch": 1.7788461538461537,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.362887955991301e-07,
+      "loss": 0.902,
+      "step": 925
+    },
+    {
+      "epoch": 1.7807692307692307,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.305801853026985e-07,
+      "loss": 0.8853,
+      "step": 926
+    },
+    {
+      "epoch": 1.7826923076923076,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.8499,
+      "step": 927
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.193046566510777e-07,
+      "loss": 0.8928,
+      "step": 928
+    },
+    {
+      "epoch": 1.7865384615384614,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.1373785230049356e-07,
+      "loss": 0.8892,
+      "step": 929
+    },
+    {
+      "epoch": 1.7884615384615383,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.0821842980601756e-07,
+      "loss": 0.8887,
+      "step": 930
+    },
+    {
+      "epoch": 1.7903846153846152,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.0274644497342133e-07,
+      "loss": 0.9069,
+      "step": 931
+    },
+    {
+      "epoch": 1.7923076923076922,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9732195312884515e-07,
+      "loss": 0.9282,
+      "step": 932
+    },
+    {
+      "epoch": 1.794230769230769,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.91945009118238e-07,
+      "loss": 0.8736,
+      "step": 933
+    },
+    {
+      "epoch": 1.796153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.866156673068016e-07,
+      "loss": 0.9549,
+      "step": 934
+    },
+    {
+      "epoch": 1.7980769230769231,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.813339815784416e-07,
+      "loss": 0.8407,
+      "step": 935
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.76100005335222e-07,
+      "loss": 0.9366,
+      "step": 936
+    },
+    {
+      "epoch": 1.801923076923077,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.8663,
+      "step": 937
+    },
+    {
+      "epoch": 1.8038461538461539,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.657753925000228e-07,
+      "loss": 0.8676,
+      "step": 938
+    },
+    {
+      "epoch": 1.8057692307692308,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.6068486029813154e-07,
+      "loss": 0.8491,
+      "step": 939
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.556422463605024e-07,
+      "loss": 0.9075,
+      "step": 940
+    },
+    {
+      "epoch": 1.8096153846153846,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.506476016719922e-07,
+      "loss": 0.8678,
+      "step": 941
+    },
+    {
+      "epoch": 1.8115384615384615,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4570097673245197e-07,
+      "loss": 0.921,
+      "step": 942
+    },
+    {
+      "epoch": 1.8134615384615385,
+      "grad_norm": 0.75,
+      "learning_rate": 2.4080242155621327e-07,
+      "loss": 0.9795,
+      "step": 943
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.3595198567158473e-07,
+      "loss": 0.9078,
+      "step": 944
+    },
+    {
+      "epoch": 1.8173076923076923,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3114971812034981e-07,
+      "loss": 0.8915,
+      "step": 945
+    },
+    {
+      "epoch": 1.8192307692307692,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.8709,
+      "step": 946
+    },
+    {
+      "epoch": 1.8211538461538461,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.2168988174960382e-07,
+      "loss": 0.9404,
+      "step": 947
+    },
+    {
+      "epoch": 1.823076923076923,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.1703240857659958e-07,
+      "loss": 0.8965,
+      "step": 948
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.124232950290367e-07,
+      "loss": 0.8497,
+      "step": 949
+    },
+    {
+      "epoch": 1.8269230769230769,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.0786258770873647e-07,
+      "loss": 0.9354,
+      "step": 950
+    },
+    {
+      "epoch": 1.828846153846154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.0335033272809612e-07,
+      "loss": 0.8639,
+      "step": 951
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9888657570961924e-07,
+      "loss": 0.8492,
+      "step": 952
+    },
+    {
+      "epoch": 1.8326923076923078,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9447136178545766e-07,
+      "loss": 0.929,
+      "step": 953
+    },
+    {
+      "epoch": 1.8346153846153848,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9010473559695376e-07,
+      "loss": 0.8714,
+      "step": 954
+    },
+    {
+      "epoch": 1.8365384615384617,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.857867412941883e-07,
+      "loss": 0.9262,
+      "step": 955
+    },
+    {
+      "epoch": 1.8384615384615386,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8151742253553483e-07,
+      "loss": 0.8796,
+      "step": 956
+    },
+    {
+      "epoch": 1.8403846153846155,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7729682248721848e-07,
+      "loss": 0.8677,
+      "step": 957
+    },
+    {
+      "epoch": 1.8423076923076924,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.731249838228799e-07,
+      "loss": 0.8699,
+      "step": 958
+    },
+    {
+      "epoch": 1.8442307692307693,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.69001948723142e-07,
+      "loss": 0.8765,
+      "step": 959
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.649277588751863e-07,
+      "loss": 0.9203,
+      "step": 960
+    },
+    {
+      "epoch": 1.8480769230769232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6090245547232707e-07,
+      "loss": 0.9009,
+      "step": 961
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5692607921360014e-07,
+      "loss": 0.8951,
+      "step": 962
+    },
+    {
+      "epoch": 1.851923076923077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.938,
+      "step": 963
+    },
+    {
+      "epoch": 1.853846153846154,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.491202684508136e-07,
+      "loss": 0.9381,
+      "step": 964
+    },
+    {
+      "epoch": 1.8557692307692308,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.9324,
+      "step": 965
+    },
+    {
+      "epoch": 1.8576923076923078,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.415106422779733e-07,
+      "loss": 0.881,
+      "step": 966
+    },
+    {
+      "epoch": 1.8596153846153847,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3777949489706898e-07,
+      "loss": 0.8983,
+      "step": 967
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3409750845191138e-07,
+      "loss": 0.9162,
+      "step": 968
+    },
+    {
+      "epoch": 1.8634615384615385,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3046472017032685e-07,
+      "loss": 0.8893,
+      "step": 969
+    },
+    {
+      "epoch": 1.8653846153846154,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2688116678270636e-07,
+      "loss": 0.9097,
+      "step": 970
+    },
+    {
+      "epoch": 1.8673076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2334688452164122e-07,
+      "loss": 0.9282,
+      "step": 971
+    },
+    {
+      "epoch": 1.8692307692307693,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.198619091215497e-07,
+      "loss": 0.8906,
+      "step": 972
+    },
+    {
+      "epoch": 1.8711538461538462,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1642627581831767e-07,
+      "loss": 0.9165,
+      "step": 973
+    },
+    {
+      "epoch": 1.873076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1304001934894393e-07,
+      "loss": 0.9424,
+      "step": 974
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0970317395119001e-07,
+      "loss": 0.9375,
+      "step": 975
+    },
+    {
+      "epoch": 1.875,
+      "eval_loss": 0.9301103949546814,
+      "eval_runtime": 34.343,
+      "eval_samples_per_second": 68.223,
+      "eval_steps_per_second": 17.063,
+      "step": 975
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.0641577336322761e-07,
+      "loss": 0.8467,
+      "step": 976
+    },
+    {
+      "epoch": 1.8788461538461538,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.0317785082330555e-07,
+      "loss": 0.9042,
+      "step": 977
+    },
+    {
+      "epoch": 1.8807692307692307,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.998943906941005e-08,
+      "loss": 0.8585,
+      "step": 978
+    },
+    {
+      "epoch": 1.8826923076923077,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.685057033892998e-08,
+      "loss": 0.9187,
+      "step": 979
+    },
+    {
+      "epoch": 1.8846153846153846,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.376127636833876e-08,
+      "loss": 0.8607,
+      "step": 980
+    },
+    {
+      "epoch": 1.8865384615384615,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.072158839286748e-08,
+      "loss": 0.9251,
+      "step": 981
+    },
+    {
+      "epoch": 1.8884615384615384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.773153714619064e-08,
+      "loss": 0.9172,
+      "step": 982
+    },
+    {
+      "epoch": 1.8903846153846153,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.479115286011752e-08,
+      "loss": 0.9146,
+      "step": 983
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.8388,
+      "step": 984
+    },
+    {
+      "epoch": 1.8942307692307692,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.905950358584768e-08,
+      "loss": 0.8973,
+      "step": 985
+    },
+    {
+      "epoch": 1.896153846153846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.626829654920732e-08,
+      "loss": 0.9986,
+      "step": 986
+    },
+    {
+      "epoch": 1.898076923076923,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.352687237569489e-08,
+      "loss": 0.9084,
+      "step": 987
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.08352587833e-08,
+      "loss": 0.8867,
+      "step": 988
+    },
+    {
+      "epoch": 1.9019230769230768,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.9154,
+      "step": 989
+    },
+    {
+      "epoch": 1.9038461538461537,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.560157169542391e-08,
+      "loss": 0.8247,
+      "step": 990
+    },
+    {
+      "epoch": 1.9057692307692307,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.305955111670204e-08,
+      "loss": 0.9041,
+      "step": 991
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.056744695208283e-08,
+      "loss": 0.9043,
+      "step": 992
+    },
+    {
+      "epoch": 1.9096153846153845,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.8125284398730666e-08,
+      "loss": 0.8587,
+      "step": 993
+    },
+    {
+      "epoch": 1.9115384615384614,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.573308814886158e-08,
+      "loss": 0.9092,
+      "step": 994
+    },
+    {
+      "epoch": 1.9134615384615383,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.339088238949186e-08,
+      "loss": 0.8529,
+      "step": 995
+    },
+    {
+      "epoch": 1.9153846153846152,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.109869080219376e-08,
+      "loss": 0.9078,
+      "step": 996
+    },
+    {
+      "epoch": 1.9173076923076922,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.885653656285627e-08,
+      "loss": 0.8423,
+      "step": 997
+    },
+    {
+      "epoch": 1.919230769230769,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.666444234145084e-08,
+      "loss": 0.8814,
+      "step": 998
+    },
+    {
+      "epoch": 1.921153846153846,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.45224303018027e-08,
+      "loss": 0.8811,
+      "step": 999
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.2430522101364894e-08,
+      "loss": 0.9339,
+      "step": 1000
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.038873889100237e-08,
+      "loss": 0.8606,
+      "step": 1001
+    },
+    {
+      "epoch": 1.926923076923077,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.839710131477492e-08,
+      "loss": 0.9095,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9288461538461539,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.9305,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9307692307692308,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.456434310570023e-08,
+      "loss": 0.9045,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9326923076923077,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.2723261225102164e-08,
+      "loss": 0.8942,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9346153846153846,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.093240248274565e-08,
+      "loss": 0.8703,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9365384615384615,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9191784985644345e-08,
+      "loss": 0.9141,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.75,
+      "learning_rate": 2.7501426332831594e-08,
+      "loss": 0.8912,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9403846153846154,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.5861343615184997e-08,
+      "loss": 0.8769,
+      "step": 1009
+    },
+    {
+      "epoch": 1.9423076923076923,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.427155341525156e-08,
+      "loss": 0.8312,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9442307692307692,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.2732071807081147e-08,
+      "loss": 0.8963,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9461538461538461,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.1242914356063292e-08,
+      "loss": 0.9095,
+      "step": 1012
+    },
+    {
+      "epoch": 1.948076923076923,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.98040961187701e-08,
+      "loss": 0.9588,
+      "step": 1013
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.841563164280413e-08,
+      "loss": 0.8922,
+      "step": 1014
+    },
+    {
+      "epoch": 1.9519230769230769,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.8975,
+      "step": 1015
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5789819619537182e-08,
+      "loss": 0.8766,
+      "step": 1016
+    },
+    {
+      "epoch": 1.955769230769231,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4552498621295264e-08,
+      "loss": 0.9691,
+      "step": 1017
+    },
+    {
+      "epoch": 1.9576923076923078,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3365584482228356e-08,
+      "loss": 0.8297,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9596153846153848,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2229089202987487e-08,
+      "loss": 0.9212,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1143024274448689e-08,
+      "loss": 0.9787,
+      "step": 1020
+    },
+    {
+      "epoch": 1.9634615384615386,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0107400677596413e-08,
+      "loss": 0.9338,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9653846153846155,
+      "grad_norm": 0.75,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.901,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9673076923076924,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.187518852771914e-09,
+      "loss": 0.8544,
+      "step": 1023
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.3032800363398435e-09,
+      "loss": 0.892,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9711538461538463,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.469521374477539e-09,
+      "loss": 0.8719,
+      "step": 1025
+    },
+    {
+      "epoch": 1.9730769230769232,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.686251297151724e-09,
+      "loss": 0.9165,
+      "step": 1026
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.9534777238485764e-09,
+      "loss": 0.9102,
+      "step": 1027
+    },
+    {
+      "epoch": 1.976923076923077,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.2712080634949024e-09,
+      "loss": 0.9305,
+      "step": 1028
+    },
+    {
+      "epoch": 1.978846153846154,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.6394492143820847e-09,
+      "loss": 0.9779,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9807692307692308,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.0582075640972487e-09,
+      "loss": 0.904,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9826923076923078,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5274889894583156e-09,
+      "loss": 0.9295,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0472988564540496e-09,
+      "loss": 0.8869,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9865384615384616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6176420201902132e-09,
+      "loss": 0.9327,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9884615384615385,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2385228248407155e-09,
+      "loss": 0.9076,
+      "step": 1034
+    },
+    {
+      "epoch": 1.9903846153846154,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.099451036048701e-10,
+      "loss": 0.9368,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9923076923076923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.319121786646509e-10,
+      "loss": 0.8626,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9942307692307693,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.0442686115582665e-10,
+      "loss": 0.9733,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9961538461538462,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.274914511374293e-10,
+      "loss": 0.824,
+      "step": 1038
+    },
+    {
+      "epoch": 1.998076923076923,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.011077375662195e-10,
+      "loss": 0.8492,
+      "step": 1039
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5276998284473608e-11,
+      "loss": 0.8473,
+      "step": 1040
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.930040180683136,
+      "eval_runtime": 34.5175,
+      "eval_samples_per_second": 67.879,
+      "eval_steps_per_second": 16.977,
+      "step": 1040
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1040,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 520,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.943620643240018e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}