diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4287 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 50,
+  "global_step": 594,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005050505050505051,
+      "grad_norm": 2.061766753199945,
+      "learning_rate": 1.0344827586206896e-06,
+      "loss": 1.9091,
+      "step": 1
+    },
+    {
+      "epoch": 0.005050505050505051,
+      "eval_loss": 1.7296912670135498,
+      "eval_runtime": 229.6516,
+      "eval_samples_per_second": 6.231,
+      "eval_steps_per_second": 0.779,
+      "step": 1
+    },
+    {
+      "epoch": 0.010101010101010102,
+      "grad_norm": 0.9340006537063892,
+      "learning_rate": 2.068965517241379e-06,
+      "loss": 1.9543,
+      "step": 2
+    },
+    {
+      "epoch": 0.015151515151515152,
+      "grad_norm": 0.8730658898573816,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 1.8991,
+      "step": 3
+    },
+    {
+      "epoch": 0.020202020202020204,
+      "grad_norm": 0.9760544939810394,
+      "learning_rate": 4.137931034482758e-06,
+      "loss": 1.9469,
+      "step": 4
+    },
+    {
+      "epoch": 0.025252525252525252,
+      "grad_norm": 1.1435345547054372,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 1.875,
+      "step": 5
+    },
+    {
+      "epoch": 0.030303030303030304,
+      "grad_norm": 0.6977064869142138,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 1.9303,
+      "step": 6
+    },
+    {
+      "epoch": 0.03535353535353535,
+      "grad_norm": 0.9852376303852637,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 1.9764,
+      "step": 7
+    },
+    {
+      "epoch": 0.04040404040404041,
+      "grad_norm": 1.5021550676895192,
+      "learning_rate": 8.275862068965517e-06,
+      "loss": 1.9526,
+      "step": 8
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 14.465654024779864,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 1.9238,
+      "step": 9
+    },
+    {
+      "epoch": 0.050505050505050504,
+      "grad_norm": 1.3153716741597974,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 1.8548,
+      "step": 10
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 0.7251965259785947,
+      "learning_rate": 1.1379310344827586e-05,
+      "loss": 1.9568,
+      "step": 11
+    },
+    {
+      "epoch": 0.06060606060606061,
+      "grad_norm": 0.675323305346893,
+      "learning_rate": 1.2413793103448277e-05,
+      "loss": 2.0394,
+      "step": 12
+    },
+    {
+      "epoch": 0.06565656565656566,
+      "grad_norm": 0.7284662389076806,
+      "learning_rate": 1.3448275862068966e-05,
+      "loss": 1.9143,
+      "step": 13
+    },
+    {
+      "epoch": 0.0707070707070707,
+      "grad_norm": 0.9273147668903492,
+      "learning_rate": 1.4482758620689657e-05,
+      "loss": 1.8514,
+      "step": 14
+    },
+    {
+      "epoch": 0.07575757575757576,
+      "grad_norm": 0.8045697974814354,
+      "learning_rate": 1.5517241379310346e-05,
+      "loss": 1.8691,
+      "step": 15
+    },
+    {
+      "epoch": 0.08080808080808081,
+      "grad_norm": 0.8617824816372583,
+      "learning_rate": 1.6551724137931033e-05,
+      "loss": 1.9208,
+      "step": 16
+    },
+    {
+      "epoch": 0.08585858585858586,
+      "grad_norm": 0.8474189127719305,
+      "learning_rate": 1.7586206896551724e-05,
+      "loss": 1.971,
+      "step": 17
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 0.7802161578292038,
+      "learning_rate": 1.8620689655172415e-05,
+      "loss": 1.9364,
+      "step": 18
+    },
+    {
+      "epoch": 0.09595959595959595,
+      "grad_norm": 0.7670275224618359,
+      "learning_rate": 1.9655172413793102e-05,
+      "loss": 1.8924,
+      "step": 19
+    },
+    {
+      "epoch": 0.10101010101010101,
+      "grad_norm": 0.8871142093374335,
+      "learning_rate": 2.0689655172413797e-05,
+      "loss": 1.8666,
+      "step": 20
+    },
+    {
+      "epoch": 0.10606060606060606,
+      "grad_norm": 0.8537219904851733,
+      "learning_rate": 2.1724137931034484e-05,
+      "loss": 1.903,
+      "step": 21
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.7103385523680423,
+      "learning_rate": 2.275862068965517e-05,
+      "loss": 1.8842,
+      "step": 22
+    },
+    {
+      "epoch": 0.11616161616161616,
+      "grad_norm": 0.7687652688491076,
+      "learning_rate": 2.3793103448275862e-05,
+      "loss": 1.9043,
+      "step": 23
+    },
+    {
+      "epoch": 0.12121212121212122,
+      "grad_norm": 0.8618641553034271,
+      "learning_rate": 2.4827586206896553e-05,
+      "loss": 1.9657,
+      "step": 24
+    },
+    {
+      "epoch": 0.12626262626262627,
+      "grad_norm": 0.6952586490192444,
+      "learning_rate": 2.586206896551724e-05,
+      "loss": 1.8744,
+      "step": 25
+    },
+    {
+      "epoch": 0.13131313131313133,
+      "grad_norm": 0.7879308602246046,
+      "learning_rate": 2.689655172413793e-05,
+      "loss": 1.9088,
+      "step": 26
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 0.6286619656859633,
+      "learning_rate": 2.793103448275862e-05,
+      "loss": 1.8769,
+      "step": 27
+    },
+    {
+      "epoch": 0.1414141414141414,
+      "grad_norm": 0.6368137307786529,
+      "learning_rate": 2.8965517241379313e-05,
+      "loss": 1.9609,
+      "step": 28
+    },
+    {
+      "epoch": 0.14646464646464646,
+      "grad_norm": 0.6997236575440393,
+      "learning_rate": 3e-05,
+      "loss": 1.8216,
+      "step": 29
+    },
+    {
+      "epoch": 0.15151515151515152,
+      "grad_norm": 0.7017124666486956,
+      "learning_rate": 2.999976812015884e-05,
+      "loss": 1.8589,
+      "step": 30
+    },
+    {
+      "epoch": 0.15656565656565657,
+      "grad_norm": 0.6473032006817286,
+      "learning_rate": 2.999907248780446e-05,
+      "loss": 1.8554,
+      "step": 31
+    },
+    {
+      "epoch": 0.16161616161616163,
+      "grad_norm": 0.5625959900004289,
+      "learning_rate": 2.9997913124443945e-05,
+      "loss": 1.8885,
+      "step": 32
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.5416656043411677,
+      "learning_rate": 2.9996290065921693e-05,
+      "loss": 1.8011,
+      "step": 33
+    },
+    {
+      "epoch": 0.1717171717171717,
+      "grad_norm": 0.6288840271054257,
+      "learning_rate": 2.9994203362418313e-05,
+      "loss": 1.8941,
+      "step": 34
+    },
+    {
+      "epoch": 0.17676767676767677,
+      "grad_norm": 0.46613508572576606,
+      "learning_rate": 2.9991653078449062e-05,
+      "loss": 1.8755,
+      "step": 35
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.483380275460621,
+      "learning_rate": 2.998863929286187e-05,
+      "loss": 1.7971,
+      "step": 36
+    },
+    {
+      "epoch": 0.18686868686868688,
+      "grad_norm": 0.5435079019515928,
+      "learning_rate": 2.9985162098834886e-05,
+      "loss": 1.8561,
+      "step": 37
+    },
+    {
+      "epoch": 0.1919191919191919,
+      "grad_norm": 0.39935840982328513,
+      "learning_rate": 2.9981221603873608e-05,
+      "loss": 1.8447,
+      "step": 38
+    },
+    {
+      "epoch": 0.19696969696969696,
+      "grad_norm": 0.4860381910271567,
+      "learning_rate": 2.9976817929807542e-05,
+      "loss": 1.8571,
+      "step": 39
+    },
+    {
+      "epoch": 0.20202020202020202,
+      "grad_norm": 0.4704297980261362,
+      "learning_rate": 2.9971951212786453e-05,
+      "loss": 1.8953,
+      "step": 40
+    },
+    {
+      "epoch": 0.20707070707070707,
+      "grad_norm": 0.45313219770954005,
+      "learning_rate": 2.996662160327616e-05,
+      "loss": 1.8858,
+      "step": 41
+    },
+    {
+      "epoch": 0.21212121212121213,
+      "grad_norm": 0.4849523903159801,
+      "learning_rate": 2.9960829266053854e-05,
+      "loss": 1.8558,
+      "step": 42
+    },
+    {
+      "epoch": 0.21717171717171718,
+      "grad_norm": 0.5296271329297302,
+      "learning_rate": 2.9954574380203036e-05,
+      "loss": 1.8819,
+      "step": 43
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.42138253356228195,
+      "learning_rate": 2.9947857139107964e-05,
+      "loss": 1.8218,
+      "step": 44
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 0.54778300283403,
+      "learning_rate": 2.994067775044768e-05,
+      "loss": 1.8718,
+      "step": 45
+    },
+    {
+      "epoch": 0.23232323232323232,
+      "grad_norm": 0.5171055027316829,
+      "learning_rate": 2.9933036436189582e-05,
+      "loss": 1.8202,
+      "step": 46
+    },
+    {
+      "epoch": 0.23737373737373738,
+      "grad_norm": 0.42755509411252407,
+      "learning_rate": 2.992493343258257e-05,
+      "loss": 1.7941,
+      "step": 47
+    },
+    {
+      "epoch": 0.24242424242424243,
+      "grad_norm": 0.5864597089331413,
+      "learning_rate": 2.9916368990149738e-05,
+      "loss": 1.8177,
+      "step": 48
+    },
+    {
+      "epoch": 0.2474747474747475,
+      "grad_norm": 0.5015977034643889,
+      "learning_rate": 2.990734337368062e-05,
+      "loss": 1.8441,
+      "step": 49
+    },
+    {
+      "epoch": 0.25252525252525254,
+      "grad_norm": 0.4223650022421803,
+      "learning_rate": 2.989785686222302e-05,
+      "loss": 1.8235,
+      "step": 50
+    },
+    {
+      "epoch": 0.25252525252525254,
+      "eval_loss": 1.5191664695739746,
+      "eval_runtime": 228.7897,
+      "eval_samples_per_second": 6.255,
+      "eval_steps_per_second": 0.782,
+      "step": 50
+    },
+    {
+      "epoch": 0.25757575757575757,
+      "grad_norm": 0.5549092161535342,
+      "learning_rate": 2.9887909749074373e-05,
+      "loss": 1.8724,
+      "step": 51
+    },
+    {
+      "epoch": 0.26262626262626265,
+      "grad_norm": 0.5808532765316217,
+      "learning_rate": 2.9877502341772687e-05,
+      "loss": 1.8668,
+      "step": 52
+    },
+    {
+      "epoch": 0.2676767676767677,
+      "grad_norm": 0.42037362125882416,
+      "learning_rate": 2.9866634962087014e-05,
+      "loss": 1.8149,
+      "step": 53
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 0.510701064713823,
+      "learning_rate": 2.9855307946007532e-05,
+      "loss": 1.8388,
+      "step": 54
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 0.4740044982938013,
+      "learning_rate": 2.984352164373513e-05,
+      "loss": 1.8898,
+      "step": 55
+    },
+    {
+      "epoch": 0.2828282828282828,
+      "grad_norm": 0.42600738162835966,
+      "learning_rate": 2.9831276419670593e-05,
+      "loss": 1.7645,
+      "step": 56
+    },
+    {
+      "epoch": 0.2878787878787879,
+      "grad_norm": 0.523957827857094,
+      "learning_rate": 2.9818572652403336e-05,
+      "loss": 1.8688,
+      "step": 57
+    },
+    {
+      "epoch": 0.29292929292929293,
+      "grad_norm": 0.4739264160838728,
+      "learning_rate": 2.9805410734699694e-05,
+      "loss": 1.8253,
+      "step": 58
+    },
+    {
+      "epoch": 0.29797979797979796,
+      "grad_norm": 0.4030312937735394,
+      "learning_rate": 2.9791791073490795e-05,
+      "loss": 1.8799,
+      "step": 59
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 0.4287715829219747,
+      "learning_rate": 2.9777714089859946e-05,
+      "loss": 1.8283,
+      "step": 60
+    },
+    {
+      "epoch": 0.30808080808080807,
+      "grad_norm": 0.41789218942140666,
+      "learning_rate": 2.976318021902965e-05,
+      "loss": 1.8135,
+      "step": 61
+    },
+    {
+      "epoch": 0.31313131313131315,
+      "grad_norm": 0.4587324787560187,
+      "learning_rate": 2.9748189910348122e-05,
+      "loss": 1.8821,
+      "step": 62
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 0.4460155322546901,
+      "learning_rate": 2.9732743627275428e-05,
+      "loss": 1.8889,
+      "step": 63
+    },
+    {
+      "epoch": 0.32323232323232326,
+      "grad_norm": 0.4157211867080571,
+      "learning_rate": 2.9716841847369106e-05,
+      "loss": 1.8345,
+      "step": 64
+    },
+    {
+      "epoch": 0.3282828282828283,
+      "grad_norm": 0.41844440976522496,
+      "learning_rate": 2.9700485062269465e-05,
+      "loss": 1.7642,
+      "step": 65
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.4520630042741615,
+      "learning_rate": 2.968367377768432e-05,
+      "loss": 1.8485,
+      "step": 66
+    },
+    {
+      "epoch": 0.3383838383838384,
+      "grad_norm": 0.4368200901934031,
+      "learning_rate": 2.966640851337342e-05,
+      "loss": 1.8617,
+      "step": 67
+    },
+    {
+      "epoch": 0.3434343434343434,
+      "grad_norm": 0.38212187200746967,
+      "learning_rate": 2.964868980313232e-05,
+      "loss": 1.7493,
+      "step": 68
+    },
+    {
+      "epoch": 0.3484848484848485,
+      "grad_norm": 0.4173981057054639,
+      "learning_rate": 2.963051819477592e-05,
+      "loss": 1.836,
+      "step": 69
+    },
+    {
+      "epoch": 0.35353535353535354,
+      "grad_norm": 0.4067519509663366,
+      "learning_rate": 2.9611894250121508e-05,
+      "loss": 1.81,
+      "step": 70
+    },
+    {
+      "epoch": 0.35858585858585856,
+      "grad_norm": 0.40355345502721407,
+      "learning_rate": 2.9592818544971394e-05,
+      "loss": 1.8126,
+      "step": 71
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.4205716149104244,
+      "learning_rate": 2.9573291669095112e-05,
+      "loss": 1.725,
+      "step": 72
+    },
+    {
+      "epoch": 0.3686868686868687,
+      "grad_norm": 0.42333778099845615,
+      "learning_rate": 2.955331422621117e-05,
+      "loss": 1.8448,
+      "step": 73
+    },
+    {
+      "epoch": 0.37373737373737376,
+      "grad_norm": 0.39602223830054895,
+      "learning_rate": 2.953288683396841e-05,
+      "loss": 1.8518,
+      "step": 74
+    },
+    {
+      "epoch": 0.3787878787878788,
+      "grad_norm": 0.42272756108809484,
+      "learning_rate": 2.95120101239269e-05,
+      "loss": 1.846,
+      "step": 75
+    },
+    {
+      "epoch": 0.3838383838383838,
+      "grad_norm": 0.38458304927883635,
+      "learning_rate": 2.9490684741538394e-05,
+      "loss": 1.7287,
+      "step": 76
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 0.4515217019544044,
+      "learning_rate": 2.9468911346126395e-05,
+      "loss": 1.9222,
+      "step": 77
+    },
+    {
+      "epoch": 0.3939393939393939,
+      "grad_norm": 0.3902694265684303,
+      "learning_rate": 2.9446690610865775e-05,
+      "loss": 1.8075,
+      "step": 78
+    },
+    {
+      "epoch": 0.398989898989899,
+      "grad_norm": 0.4621030399414333,
+      "learning_rate": 2.9424023222761938e-05,
+      "loss": 1.8307,
+      "step": 79
+    },
+    {
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.4164626377683268,
+      "learning_rate": 2.9400909882629595e-05,
+      "loss": 1.826,
+      "step": 80
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 0.4989467879911725,
+      "learning_rate": 2.9377351305071097e-05,
+      "loss": 1.7694,
+      "step": 81
+    },
+    {
+      "epoch": 0.41414141414141414,
+      "grad_norm": 0.38502918266185665,
+      "learning_rate": 2.935334821845434e-05,
+      "loss": 1.8558,
+      "step": 82
+    },
+    {
+      "epoch": 0.41919191919191917,
+      "grad_norm": 0.43876163502583143,
+      "learning_rate": 2.9328901364890253e-05,
+      "loss": 1.9031,
+      "step": 83
+    },
+    {
+      "epoch": 0.42424242424242425,
+      "grad_norm": 0.44274327469566893,
+      "learning_rate": 2.930401150020983e-05,
+      "loss": 1.8486,
+      "step": 84
+    },
+    {
+      "epoch": 0.4292929292929293,
+      "grad_norm": 0.42306669374100425,
+      "learning_rate": 2.9278679393940794e-05,
+      "loss": 1.8095,
+      "step": 85
+    },
+    {
+      "epoch": 0.43434343434343436,
+      "grad_norm": 0.43069541482459783,
+      "learning_rate": 2.9252905829283778e-05,
+      "loss": 1.8091,
+      "step": 86
+    },
+    {
+      "epoch": 0.4393939393939394,
+      "grad_norm": 0.524368528081938,
+      "learning_rate": 2.9226691603088124e-05,
+      "loss": 1.8407,
+      "step": 87
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.4306578292725837,
+      "learning_rate": 2.9200037525827255e-05,
+      "loss": 1.8075,
+      "step": 88
+    },
+    {
+      "epoch": 0.4494949494949495,
+      "grad_norm": 0.401793293233711,
+      "learning_rate": 2.9172944421573587e-05,
+      "loss": 1.8405,
+      "step": 89
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 0.48084582939870474,
+      "learning_rate": 2.9145413127973085e-05,
+      "loss": 1.8596,
+      "step": 90
+    },
+    {
+      "epoch": 0.4595959595959596,
+      "grad_norm": 0.4259765113516244,
+      "learning_rate": 2.911744449621935e-05,
+      "loss": 1.7931,
+      "step": 91
+    },
+    {
+      "epoch": 0.46464646464646464,
+      "grad_norm": 0.41686644545343865,
+      "learning_rate": 2.90890393910273e-05,
+      "loss": 1.8693,
+      "step": 92
+    },
+    {
+      "epoch": 0.4696969696969697,
+      "grad_norm": 0.49645836560093654,
+      "learning_rate": 2.9060198690606438e-05,
+      "loss": 1.8567,
+      "step": 93
+    },
+    {
+      "epoch": 0.47474747474747475,
+      "grad_norm": 0.4122894943723907,
+      "learning_rate": 2.9030923286633703e-05,
+      "loss": 1.7979,
+      "step": 94
+    },
+    {
+      "epoch": 0.4797979797979798,
+      "grad_norm": 0.41291814598709514,
+      "learning_rate": 2.9001214084225898e-05,
+      "loss": 1.8409,
+      "step": 95
+    },
+    {
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.4346649991553492,
+      "learning_rate": 2.89710720019117e-05,
+      "loss": 1.8904,
+      "step": 96
+    },
+    {
+      "epoch": 0.4898989898989899,
+      "grad_norm": 0.41818341954882543,
+      "learning_rate": 2.8940497971603288e-05,
+      "loss": 1.8695,
+      "step": 97
+    },
+    {
+      "epoch": 0.494949494949495,
+      "grad_norm": 0.4420144131423374,
+      "learning_rate": 2.890949293856749e-05,
+      "loss": 1.8021,
+      "step": 98
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43880248651128445,
+      "learning_rate": 2.8878057861396606e-05,
+      "loss": 1.8329,
+      "step": 99
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "grad_norm": 0.40444053098152477,
+      "learning_rate": 2.8846193711978717e-05,
+      "loss": 1.8201,
+      "step": 100
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "eval_loss": 1.4966636896133423,
+      "eval_runtime": 229.0804,
+      "eval_samples_per_second": 6.247,
+      "eval_steps_per_second": 0.781,
+      "step": 100
+    },
+    {
+      "epoch": 0.51010101010101,
+      "grad_norm": 0.45960433784465665,
+      "learning_rate": 2.881390147546768e-05,
+      "loss": 1.8183,
+      "step": 101
+    },
+    {
+      "epoch": 0.5151515151515151,
+      "grad_norm": 0.38763847079359476,
+      "learning_rate": 2.878118215025265e-05,
+      "loss": 1.8208,
+      "step": 102
+    },
+    {
+      "epoch": 0.5202020202020202,
+      "grad_norm": 0.41520378471101016,
+      "learning_rate": 2.874803674792722e-05,
+      "loss": 1.8714,
+      "step": 103
+    },
+    {
+      "epoch": 0.5252525252525253,
+      "grad_norm": 0.40206269424525176,
+      "learning_rate": 2.8714466293258142e-05,
+      "loss": 1.7878,
+      "step": 104
+    },
+    {
+      "epoch": 0.5303030303030303,
+      "grad_norm": 0.4373758839964535,
+      "learning_rate": 2.868047182415364e-05,
+      "loss": 1.8093,
+      "step": 105
+    },
+    {
+      "epoch": 0.5353535353535354,
+      "grad_norm": 0.3875079810746361,
+      "learning_rate": 2.864605439163133e-05,
+      "loss": 1.8606,
+      "step": 106
+    },
+    {
+      "epoch": 0.5404040404040404,
+      "grad_norm": 0.4303262268327845,
+      "learning_rate": 2.8611215059785706e-05,
+      "loss": 1.8233,
+      "step": 107
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.45649878897921986,
+      "learning_rate": 2.8575954905755278e-05,
+      "loss": 1.8817,
+      "step": 108
+    },
+    {
+      "epoch": 0.5505050505050505,
+      "grad_norm": 0.44045126098579024,
+      "learning_rate": 2.8540275019689237e-05,
+      "loss": 1.7945,
+      "step": 109
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.43144466038537,
+      "learning_rate": 2.8504176504713745e-05,
+      "loss": 1.7419,
+      "step": 110
+    },
+    {
+      "epoch": 0.5606060606060606,
+      "grad_norm": 0.38271501972768057,
+      "learning_rate": 2.846766047689787e-05,
+      "loss": 1.8557,
+      "step": 111
+    },
+    {
+      "epoch": 0.5656565656565656,
+      "grad_norm": 0.4193952465924727,
+      "learning_rate": 2.8430728065219035e-05,
+      "loss": 1.8245,
+      "step": 112
+    },
+    {
+      "epoch": 0.5707070707070707,
+      "grad_norm": 0.41096403730883174,
+      "learning_rate": 2.839338041152814e-05,
+      "loss": 1.8117,
+      "step": 113
+    },
+    {
+      "epoch": 0.5757575757575758,
+      "grad_norm": 0.41746315731568373,
+      "learning_rate": 2.835561867051426e-05,
+      "loss": 1.7859,
+      "step": 114
+    },
+    {
+      "epoch": 0.5808080808080808,
+      "grad_norm": 0.35796806338294784,
+      "learning_rate": 2.8317444009668916e-05,
+      "loss": 1.8125,
+      "step": 115
+    },
+    {
+      "epoch": 0.5858585858585859,
+      "grad_norm": 0.3782772183418781,
+      "learning_rate": 2.8278857609250033e-05,
+      "loss": 1.8551,
+      "step": 116
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 0.4563811308324174,
+      "learning_rate": 2.823986066224538e-05,
+      "loss": 1.8711,
+      "step": 117
+    },
+    {
+      "epoch": 0.5959595959595959,
+      "grad_norm": 0.42499726957188555,
+      "learning_rate": 2.820045437433575e-05,
+      "loss": 1.8827,
+      "step": 118
+    },
+    {
+      "epoch": 0.601010101010101,
+      "grad_norm": 0.35463970926383165,
+      "learning_rate": 2.816063996385765e-05,
+      "loss": 1.8511,
+      "step": 119
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 0.39828700116760574,
+      "learning_rate": 2.8120418661765624e-05,
+      "loss": 1.8386,
+      "step": 120
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 0.406542351713585,
+      "learning_rate": 2.8079791711594236e-05,
+      "loss": 1.8035,
+      "step": 121
+    },
+    {
+      "epoch": 0.6161616161616161,
+      "grad_norm": 0.36243471155718693,
+      "learning_rate": 2.8038760369419583e-05,
+      "loss": 1.8014,
+      "step": 122
+    },
+    {
+      "epoch": 0.6212121212121212,
+      "grad_norm": 0.3845741639511099,
+      "learning_rate": 2.7997325903820478e-05,
+      "loss": 1.9082,
+      "step": 123
+    },
+    {
+      "epoch": 0.6262626262626263,
+      "grad_norm": 0.4004906560962447,
+      "learning_rate": 2.7955489595839228e-05,
+      "loss": 1.828,
+      "step": 124
+    },
+    {
+      "epoch": 0.6313131313131313,
+      "grad_norm": 0.40745184195706274,
+      "learning_rate": 2.7913252738942027e-05,
+      "loss": 1.8285,
+      "step": 125
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 0.4701862294606233,
+      "learning_rate": 2.787061663897896e-05,
+      "loss": 1.7612,
+      "step": 126
+    },
+    {
+      "epoch": 0.6414141414141414,
+      "grad_norm": 0.3845304191870889,
+      "learning_rate": 2.782758261414365e-05,
+      "loss": 1.9005,
+      "step": 127
+    },
+    {
+      "epoch": 0.6464646464646465,
+      "grad_norm": 0.45387333682470754,
+      "learning_rate": 2.7784151994932462e-05,
+      "loss": 1.8369,
+      "step": 128
+    },
+    {
+      "epoch": 0.6515151515151515,
+      "grad_norm": 0.402744943135056,
+      "learning_rate": 2.7740326124103416e-05,
+      "loss": 1.8382,
+      "step": 129
+    },
+    {
+      "epoch": 0.6565656565656566,
+      "grad_norm": 0.39184387630637835,
+      "learning_rate": 2.7696106356634637e-05,
+      "loss": 1.8659,
+      "step": 130
+    },
+    {
+      "epoch": 0.6616161616161617,
+      "grad_norm": 0.4626435038267262,
+      "learning_rate": 2.7651494059682485e-05,
+      "loss": 1.8667,
+      "step": 131
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.40353571201115057,
+      "learning_rate": 2.7606490612539262e-05,
+      "loss": 1.8209,
+      "step": 132
+    },
+    {
+      "epoch": 0.6717171717171717,
+      "grad_norm": 0.4162695579854718,
+      "learning_rate": 2.7561097406590595e-05,
+      "loss": 1.8002,
+      "step": 133
+    },
+    {
+      "epoch": 0.6767676767676768,
+      "grad_norm": 0.3909561466126897,
+      "learning_rate": 2.751531584527241e-05,
+      "loss": 1.8076,
+      "step": 134
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 0.41533705192980835,
+      "learning_rate": 2.746914734402752e-05,
+      "loss": 1.9099,
+      "step": 135
+    },
+    {
+      "epoch": 0.6868686868686869,
+      "grad_norm": 0.38207500499538627,
+      "learning_rate": 2.7422593330261888e-05,
+      "loss": 1.7688,
+      "step": 136
+    },
+    {
+      "epoch": 0.6919191919191919,
+      "grad_norm": 0.35742242270800784,
+      "learning_rate": 2.7375655243300493e-05,
+      "loss": 1.8651,
+      "step": 137
+    },
+    {
+      "epoch": 0.696969696969697,
+      "grad_norm": 0.41421312272061606,
+      "learning_rate": 2.7328334534342827e-05,
+      "loss": 1.8785,
+      "step": 138
+    },
+    {
+      "epoch": 0.702020202020202,
+      "grad_norm": 0.4171193053836604,
+      "learning_rate": 2.7280632666418013e-05,
+      "loss": 1.8781,
+      "step": 139
+    },
+    {
+      "epoch": 0.7070707070707071,
+      "grad_norm": 0.40782157651436396,
+      "learning_rate": 2.723255111433959e-05,
+      "loss": 1.8247,
+      "step": 140
+    },
+    {
+      "epoch": 0.7121212121212122,
+      "grad_norm": 0.4281506276367188,
+      "learning_rate": 2.7184091364659923e-05,
+      "loss": 1.733,
+      "step": 141
+    },
+    {
+      "epoch": 0.7171717171717171,
+      "grad_norm": 0.39479511690560315,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 1.8401,
+      "step": 142
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 0.42100536218468343,
+      "learning_rate": 2.70860432771242e-05,
+      "loss": 1.8484,
+      "step": 143
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.3855161042933354,
+      "learning_rate": 2.703645797065147e-05,
+      "loss": 1.8197,
+      "step": 144
+    },
+    {
+      "epoch": 0.7323232323232324,
+      "grad_norm": 0.3920745691095182,
+      "learning_rate": 2.6986500529250427e-05,
+      "loss": 1.8345,
+      "step": 145
+    },
+    {
+      "epoch": 0.7373737373737373,
+      "grad_norm": 0.42204478492867153,
+      "learning_rate": 2.6936172497470874e-05,
+      "loss": 1.8368,
+      "step": 146
+    },
+    {
+      "epoch": 0.7424242424242424,
+      "grad_norm": 0.40405492226523393,
+      "learning_rate": 2.688547543132029e-05,
+      "loss": 1.8182,
+      "step": 147
+    },
+    {
+      "epoch": 0.7474747474747475,
+      "grad_norm": 0.36686029807824033,
+      "learning_rate": 2.6834410898215688e-05,
+      "loss": 1.8581,
+      "step": 148
+    },
+    {
+      "epoch": 0.7525252525252525,
+      "grad_norm": 0.36859567357522177,
+      "learning_rate": 2.678298047693518e-05,
+      "loss": 1.7276,
+      "step": 149
+    },
+    {
+      "epoch": 0.7575757575757576,
+      "grad_norm": 0.37761559772659303,
+      "learning_rate": 2.6731185757569153e-05,
+      "loss": 1.8321,
+      "step": 150
+    },
+    {
+      "epoch": 0.7575757575757576,
+      "eval_loss": 1.484434962272644,
+      "eval_runtime": 229.048,
+      "eval_samples_per_second": 6.248,
+      "eval_steps_per_second": 0.781,
+      "step": 150
+    },
+    {
+      "epoch": 0.7626262626262627,
+      "grad_norm": 0.40657370322541503,
+      "learning_rate": 2.6679028341471114e-05,
+      "loss": 1.8364,
+      "step": 151
+    },
+    {
+      "epoch": 0.7676767676767676,
+      "grad_norm": 0.3852020973392934,
+      "learning_rate": 2.6626509841208177e-05,
+      "loss": 1.8723,
+      "step": 152
+    },
+    {
+      "epoch": 0.7727272727272727,
+      "grad_norm": 0.4471085311517623,
+      "learning_rate": 2.6573631880511214e-05,
+      "loss": 1.9017,
+      "step": 153
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.3858532089668969,
+      "learning_rate": 2.652039609422463e-05,
+      "loss": 1.7849,
+      "step": 154
+    },
+    {
+      "epoch": 0.7828282828282829,
+      "grad_norm": 0.427216562758674,
+      "learning_rate": 2.6466804128255865e-05,
+      "loss": 1.7743,
+      "step": 155
+    },
+    {
+      "epoch": 0.7878787878787878,
+      "grad_norm": 0.38828532121415243,
+      "learning_rate": 2.6412857639524442e-05,
+      "loss": 1.8294,
+      "step": 156
+    },
+    {
+      "epoch": 0.7929292929292929,
+      "grad_norm": 0.40962967785771864,
+      "learning_rate": 2.6358558295910805e-05,
+      "loss": 1.8314,
+      "step": 157
+    },
+    {
+      "epoch": 0.797979797979798,
+      "grad_norm": 0.7209413951413858,
+      "learning_rate": 2.6303907776204706e-05,
+      "loss": 1.8388,
+      "step": 158
+    },
+    {
+      "epoch": 0.803030303030303,
+      "grad_norm": 0.3753158677183495,
+      "learning_rate": 2.624890777005332e-05,
+      "loss": 1.9189,
+      "step": 159
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.3683405444328328,
+      "learning_rate": 2.6193559977909008e-05,
+      "loss": 1.7511,
+      "step": 160
+    },
+    {
+      "epoch": 0.8131313131313131,
+      "grad_norm": 0.420103866946048,
+      "learning_rate": 2.6137866110976742e-05,
+      "loss": 1.8447,
+      "step": 161
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 0.4025395036734605,
+      "learning_rate": 2.608182789116118e-05,
+      "loss": 1.7567,
+      "step": 162
+    },
+    {
+      "epoch": 0.8232323232323232,
+      "grad_norm": 0.41152519566999235,
+      "learning_rate": 2.6025447051013466e-05,
+      "loss": 1.8844,
+      "step": 163
+    },
+    {
+      "epoch": 0.8282828282828283,
+      "grad_norm": 0.37456667248254777,
+      "learning_rate": 2.5968725333677628e-05,
+      "loss": 1.8316,
+      "step": 164
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.3989107911221168,
+      "learning_rate": 2.5911664492836714e-05,
+      "loss": 1.8336,
+      "step": 165
+    },
+    {
+      "epoch": 0.8383838383838383,
+      "grad_norm": 0.4192570186399055,
+      "learning_rate": 2.585426629265854e-05,
+      "loss": 1.8368,
+      "step": 166
+    },
+    {
+      "epoch": 0.8434343434343434,
+      "grad_norm": 0.4039992281383353,
+      "learning_rate": 2.579653250774119e-05,
+      "loss": 1.8474,
+      "step": 167
+    },
+    {
+      "epoch": 0.8484848484848485,
+      "grad_norm": 0.3860934737365767,
+      "learning_rate": 2.5738464923058118e-05,
+      "loss": 1.9136,
+      "step": 168
+    },
+    {
+      "epoch": 0.8535353535353535,
+      "grad_norm": 0.3766141715707985,
+      "learning_rate": 2.568006533390295e-05,
+      "loss": 1.7905,
+      "step": 169
+    },
+    {
+      "epoch": 0.8585858585858586,
+      "grad_norm": 0.43692452251495456,
+      "learning_rate": 2.562133554583402e-05,
+      "loss": 1.8692,
+      "step": 170
+    },
+    {
+      "epoch": 0.8636363636363636,
+      "grad_norm": 0.38506171479168916,
+      "learning_rate": 2.556227737461852e-05,
+      "loss": 1.8802,
+      "step": 171
+    },
+    {
+      "epoch": 0.8686868686868687,
+      "grad_norm": 0.4139103090309799,
+      "learning_rate": 2.5502892646176364e-05,
+      "loss": 1.9145,
+      "step": 172
+    },
+    {
+      "epoch": 0.8737373737373737,
+      "grad_norm": 0.41202711286858845,
+      "learning_rate": 2.5443183196523744e-05,
+      "loss": 1.8799,
+      "step": 173
+    },
+    {
+      "epoch": 0.8787878787878788,
+      "grad_norm": 0.441012649010044,
+      "learning_rate": 2.5383150871716342e-05,
+      "loss": 1.7964,
+      "step": 174
+    },
+    {
+      "epoch": 0.8838383838383839,
+      "grad_norm": 0.3924859380226668,
+      "learning_rate": 2.5322797527792297e-05,
+      "loss": 1.8196,
+      "step": 175
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.484116113199706,
+      "learning_rate": 2.526212503071477e-05,
+      "loss": 1.8211,
+      "step": 176
+    },
+    {
+      "epoch": 0.8939393939393939,
+      "grad_norm": 0.4214268259949748,
+      "learning_rate": 2.52011352563143e-05,
+      "loss": 1.8756,
+      "step": 177
+    },
+    {
+      "epoch": 0.898989898989899,
+      "grad_norm": 0.3761454305197214,
+      "learning_rate": 2.5139830090230776e-05,
+      "loss": 1.7887,
+      "step": 178
+    },
+    {
+      "epoch": 0.9040404040404041,
+      "grad_norm": 0.38472045617953676,
+      "learning_rate": 2.507821142785516e-05,
+      "loss": 1.9109,
+      "step": 179
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.4437224782187781,
+      "learning_rate": 2.501628117427087e-05,
+      "loss": 1.8024,
+      "step": 180
+    },
+    {
+      "epoch": 0.9141414141414141,
+      "grad_norm": 0.3928636586609788,
+      "learning_rate": 2.4954041244194883e-05,
+      "loss": 1.8177,
+      "step": 181
+    },
+    {
+      "epoch": 0.9191919191919192,
+      "grad_norm": 0.40978302308222725,
+      "learning_rate": 2.4891493561918545e-05,
+      "loss": 1.8238,
+      "step": 182
+    },
+    {
+      "epoch": 0.9242424242424242,
+      "grad_norm": 0.41578094468387805,
+      "learning_rate": 2.482864006124808e-05,
+      "loss": 1.7558,
+      "step": 183
+    },
+    {
+      "epoch": 0.9292929292929293,
+      "grad_norm": 0.47459329104171105,
+      "learning_rate": 2.4765482685444786e-05,
+      "loss": 1.8982,
+      "step": 184
+    },
+    {
+      "epoch": 0.9343434343434344,
+      "grad_norm": 0.4584785873115325,
+      "learning_rate": 2.470202338716497e-05,
+      "loss": 1.8925,
+      "step": 185
+    },
+    {
+      "epoch": 0.9393939393939394,
+      "grad_norm": 0.39624647908908,
+      "learning_rate": 2.4638264128399555e-05,
+      "loss": 1.8032,
+      "step": 186
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 0.38536269925326294,
+      "learning_rate": 2.457420688041345e-05,
+      "loss": 1.7927,
+      "step": 187
+    },
+    {
+      "epoch": 0.9494949494949495,
+      "grad_norm": 0.44911468803951493,
+      "learning_rate": 2.4509853623684598e-05,
+      "loss": 1.8635,
+      "step": 188
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 0.38483613716276227,
+      "learning_rate": 2.4445206347842714e-05,
+      "loss": 1.828,
+      "step": 189
+    },
+    {
+      "epoch": 0.9595959595959596,
+      "grad_norm": 0.4134516508007333,
+      "learning_rate": 2.43802670516078e-05,
+      "loss": 1.7766,
+      "step": 190
+    },
+    {
+      "epoch": 0.9646464646464646,
+      "grad_norm": 0.4205305313105495,
+      "learning_rate": 2.4315037742728366e-05,
+      "loss": 1.8162,
+      "step": 191
+    },
+    {
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.4019754805551098,
+      "learning_rate": 2.4249520437919307e-05,
+      "loss": 1.7429,
+      "step": 192
+    },
+    {
+      "epoch": 0.9747474747474747,
+      "grad_norm": 0.38403144394408584,
+      "learning_rate": 2.4183717162799587e-05,
+      "loss": 1.731,
+      "step": 193
+    },
+    {
+      "epoch": 0.9797979797979798,
+      "grad_norm": 0.3807338127364377,
+      "learning_rate": 2.4117629951829602e-05,
+      "loss": 1.8942,
+      "step": 194
+    },
+    {
+      "epoch": 0.9848484848484849,
+      "grad_norm": 0.44761141331014614,
+      "learning_rate": 2.4051260848248286e-05,
+      "loss": 1.8313,
+      "step": 195
+    },
+    {
+      "epoch": 0.98989898989899,
+      "grad_norm": 0.38554593665783987,
+      "learning_rate": 2.398461190400993e-05,
+      "loss": 1.8394,
+      "step": 196
+    },
+    {
+      "epoch": 0.9949494949494949,
+      "grad_norm": 0.403968521425789,
+      "learning_rate": 2.3917685179720752e-05,
+      "loss": 1.8287,
+      "step": 197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36274397859542107,
+      "learning_rate": 2.3850482744575177e-05,
+      "loss": 1.7637,
+      "step": 198
+    },
+    {
+      "epoch": 1.005050505050505,
+      "grad_norm": 0.5408632160019302,
+      "learning_rate": 2.3783006676291866e-05,
+      "loss": 1.6493,
+      "step": 199
+    },
+    {
+      "epoch": 1.0101010101010102,
+      "grad_norm": 0.4273958454403075,
+      "learning_rate": 2.3715259061049487e-05,
+      "loss": 1.5513,
+      "step": 200
+    },
+    {
+      "epoch": 1.0101010101010102,
+      "eval_loss": 1.4980653524398804,
+      "eval_runtime": 229.1221,
+      "eval_samples_per_second": 6.246,
+      "eval_steps_per_second": 0.781,
+      "step": 200
+    },
+    {
+      "epoch": 1.0151515151515151,
+      "grad_norm": 0.9048049507315692,
+      "learning_rate": 2.3647241993422208e-05,
+      "loss": 1.6547,
+      "step": 201
+    },
+    {
+      "epoch": 1.02020202020202,
+      "grad_norm": 0.3848882038249155,
+      "learning_rate": 2.3578957576314944e-05,
+      "loss": 1.5279,
+      "step": 202
+    },
+    {
+      "epoch": 1.0252525252525253,
+      "grad_norm": 0.40317657177333205,
+      "learning_rate": 2.3510407920898327e-05,
+      "loss": 1.6036,
+      "step": 203
+    },
+    {
+      "epoch": 1.0303030303030303,
+      "grad_norm": 0.43727100501923616,
+      "learning_rate": 2.3441595146543458e-05,
+      "loss": 1.6581,
+      "step": 204
+    },
+    {
+      "epoch": 1.0353535353535352,
+      "grad_norm": 0.3912845389083466,
+      "learning_rate": 2.337252138075636e-05,
+      "loss": 1.5399,
+      "step": 205
+    },
+    {
+      "epoch": 1.0404040404040404,
+      "grad_norm": 0.3961929212943775,
+      "learning_rate": 2.3303188759112213e-05,
+      "loss": 1.6788,
+      "step": 206
+    },
+    {
+      "epoch": 1.0454545454545454,
+      "grad_norm": 0.4307375964156506,
+      "learning_rate": 2.3233599425189317e-05,
+      "loss": 1.6081,
+      "step": 207
+    },
+    {
+      "epoch": 1.0505050505050506,
+      "grad_norm": 0.3912142497081324,
+      "learning_rate": 2.316375553050284e-05,
+      "loss": 1.5678,
+      "step": 208
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 0.3763414411950308,
+      "learning_rate": 2.3093659234438266e-05,
+      "loss": 1.5547,
+      "step": 209
+    },
+    {
+      "epoch": 1.0606060606060606,
+      "grad_norm": 0.36800317307295144,
+      "learning_rate": 2.3023312704184676e-05,
+      "loss": 1.4771,
+      "step": 210
+    },
+    {
+      "epoch": 1.0656565656565657,
+      "grad_norm": 0.3910877207089002,
+      "learning_rate": 2.295271811466769e-05,
+      "loss": 1.6495,
+      "step": 211
+    },
+    {
+      "epoch": 1.0707070707070707,
+      "grad_norm": 0.39199841864196705,
+      "learning_rate": 2.2881877648482274e-05,
+      "loss": 1.6182,
+      "step": 212
+    },
+    {
+      "epoch": 1.0757575757575757,
+      "grad_norm": 0.4188630635601797,
+      "learning_rate": 2.281079349582524e-05,
+      "loss": 1.5539,
+      "step": 213
+    },
+    {
+      "epoch": 1.0808080808080809,
+      "grad_norm": 0.4321942331098238,
+      "learning_rate": 2.2739467854427512e-05,
+      "loss": 1.7017,
+      "step": 214
+    },
+    {
+      "epoch": 1.0858585858585859,
+      "grad_norm": 0.34896226538754127,
+      "learning_rate": 2.266790292948622e-05,
+      "loss": 1.6232,
+      "step": 215
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 0.39588469087020545,
+      "learning_rate": 2.2596100933596498e-05,
+      "loss": 1.633,
+      "step": 216
+    },
+    {
+      "epoch": 1.095959595959596,
+      "grad_norm": 0.3680904345345009,
+      "learning_rate": 2.252406408668304e-05,
+      "loss": 1.5771,
+      "step": 217
+    },
+    {
+      "epoch": 1.101010101010101,
+      "grad_norm": 0.3635291050534616,
+      "learning_rate": 2.2451794615931542e-05,
+      "loss": 1.5251,
+      "step": 218
+    },
+    {
+      "epoch": 1.106060606060606,
+      "grad_norm": 0.39495821190617536,
+      "learning_rate": 2.237929475571979e-05,
+      "loss": 1.6461,
+      "step": 219
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.3741651687847982,
+      "learning_rate": 2.2306566747548604e-05,
+      "loss": 1.5491,
+      "step": 220
+    },
+    {
+      "epoch": 1.1161616161616161,
+      "grad_norm": 0.3482093972061835,
+      "learning_rate": 2.2233612839972497e-05,
+      "loss": 1.4896,
+      "step": 221
+    },
+    {
+      "epoch": 1.121212121212121,
+      "grad_norm": 0.39618420354024975,
+      "learning_rate": 2.2160435288530208e-05,
+      "loss": 1.6567,
+      "step": 222
+    },
+    {
+      "epoch": 1.1262626262626263,
+      "grad_norm": 0.3660259252156906,
+      "learning_rate": 2.2087036355674947e-05,
+      "loss": 1.5476,
+      "step": 223
+    },
+    {
+      "epoch": 1.1313131313131313,
+      "grad_norm": 0.39507488097490023,
+      "learning_rate": 2.2013418310704422e-05,
+      "loss": 1.5749,
+      "step": 224
+    },
+    {
+      "epoch": 1.1363636363636362,
+      "grad_norm": 0.4312941458010281,
+      "learning_rate": 2.1939583429690716e-05,
+      "loss": 1.5716,
+      "step": 225
+    },
+    {
+      "epoch": 1.1414141414141414,
+      "grad_norm": 0.35158839668333186,
+      "learning_rate": 2.1865533995409887e-05,
+      "loss": 1.5532,
+      "step": 226
+    },
+    {
+      "epoch": 1.1464646464646464,
+      "grad_norm": 0.3641625201369256,
+      "learning_rate": 2.1791272297271416e-05,
+      "loss": 1.5972,
+      "step": 227
+    },
+    {
+      "epoch": 1.1515151515151516,
+      "grad_norm": 0.4250517041163231,
+      "learning_rate": 2.1716800631247403e-05,
+      "loss": 1.6134,
+      "step": 228
+    },
+    {
+      "epoch": 1.1565656565656566,
+      "grad_norm": 0.4357327884434021,
+      "learning_rate": 2.1642121299801594e-05,
+      "loss": 1.6299,
+      "step": 229
+    },
+    {
+      "epoch": 1.1616161616161615,
+      "grad_norm": 0.36361246043809753,
+      "learning_rate": 2.1567236611818187e-05,
+      "loss": 1.6186,
+      "step": 230
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.38940692200146193,
+      "learning_rate": 2.149214888253046e-05,
+      "loss": 1.5636,
+      "step": 231
+    },
+    {
+      "epoch": 1.1717171717171717,
+      "grad_norm": 0.4479735326446513,
+      "learning_rate": 2.1416860433449177e-05,
+      "loss": 1.6342,
+      "step": 232
+    },
+    {
+      "epoch": 1.1767676767676767,
+      "grad_norm": 0.36500909652831554,
+      "learning_rate": 2.1341373592290822e-05,
+      "loss": 1.545,
+      "step": 233
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 0.41140999886470425,
+      "learning_rate": 2.126569069290562e-05,
+      "loss": 1.6267,
+      "step": 234
+    },
+    {
+      "epoch": 1.1868686868686869,
+      "grad_norm": 0.37095591557372914,
+      "learning_rate": 2.1189814075205406e-05,
+      "loss": 1.5176,
+      "step": 235
+    },
+    {
+      "epoch": 1.1919191919191918,
+      "grad_norm": 0.3754937123457338,
+      "learning_rate": 2.1113746085091246e-05,
+      "loss": 1.6441,
+      "step": 236
+    },
+    {
+      "epoch": 1.196969696969697,
+      "grad_norm": 0.42464228140478444,
+      "learning_rate": 2.1037489074380934e-05,
+      "loss": 1.574,
+      "step": 237
+    },
+    {
+      "epoch": 1.202020202020202,
+      "grad_norm": 0.3732085771180313,
+      "learning_rate": 2.0961045400736286e-05,
+      "loss": 1.5687,
+      "step": 238
+    },
+    {
+      "epoch": 1.2070707070707072,
+      "grad_norm": 0.4174666906233983,
+      "learning_rate": 2.0884417427590217e-05,
+      "loss": 1.5579,
+      "step": 239
+    },
+    {
+      "epoch": 1.2121212121212122,
+      "grad_norm": 0.414246878313963,
+      "learning_rate": 2.0807607524073684e-05,
+      "loss": 1.6069,
+      "step": 240
+    },
+    {
+      "epoch": 1.2171717171717171,
+      "grad_norm": 0.35489448104499416,
+      "learning_rate": 2.073061806494246e-05,
+      "loss": 1.5943,
+      "step": 241
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.3961867107004295,
+      "learning_rate": 2.0653451430503686e-05,
+      "loss": 1.5851,
+      "step": 242
+    },
+    {
+      "epoch": 1.2272727272727273,
+      "grad_norm": 0.4264411067539511,
+      "learning_rate": 2.0576110006542278e-05,
+      "loss": 1.5436,
+      "step": 243
+    },
+    {
+      "epoch": 1.2323232323232323,
+      "grad_norm": 0.43088202335465686,
+      "learning_rate": 2.0498596184247196e-05,
+      "loss": 1.642,
+      "step": 244
+    },
+    {
+      "epoch": 1.2373737373737375,
+      "grad_norm": 0.36469200409321884,
+      "learning_rate": 2.0420912360137466e-05,
+      "loss": 1.6031,
+      "step": 245
+    },
+    {
+      "epoch": 1.2424242424242424,
+      "grad_norm": 0.4153699560933353,
+      "learning_rate": 2.0343060935988136e-05,
+      "loss": 1.6385,
+      "step": 246
+    },
+    {
+      "epoch": 1.2474747474747474,
+      "grad_norm": 0.39846520447706185,
+      "learning_rate": 2.0265044318755988e-05,
+      "loss": 1.5288,
+      "step": 247
+    },
+    {
+      "epoch": 1.2525252525252526,
+      "grad_norm": 0.35709349697159576,
+      "learning_rate": 2.018686492050513e-05,
+      "loss": 1.6124,
+      "step": 248
+    },
+    {
+      "epoch": 1.2575757575757576,
+      "grad_norm": 0.38319549330931446,
+      "learning_rate": 2.010852515833242e-05,
+      "loss": 1.535,
+      "step": 249
+    },
+    {
+      "epoch": 1.2626262626262625,
+      "grad_norm": 0.46317037418448487,
+      "learning_rate": 2.003002745429274e-05,
+      "loss": 1.6523,
+      "step": 250
+    },
+    {
+      "epoch": 1.2626262626262625,
+      "eval_loss": 1.498921275138855,
+      "eval_runtime": 229.2608,
+      "eval_samples_per_second": 6.242,
+      "eval_steps_per_second": 0.781,
+      "step": 250
+    },
+    {
+      "epoch": 1.2676767676767677,
+      "grad_norm": 0.386123708626963,
+      "learning_rate": 1.9951374235324105e-05,
+      "loss": 1.6125,
+      "step": 251
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 0.3970356471812719,
+      "learning_rate": 1.9872567933172647e-05,
+      "loss": 1.586,
+      "step": 252
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 0.4033780506493091,
+      "learning_rate": 1.97936109843174e-05,
+      "loss": 1.5574,
+      "step": 253
+    },
+    {
+      "epoch": 1.2828282828282829,
+      "grad_norm": 0.39482348113655924,
+      "learning_rate": 1.9714505829895004e-05,
+      "loss": 1.6136,
+      "step": 254
+    },
+    {
+      "epoch": 1.2878787878787878,
+      "grad_norm": 0.45725538920837333,
+      "learning_rate": 1.963525491562421e-05,
+      "loss": 1.6221,
+      "step": 255
+    },
+    {
+      "epoch": 1.2929292929292928,
+      "grad_norm": 0.4259870746825829,
+      "learning_rate": 1.9555860691730277e-05,
+      "loss": 1.5543,
+      "step": 256
+    },
+    {
+      "epoch": 1.297979797979798,
+      "grad_norm": 0.38172206665482866,
+      "learning_rate": 1.9476325612869202e-05,
+      "loss": 1.4709,
+      "step": 257
+    },
+    {
+      "epoch": 1.303030303030303,
+      "grad_norm": 0.3996792794666037,
+      "learning_rate": 1.9396652138051844e-05,
+      "loss": 1.6108,
+      "step": 258
+    },
+    {
+      "epoch": 1.308080808080808,
+      "grad_norm": 0.4433312059533111,
+      "learning_rate": 1.9316842730567902e-05,
+      "loss": 1.627,
+      "step": 259
+    },
+    {
+      "epoch": 1.3131313131313131,
+      "grad_norm": 0.36892676364281046,
+      "learning_rate": 1.923689985790974e-05,
+      "loss": 1.4977,
+      "step": 260
+    },
+    {
+      "epoch": 1.3181818181818181,
+      "grad_norm": 0.3785785048742356,
+      "learning_rate": 1.9156825991696096e-05,
+      "loss": 1.5503,
+      "step": 261
+    },
+    {
+      "epoch": 1.3232323232323233,
+      "grad_norm": 0.4339005608639186,
+      "learning_rate": 1.9076623607595696e-05,
+      "loss": 1.5747,
+      "step": 262
+    },
+    {
+      "epoch": 1.3282828282828283,
+      "grad_norm": 0.37171089149936926,
+      "learning_rate": 1.8996295185250682e-05,
+      "loss": 1.4574,
+      "step": 263
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.383278000804476,
+      "learning_rate": 1.8915843208199967e-05,
+      "loss": 1.6036,
+      "step": 264
+    },
+    {
+      "epoch": 1.3383838383838385,
+      "grad_norm": 0.41536599243507505,
+      "learning_rate": 1.8835270163802433e-05,
+      "loss": 1.6381,
+      "step": 265
+    },
+    {
+      "epoch": 1.3434343434343434,
+      "grad_norm": 0.38161228496015287,
+      "learning_rate": 1.8754578543160045e-05,
+      "loss": 1.5011,
+      "step": 266
+    },
+    {
+      "epoch": 1.3484848484848486,
+      "grad_norm": 0.3744640963542385,
+      "learning_rate": 1.867377084104083e-05,
+      "loss": 1.6644,
+      "step": 267
+    },
+    {
+      "epoch": 1.3535353535353536,
+      "grad_norm": 0.3733108852557401,
+      "learning_rate": 1.8592849555801746e-05,
+      "loss": 1.6011,
+      "step": 268
+    },
+    {
+      "epoch": 1.3585858585858586,
+      "grad_norm": 0.4411122207695828,
+      "learning_rate": 1.851181718931141e-05,
+      "loss": 1.5586,
+      "step": 269
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.38080346450070857,
+      "learning_rate": 1.84306762468728e-05,
+      "loss": 1.5985,
+      "step": 270
+    },
+    {
+      "epoch": 1.3686868686868687,
+      "grad_norm": 0.3491871213017396,
+      "learning_rate": 1.8349429237145776e-05,
+      "loss": 1.5688,
+      "step": 271
+    },
+    {
+      "epoch": 1.3737373737373737,
+      "grad_norm": 0.37833218710396194,
+      "learning_rate": 1.8268078672069478e-05,
+      "loss": 1.5583,
+      "step": 272
+    },
+    {
+      "epoch": 1.378787878787879,
+      "grad_norm": 0.5250217329172516,
+      "learning_rate": 1.818662706678473e-05,
+      "loss": 1.5692,
+      "step": 273
+    },
+    {
+      "epoch": 1.3838383838383839,
+      "grad_norm": 0.37119240386882224,
+      "learning_rate": 1.8105076939556238e-05,
+      "loss": 1.5941,
+      "step": 274
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 0.3439611330263627,
+      "learning_rate": 1.8023430811694746e-05,
+      "loss": 1.6779,
+      "step": 275
+    },
+    {
+      "epoch": 1.393939393939394,
+      "grad_norm": 0.3327777793630774,
+      "learning_rate": 1.7941691207479067e-05,
+      "loss": 1.6037,
+      "step": 276
+    },
+    {
+      "epoch": 1.398989898989899,
+      "grad_norm": 0.34323800459831283,
+      "learning_rate": 1.7859860654078065e-05,
+      "loss": 1.6266,
+      "step": 277
+    },
+    {
+      "epoch": 1.404040404040404,
+      "grad_norm": 0.3647890764031767,
+      "learning_rate": 1.77779416814725e-05,
+      "loss": 1.5636,
+      "step": 278
+    },
+    {
+      "epoch": 1.4090909090909092,
+      "grad_norm": 0.3686098374668375,
+      "learning_rate": 1.769593682237682e-05,
+      "loss": 1.6069,
+      "step": 279
+    },
+    {
+      "epoch": 1.4141414141414141,
+      "grad_norm": 0.33500372009414925,
+      "learning_rate": 1.7613848612160857e-05,
+      "loss": 1.5403,
+      "step": 280
+    },
+    {
+      "epoch": 1.4191919191919191,
+      "grad_norm": 0.35011148506577505,
+      "learning_rate": 1.753167958877143e-05,
+      "loss": 1.6803,
+      "step": 281
+    },
+    {
+      "epoch": 1.4242424242424243,
+      "grad_norm": 0.3903537969599135,
+      "learning_rate": 1.7449432292653875e-05,
+      "loss": 1.5327,
+      "step": 282
+    },
+    {
+      "epoch": 1.4292929292929293,
+      "grad_norm": 0.3511906044729009,
+      "learning_rate": 1.736710926667352e-05,
+      "loss": 1.5245,
+      "step": 283
+    },
+    {
+      "epoch": 1.4343434343434343,
+      "grad_norm": 0.36248283366049505,
+      "learning_rate": 1.7284713056037074e-05,
+      "loss": 1.6522,
+      "step": 284
+    },
+    {
+      "epoch": 1.4393939393939394,
+      "grad_norm": 0.3698906007038963,
+      "learning_rate": 1.720224620821389e-05,
+      "loss": 1.612,
+      "step": 285
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.36917069694620513,
+      "learning_rate": 1.7119711272857242e-05,
+      "loss": 1.6716,
+      "step": 286
+    },
+    {
+      "epoch": 1.4494949494949494,
+      "grad_norm": 0.3504632139356775,
+      "learning_rate": 1.7037110801725498e-05,
+      "loss": 1.567,
+      "step": 287
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.38102242984213275,
+      "learning_rate": 1.69544473486032e-05,
+      "loss": 1.5831,
+      "step": 288
+    },
+    {
+      "epoch": 1.4595959595959596,
+      "grad_norm": 0.3490442364729326,
+      "learning_rate": 1.687172346922213e-05,
+      "loss": 1.5999,
+      "step": 289
+    },
+    {
+      "epoch": 1.4646464646464645,
+      "grad_norm": 0.36706394830835837,
+      "learning_rate": 1.67889417211823e-05,
+      "loss": 1.6264,
+      "step": 290
+    },
+    {
+      "epoch": 1.4696969696969697,
+      "grad_norm": 0.330869571421232,
+      "learning_rate": 1.670610466387285e-05,
+      "loss": 1.5905,
+      "step": 291
+    },
+    {
+      "epoch": 1.4747474747474747,
+      "grad_norm": 0.35481611568449467,
+      "learning_rate": 1.662321485839294e-05,
+      "loss": 1.5783,
+      "step": 292
+    },
+    {
+      "epoch": 1.4797979797979797,
+      "grad_norm": 0.420284177177337,
+      "learning_rate": 1.6540274867472554e-05,
+      "loss": 1.5323,
+      "step": 293
+    },
+    {
+      "epoch": 1.4848484848484849,
+      "grad_norm": 0.3733856344040281,
+      "learning_rate": 1.645728725539329e-05,
+      "loss": 1.5342,
+      "step": 294
+    },
+    {
+      "epoch": 1.4898989898989898,
+      "grad_norm": 0.35109601539946006,
+      "learning_rate": 1.637425458790905e-05,
+      "loss": 1.5483,
+      "step": 295
+    },
+    {
+      "epoch": 1.494949494949495,
+      "grad_norm": 0.341135998502388,
+      "learning_rate": 1.6291179432166737e-05,
+      "loss": 1.6149,
+      "step": 296
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.35535095307312015,
+      "learning_rate": 1.620806435662687e-05,
+      "loss": 1.5373,
+      "step": 297
+    },
+    {
+      "epoch": 1.5050505050505052,
+      "grad_norm": 0.36290854386190635,
+      "learning_rate": 1.612491193098419e-05,
+      "loss": 1.6085,
+      "step": 298
+    },
+    {
+      "epoch": 1.51010101010101,
+      "grad_norm": 0.3571923405167369,
+      "learning_rate": 1.6041724726088187e-05,
+      "loss": 1.5928,
+      "step": 299
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "grad_norm": 0.36396420706496496,
+      "learning_rate": 1.5958505313863654e-05,
+      "loss": 1.5806,
+      "step": 300
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "eval_loss": 1.4940327405929565,
+      "eval_runtime": 229.2657,
+      "eval_samples_per_second": 6.242,
+      "eval_steps_per_second": 0.781,
+      "step": 300
+    },
+    {
+      "epoch": 1.5202020202020203,
+      "grad_norm": 0.3431930051075577,
+      "learning_rate": 1.587525626723113e-05,
+      "loss": 1.5698,
+      "step": 301
+    },
+    {
+      "epoch": 1.5252525252525253,
+      "grad_norm": 0.3620485249342895,
+      "learning_rate": 1.5791980160027376e-05,
+      "loss": 1.5985,
+      "step": 302
+    },
+    {
+      "epoch": 1.5303030303030303,
+      "grad_norm": 0.34146120936201213,
+      "learning_rate": 1.570867956692579e-05,
+      "loss": 1.5825,
+      "step": 303
+    },
+    {
+      "epoch": 1.5353535353535355,
+      "grad_norm": 0.3545235530221104,
+      "learning_rate": 1.5625357063356825e-05,
+      "loss": 1.5798,
+      "step": 304
+    },
+    {
+      "epoch": 1.5404040404040404,
+      "grad_norm": 0.34094408848239416,
+      "learning_rate": 1.5542015225428314e-05,
+      "loss": 1.6249,
+      "step": 305
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 0.3574679152501359,
+      "learning_rate": 1.545865662984589e-05,
+      "loss": 1.6048,
+      "step": 306
+    },
+    {
+      "epoch": 1.5505050505050506,
+      "grad_norm": 0.3581945699930548,
+      "learning_rate": 1.5375283853833272e-05,
+      "loss": 1.7071,
+      "step": 307
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 0.34332771314631755,
+      "learning_rate": 1.5291899475052596e-05,
+      "loss": 1.563,
+      "step": 308
+    },
+    {
+      "epoch": 1.5606060606060606,
+      "grad_norm": 0.3585885575803054,
+      "learning_rate": 1.5208506071524727e-05,
+      "loss": 1.5551,
+      "step": 309
+    },
+    {
+      "epoch": 1.5656565656565657,
+      "grad_norm": 0.3478032073365952,
+      "learning_rate": 1.5125106221549567e-05,
+      "loss": 1.6227,
+      "step": 310
+    },
+    {
+      "epoch": 1.5707070707070707,
+      "grad_norm": 0.34451168320759784,
+      "learning_rate": 1.5041702503626296e-05,
+      "loss": 1.5625,
+      "step": 311
+    },
+    {
+      "epoch": 1.5757575757575757,
+      "grad_norm": 0.3625884247911658,
+      "learning_rate": 1.4958297496373708e-05,
+      "loss": 1.622,
+      "step": 312
+    },
+    {
+      "epoch": 1.5808080808080809,
+      "grad_norm": 0.3390738782534803,
+      "learning_rate": 1.4874893778450436e-05,
+      "loss": 1.5539,
+      "step": 313
+    },
+    {
+      "epoch": 1.5858585858585859,
+      "grad_norm": 0.36415276028038074,
+      "learning_rate": 1.4791493928475275e-05,
+      "loss": 1.5946,
+      "step": 314
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 0.3447466909227647,
+      "learning_rate": 1.4708100524947413e-05,
+      "loss": 1.6482,
+      "step": 315
+    },
+    {
+      "epoch": 1.595959595959596,
+      "grad_norm": 0.3573366580845515,
+      "learning_rate": 1.4624716146166734e-05,
+      "loss": 1.6493,
+      "step": 316
+    },
+    {
+      "epoch": 1.601010101010101,
+      "grad_norm": 0.35959249143609046,
+      "learning_rate": 1.4541343370154115e-05,
+      "loss": 1.6135,
+      "step": 317
+    },
+    {
+      "epoch": 1.606060606060606,
+      "grad_norm": 0.3323927118670595,
+      "learning_rate": 1.4457984774571692e-05,
+      "loss": 1.6435,
+      "step": 318
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 0.34075152367107814,
+      "learning_rate": 1.437464293664318e-05,
+      "loss": 1.6634,
+      "step": 319
+    },
+    {
+      "epoch": 1.6161616161616161,
+      "grad_norm": 0.3399955824391449,
+      "learning_rate": 1.4291320433074213e-05,
+      "loss": 1.6254,
+      "step": 320
+    },
+    {
+      "epoch": 1.621212121212121,
+      "grad_norm": 0.3427914203743951,
+      "learning_rate": 1.4208019839972627e-05,
+      "loss": 1.5906,
+      "step": 321
+    },
+    {
+      "epoch": 1.6262626262626263,
+      "grad_norm": 0.3394828693597387,
+      "learning_rate": 1.4124743732768873e-05,
+      "loss": 1.6156,
+      "step": 322
+    },
+    {
+      "epoch": 1.6313131313131313,
+      "grad_norm": 0.3854380971355761,
+      "learning_rate": 1.4041494686136348e-05,
+      "loss": 1.6758,
+      "step": 323
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 0.3391843715071689,
+      "learning_rate": 1.3958275273911812e-05,
+      "loss": 1.6073,
+      "step": 324
+    },
+    {
+      "epoch": 1.6414141414141414,
+      "grad_norm": 0.34941530909401375,
+      "learning_rate": 1.3875088069015815e-05,
+      "loss": 1.7296,
+      "step": 325
+    },
+    {
+      "epoch": 1.6464646464646466,
+      "grad_norm": 0.3645855591600974,
+      "learning_rate": 1.3791935643373133e-05,
+      "loss": 1.5878,
+      "step": 326
+    },
+    {
+      "epoch": 1.6515151515151514,
+      "grad_norm": 0.3818786490819524,
+      "learning_rate": 1.3708820567833266e-05,
+      "loss": 1.6895,
+      "step": 327
+    },
+    {
+      "epoch": 1.6565656565656566,
+      "grad_norm": 0.3918329456403931,
+      "learning_rate": 1.3625745412090953e-05,
+      "loss": 1.5491,
+      "step": 328
+    },
+    {
+      "epoch": 1.6616161616161618,
+      "grad_norm": 0.3685269378673387,
+      "learning_rate": 1.3542712744606712e-05,
+      "loss": 1.5422,
+      "step": 329
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.3784489076280344,
+      "learning_rate": 1.3459725132527448e-05,
+      "loss": 1.516,
+      "step": 330
+    },
+    {
+      "epoch": 1.6717171717171717,
+      "grad_norm": 0.35908318066713363,
+      "learning_rate": 1.3376785141607067e-05,
+      "loss": 1.5082,
+      "step": 331
+    },
+    {
+      "epoch": 1.676767676767677,
+      "grad_norm": 0.3600676479254767,
+      "learning_rate": 1.329389533612715e-05,
+      "loss": 1.6618,
+      "step": 332
+    },
+    {
+      "epoch": 1.6818181818181817,
+      "grad_norm": 0.3408967547242199,
+      "learning_rate": 1.3211058278817705e-05,
+      "loss": 1.5369,
+      "step": 333
+    },
+    {
+      "epoch": 1.6868686868686869,
+      "grad_norm": 0.3397113886527206,
+      "learning_rate": 1.3128276530777874e-05,
+      "loss": 1.5655,
+      "step": 334
+    },
+    {
+      "epoch": 1.691919191919192,
+      "grad_norm": 0.3552202693938084,
+      "learning_rate": 1.3045552651396805e-05,
+      "loss": 1.5913,
+      "step": 335
+    },
+    {
+      "epoch": 1.696969696969697,
+      "grad_norm": 0.3744742946582953,
+      "learning_rate": 1.2962889198274506e-05,
+      "loss": 1.5468,
+      "step": 336
+    },
+    {
+      "epoch": 1.702020202020202,
+      "grad_norm": 0.34653332317369845,
+      "learning_rate": 1.2880288727142757e-05,
+      "loss": 1.6004,
+      "step": 337
+    },
+    {
+      "epoch": 1.7070707070707072,
+      "grad_norm": 0.31916935496978377,
+      "learning_rate": 1.2797753791786112e-05,
+      "loss": 1.5545,
+      "step": 338
+    },
+    {
+      "epoch": 1.7121212121212122,
+      "grad_norm": 0.34268191483789684,
+      "learning_rate": 1.2715286943962925e-05,
+      "loss": 1.5286,
+      "step": 339
+    },
+    {
+      "epoch": 1.7171717171717171,
+      "grad_norm": 0.34934463675899574,
+      "learning_rate": 1.2632890733326475e-05,
+      "loss": 1.6462,
+      "step": 340
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 0.35309417659426007,
+      "learning_rate": 1.255056770734613e-05,
+      "loss": 1.6459,
+      "step": 341
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 0.36011757823593665,
+      "learning_rate": 1.2468320411228579e-05,
+      "loss": 1.5716,
+      "step": 342
+    },
+    {
+      "epoch": 1.7323232323232323,
+      "grad_norm": 0.31144567675642815,
+      "learning_rate": 1.2386151387839145e-05,
+      "loss": 1.4455,
+      "step": 343
+    },
+    {
+      "epoch": 1.7373737373737375,
+      "grad_norm": 0.3723102104349624,
+      "learning_rate": 1.2304063177623182e-05,
+      "loss": 1.6399,
+      "step": 344
+    },
+    {
+      "epoch": 1.7424242424242424,
+      "grad_norm": 0.38916725203253794,
+      "learning_rate": 1.2222058318527502e-05,
+      "loss": 1.553,
+      "step": 345
+    },
+    {
+      "epoch": 1.7474747474747474,
+      "grad_norm": 0.3322536259081415,
+      "learning_rate": 1.214013934592194e-05,
+      "loss": 1.492,
+      "step": 346
+    },
+    {
+      "epoch": 1.7525252525252526,
+      "grad_norm": 0.37090482924070856,
+      "learning_rate": 1.2058308792520937e-05,
+      "loss": 1.5703,
+      "step": 347
+    },
+    {
+      "epoch": 1.7575757575757576,
+      "grad_norm": 0.38587618815415825,
+      "learning_rate": 1.1976569188305255e-05,
+      "loss": 1.6146,
+      "step": 348
+    },
+    {
+      "epoch": 1.7626262626262625,
+      "grad_norm": 0.37194117916485847,
+      "learning_rate": 1.1894923060443763e-05,
+      "loss": 1.5854,
+      "step": 349
+    },
+    {
+      "epoch": 1.7676767676767677,
+      "grad_norm": 0.39535022296888556,
+      "learning_rate": 1.1813372933215274e-05,
+      "loss": 1.5918,
+      "step": 350
+    },
+    {
+      "epoch": 1.7676767676767677,
+      "eval_loss": 1.4877514839172363,
+      "eval_runtime": 229.0648,
+      "eval_samples_per_second": 6.247,
+      "eval_steps_per_second": 0.781,
+      "step": 350
+    },
+    {
+      "epoch": 1.7727272727272727,
+      "grad_norm": 0.35051004852162915,
+      "learning_rate": 1.1731921327930523e-05,
+      "loss": 1.6095,
+      "step": 351
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.34675700639977675,
+      "learning_rate": 1.165057076285423e-05,
+      "loss": 1.5565,
+      "step": 352
+    },
+    {
+      "epoch": 1.7828282828282829,
+      "grad_norm": 0.325905557218622,
+      "learning_rate": 1.1569323753127196e-05,
+      "loss": 1.5533,
+      "step": 353
+    },
+    {
+      "epoch": 1.7878787878787878,
+      "grad_norm": 0.3713917306740112,
+      "learning_rate": 1.1488182810688593e-05,
+      "loss": 1.5455,
+      "step": 354
+    },
+    {
+      "epoch": 1.7929292929292928,
+      "grad_norm": 0.405573404381939,
+      "learning_rate": 1.1407150444198262e-05,
+      "loss": 1.6257,
+      "step": 355
+    },
+    {
+      "epoch": 1.797979797979798,
+      "grad_norm": 0.34148070848500733,
+      "learning_rate": 1.132622915895917e-05,
+      "loss": 1.6975,
+      "step": 356
+    },
+    {
+      "epoch": 1.803030303030303,
+      "grad_norm": 0.34138980212543457,
+      "learning_rate": 1.1245421456839954e-05,
+      "loss": 1.5417,
+      "step": 357
+    },
+    {
+      "epoch": 1.808080808080808,
+      "grad_norm": 0.3458043789729876,
+      "learning_rate": 1.116472983619757e-05,
+      "loss": 1.6212,
+      "step": 358
+    },
+    {
+      "epoch": 1.8131313131313131,
+      "grad_norm": 0.34821285941193825,
+      "learning_rate": 1.1084156791800036e-05,
+      "loss": 1.6329,
+      "step": 359
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 0.35583032341592974,
+      "learning_rate": 1.100370481474932e-05,
+      "loss": 1.6058,
+      "step": 360
+    },
+    {
+      "epoch": 1.823232323232323,
+      "grad_norm": 0.3347559695941822,
+      "learning_rate": 1.0923376392404302e-05,
+      "loss": 1.5184,
+      "step": 361
+    },
+    {
+      "epoch": 1.8282828282828283,
+      "grad_norm": 0.3914463684785098,
+      "learning_rate": 1.0843174008303908e-05,
+      "loss": 1.6334,
+      "step": 362
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.3665474032055003,
+      "learning_rate": 1.0763100142090267e-05,
+      "loss": 1.6412,
+      "step": 363
+    },
+    {
+      "epoch": 1.8383838383838382,
+      "grad_norm": 0.3342017672784236,
+      "learning_rate": 1.0683157269432097e-05,
+      "loss": 1.6167,
+      "step": 364
+    },
+    {
+      "epoch": 1.8434343434343434,
+      "grad_norm": 0.3300768200411088,
+      "learning_rate": 1.0603347861948155e-05,
+      "loss": 1.6253,
+      "step": 365
+    },
+    {
+      "epoch": 1.8484848484848486,
+      "grad_norm": 0.3573589534814461,
+      "learning_rate": 1.0523674387130806e-05,
+      "loss": 1.5936,
+      "step": 366
+    },
+    {
+      "epoch": 1.8535353535353534,
+      "grad_norm": 0.3476694128643829,
+      "learning_rate": 1.0444139308269725e-05,
+      "loss": 1.6644,
+      "step": 367
+    },
+    {
+      "epoch": 1.8585858585858586,
+      "grad_norm": 0.38035554483291945,
+      "learning_rate": 1.036474508437579e-05,
+      "loss": 1.6168,
+      "step": 368
+    },
+    {
+      "epoch": 1.8636363636363638,
+      "grad_norm": 0.3334234919513792,
+      "learning_rate": 1.0285494170104996e-05,
+      "loss": 1.6633,
+      "step": 369
+    },
+    {
+      "epoch": 1.8686868686868687,
+      "grad_norm": 0.3265228255184419,
+      "learning_rate": 1.0206389015682601e-05,
+      "loss": 1.5896,
+      "step": 370
+    },
+    {
+      "epoch": 1.8737373737373737,
+      "grad_norm": 0.3691401309714038,
+      "learning_rate": 1.0127432066827357e-05,
+      "loss": 1.5916,
+      "step": 371
+    },
+    {
+      "epoch": 1.878787878787879,
+      "grad_norm": 0.33985591597069215,
+      "learning_rate": 1.0048625764675896e-05,
+      "loss": 1.6083,
+      "step": 372
+    },
+    {
+      "epoch": 1.8838383838383839,
+      "grad_norm": 0.3654336922923449,
+      "learning_rate": 9.969972545707266e-06,
+      "loss": 1.6688,
+      "step": 373
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.3474815931456802,
+      "learning_rate": 9.891474841667585e-06,
+      "loss": 1.6077,
+      "step": 374
+    },
+    {
+      "epoch": 1.893939393939394,
+      "grad_norm": 0.35105409408446747,
+      "learning_rate": 9.813135079494872e-06,
+      "loss": 1.6133,
+      "step": 375
+    },
+    {
+      "epoch": 1.898989898989899,
+      "grad_norm": 0.3412933734315999,
+      "learning_rate": 9.734955681244016e-06,
+      "loss": 1.6069,
+      "step": 376
+    },
+    {
+      "epoch": 1.904040404040404,
+      "grad_norm": 0.32783066972147346,
+      "learning_rate": 9.656939064011861e-06,
+      "loss": 1.4873,
+      "step": 377
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 0.3478856883264074,
+      "learning_rate": 9.579087639862538e-06,
+      "loss": 1.5366,
+      "step": 378
+    },
+    {
+      "epoch": 1.9141414141414141,
+      "grad_norm": 0.3539395477762644,
+      "learning_rate": 9.501403815752813e-06,
+      "loss": 1.5213,
+      "step": 379
+    },
+    {
+      "epoch": 1.9191919191919191,
+      "grad_norm": 0.3581622590288963,
+      "learning_rate": 9.423889993457721e-06,
+      "loss": 1.6273,
+      "step": 380
+    },
+    {
+      "epoch": 1.9242424242424243,
+      "grad_norm": 0.3335926440610502,
+      "learning_rate": 9.346548569496318e-06,
+      "loss": 1.6772,
+      "step": 381
+    },
+    {
+      "epoch": 1.9292929292929293,
+      "grad_norm": 0.3413862529782368,
+      "learning_rate": 9.269381935057546e-06,
+      "loss": 1.5744,
+      "step": 382
+    },
+    {
+      "epoch": 1.9343434343434343,
+      "grad_norm": 0.378369060255768,
+      "learning_rate": 9.192392475926315e-06,
+      "loss": 1.6305,
+      "step": 383
+    },
+    {
+      "epoch": 1.9393939393939394,
+      "grad_norm": 0.3508043937268879,
+      "learning_rate": 9.115582572409789e-06,
+      "loss": 1.6425,
+      "step": 384
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 0.3566255219446304,
+      "learning_rate": 9.038954599263713e-06,
+      "loss": 1.5535,
+      "step": 385
+    },
+    {
+      "epoch": 1.9494949494949494,
+      "grad_norm": 0.536928554074425,
+      "learning_rate": 8.962510925619065e-06,
+      "loss": 1.5138,
+      "step": 386
+    },
+    {
+      "epoch": 1.9545454545454546,
+      "grad_norm": 0.37533231697996555,
+      "learning_rate": 8.88625391490876e-06,
+      "loss": 1.5625,
+      "step": 387
+    },
+    {
+      "epoch": 1.9595959595959596,
+      "grad_norm": 0.3431698476036675,
+      "learning_rate": 8.8101859247946e-06,
+      "loss": 1.5796,
+      "step": 388
+    },
+    {
+      "epoch": 1.9646464646464645,
+      "grad_norm": 0.38914902758530917,
+      "learning_rate": 8.734309307094382e-06,
+      "loss": 1.6072,
+      "step": 389
+    },
+    {
+      "epoch": 1.9696969696969697,
+      "grad_norm": 0.3184503763537262,
+      "learning_rate": 8.658626407709182e-06,
+      "loss": 1.5288,
+      "step": 390
+    },
+    {
+      "epoch": 1.9747474747474747,
+      "grad_norm": 0.32890220573653406,
+      "learning_rate": 8.583139566550827e-06,
+      "loss": 1.592,
+      "step": 391
+    },
+    {
+      "epoch": 1.9797979797979797,
+      "grad_norm": 0.3395800990661379,
+      "learning_rate": 8.507851117469546e-06,
+      "loss": 1.5736,
+      "step": 392
+    },
+    {
+      "epoch": 1.9848484848484849,
+      "grad_norm": 0.34733690681577234,
+      "learning_rate": 8.432763388181812e-06,
+      "loss": 1.5791,
+      "step": 393
+    },
+    {
+      "epoch": 1.98989898989899,
+      "grad_norm": 0.37942192437254635,
+      "learning_rate": 8.357878700198407e-06,
+      "loss": 1.5927,
+      "step": 394
+    },
+    {
+      "epoch": 1.9949494949494948,
+      "grad_norm": 0.3285509965028684,
+      "learning_rate": 8.283199368752598e-06,
+      "loss": 1.5848,
+      "step": 395
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3237504993576643,
+      "learning_rate": 8.208727702728586e-06,
+      "loss": 1.5468,
+      "step": 396
+    },
+    {
+      "epoch": 2.005050505050505,
+      "grad_norm": 0.6489959146891373,
+      "learning_rate": 8.134466004590116e-06,
+      "loss": 1.3495,
+      "step": 397
+    },
+    {
+      "epoch": 2.01010101010101,
+      "grad_norm": 0.5603926475365507,
+      "learning_rate": 8.060416570309291e-06,
+      "loss": 1.3804,
+      "step": 398
+    },
+    {
+      "epoch": 2.015151515151515,
+      "grad_norm": 0.49584097590048715,
+      "learning_rate": 7.986581689295577e-06,
+      "loss": 1.3215,
+      "step": 399
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "grad_norm": 0.4392542856654279,
+      "learning_rate": 7.912963644325057e-06,
+      "loss": 1.3675,
+      "step": 400
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "eval_loss": 1.5459927320480347,
+      "eval_runtime": 229.9347,
+      "eval_samples_per_second": 6.224,
+      "eval_steps_per_second": 0.778,
+      "step": 400
+    },
+    {
+      "epoch": 2.025252525252525,
+      "grad_norm": 0.5550727689847038,
+      "learning_rate": 7.839564711469786e-06,
+      "loss": 1.3907,
+      "step": 401
+    },
+    {
+      "epoch": 2.0303030303030303,
+      "grad_norm": 0.6935079495968718,
+      "learning_rate": 7.766387160027504e-06,
+      "loss": 1.4597,
+      "step": 402
+    },
+    {
+      "epoch": 2.0353535353535355,
+      "grad_norm": 0.5919993820181615,
+      "learning_rate": 7.693433252451404e-06,
+      "loss": 1.3068,
+      "step": 403
+    },
+    {
+      "epoch": 2.04040404040404,
+      "grad_norm": 0.4444393538098348,
+      "learning_rate": 7.620705244280208e-06,
+      "loss": 1.3455,
+      "step": 404
+    },
+    {
+      "epoch": 2.0454545454545454,
+      "grad_norm": 0.44018034109403575,
+      "learning_rate": 7.548205384068461e-06,
+      "loss": 1.4318,
+      "step": 405
+    },
+    {
+      "epoch": 2.0505050505050506,
+      "grad_norm": 0.47554567734474,
+      "learning_rate": 7.475935913316967e-06,
+      "loss": 1.3565,
+      "step": 406
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 0.4485832265306745,
+      "learning_rate": 7.403899066403506e-06,
+      "loss": 1.3344,
+      "step": 407
+    },
+    {
+      "epoch": 2.0606060606060606,
+      "grad_norm": 0.4155215847745916,
+      "learning_rate": 7.332097070513776e-06,
+      "loss": 1.3103,
+      "step": 408
+    },
+    {
+      "epoch": 2.0656565656565657,
+      "grad_norm": 0.3796385213739081,
+      "learning_rate": 7.260532145572487e-06,
+      "loss": 1.3414,
+      "step": 409
+    },
+    {
+      "epoch": 2.0707070707070705,
+      "grad_norm": 0.40547334324608847,
+      "learning_rate": 7.1892065041747645e-06,
+      "loss": 1.3815,
+      "step": 410
+    },
+    {
+      "epoch": 2.0757575757575757,
+      "grad_norm": 0.39981507933068694,
+      "learning_rate": 7.118122351517729e-06,
+      "loss": 1.3698,
+      "step": 411
+    },
+    {
+      "epoch": 2.080808080808081,
+      "grad_norm": 0.380701129157713,
+      "learning_rate": 7.047281885332311e-06,
+      "loss": 1.348,
+      "step": 412
+    },
+    {
+      "epoch": 2.0858585858585856,
+      "grad_norm": 0.3836974581541541,
+      "learning_rate": 6.976687295815327e-06,
+      "loss": 1.3143,
+      "step": 413
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 0.3778921933523212,
+      "learning_rate": 6.906340765561734e-06,
+      "loss": 1.3396,
+      "step": 414
+    },
+    {
+      "epoch": 2.095959595959596,
+      "grad_norm": 0.38922397224846605,
+      "learning_rate": 6.836244469497159e-06,
+      "loss": 1.4238,
+      "step": 415
+    },
+    {
+      "epoch": 2.101010101010101,
+      "grad_norm": 0.35783821916056857,
+      "learning_rate": 6.7664005748106875e-06,
+      "loss": 1.3582,
+      "step": 416
+    },
+    {
+      "epoch": 2.106060606060606,
+      "grad_norm": 0.35939456738266723,
+      "learning_rate": 6.69681124088779e-06,
+      "loss": 1.3783,
+      "step": 417
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 0.3498463091785103,
+      "learning_rate": 6.627478619243642e-06,
+      "loss": 1.3671,
+      "step": 418
+    },
+    {
+      "epoch": 2.1161616161616164,
+      "grad_norm": 0.35085992990906917,
+      "learning_rate": 6.558404853456545e-06,
+      "loss": 1.2929,
+      "step": 419
+    },
+    {
+      "epoch": 2.121212121212121,
+      "grad_norm": 0.36844923629881426,
+      "learning_rate": 6.489592079101671e-06,
+      "loss": 1.422,
+      "step": 420
+    },
+    {
+      "epoch": 2.1262626262626263,
+      "grad_norm": 0.3459330366665105,
+      "learning_rate": 6.421042423685059e-06,
+      "loss": 1.3818,
+      "step": 421
+    },
+    {
+      "epoch": 2.1313131313131315,
+      "grad_norm": 0.34989994293169857,
+      "learning_rate": 6.352758006577794e-06,
+      "loss": 1.4177,
+      "step": 422
+    },
+    {
+      "epoch": 2.1363636363636362,
+      "grad_norm": 0.3267726260945404,
+      "learning_rate": 6.284740938950517e-06,
+      "loss": 1.3085,
+      "step": 423
+    },
+    {
+      "epoch": 2.1414141414141414,
+      "grad_norm": 0.3513152908110021,
+      "learning_rate": 6.2169933237081386e-06,
+      "loss": 1.4144,
+      "step": 424
+    },
+    {
+      "epoch": 2.1464646464646466,
+      "grad_norm": 0.35199489451607285,
+      "learning_rate": 6.149517255424823e-06,
+      "loss": 1.3202,
+      "step": 425
+    },
+    {
+      "epoch": 2.1515151515151514,
+      "grad_norm": 0.35307815632276385,
+      "learning_rate": 6.0823148202792474e-06,
+      "loss": 1.3623,
+      "step": 426
+    },
+    {
+      "epoch": 2.1565656565656566,
+      "grad_norm": 0.3470711355464423,
+      "learning_rate": 6.015388095990069e-06,
+      "loss": 1.3437,
+      "step": 427
+    },
+    {
+      "epoch": 2.1616161616161618,
+      "grad_norm": 0.3362602033819937,
+      "learning_rate": 5.948739151751716e-06,
+      "loss": 1.3541,
+      "step": 428
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.34033946687573235,
+      "learning_rate": 5.882370048170403e-06,
+      "loss": 1.4119,
+      "step": 429
+    },
+    {
+      "epoch": 2.1717171717171717,
+      "grad_norm": 0.3532405636078066,
+      "learning_rate": 5.81628283720042e-06,
+      "loss": 1.4835,
+      "step": 430
+    },
+    {
+      "epoch": 2.176767676767677,
+      "grad_norm": 0.32887202826818873,
+      "learning_rate": 5.7504795620806964e-06,
+      "loss": 1.3296,
+      "step": 431
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.3411420041143714,
+      "learning_rate": 5.684962257271636e-06,
+      "loss": 1.3661,
+      "step": 432
+    },
+    {
+      "epoch": 2.186868686868687,
+      "grad_norm": 0.31237309569104227,
+      "learning_rate": 5.619732948392197e-06,
+      "loss": 1.2338,
+      "step": 433
+    },
+    {
+      "epoch": 2.191919191919192,
+      "grad_norm": 0.3413488930865825,
+      "learning_rate": 5.55479365215729e-06,
+      "loss": 1.3423,
+      "step": 434
+    },
+    {
+      "epoch": 2.196969696969697,
+      "grad_norm": 0.3569265566015804,
+      "learning_rate": 5.490146376315405e-06,
+      "loss": 1.3593,
+      "step": 435
+    },
+    {
+      "epoch": 2.202020202020202,
+      "grad_norm": 0.36112224488349237,
+      "learning_rate": 5.42579311958655e-06,
+      "loss": 1.302,
+      "step": 436
+    },
+    {
+      "epoch": 2.207070707070707,
+      "grad_norm": 0.3276471583119553,
+      "learning_rate": 5.361735871600451e-06,
+      "loss": 1.3311,
+      "step": 437
+    },
+    {
+      "epoch": 2.212121212121212,
+      "grad_norm": 0.3156487456237961,
+      "learning_rate": 5.297976612835038e-06,
+      "loss": 1.3156,
+      "step": 438
+    },
+    {
+      "epoch": 2.217171717171717,
+      "grad_norm": 0.36001007809681823,
+      "learning_rate": 5.234517314555213e-06,
+      "loss": 1.2172,
+      "step": 439
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.35044457797610973,
+      "learning_rate": 5.171359938751919e-06,
+      "loss": 1.3586,
+      "step": 440
+    },
+    {
+      "epoch": 2.227272727272727,
+      "grad_norm": 0.3408481018485263,
+      "learning_rate": 5.108506438081457e-06,
+      "loss": 1.386,
+      "step": 441
+    },
+    {
+      "epoch": 2.2323232323232323,
+      "grad_norm": 0.5110825954105327,
+      "learning_rate": 5.045958755805123e-06,
+      "loss": 1.3849,
+      "step": 442
+    },
+    {
+      "epoch": 2.2373737373737375,
+      "grad_norm": 0.33198011548599254,
+      "learning_rate": 4.983718825729138e-06,
+      "loss": 1.3877,
+      "step": 443
+    },
+    {
+      "epoch": 2.242424242424242,
+      "grad_norm": 0.338000968148193,
+      "learning_rate": 4.921788572144841e-06,
+      "loss": 1.328,
+      "step": 444
+    },
+    {
+      "epoch": 2.2474747474747474,
+      "grad_norm": 0.3381841613090944,
+      "learning_rate": 4.860169909769223e-06,
+      "loss": 1.3864,
+      "step": 445
+    },
+    {
+      "epoch": 2.2525252525252526,
+      "grad_norm": 0.33987310454638214,
+      "learning_rate": 4.798864743685704e-06,
+      "loss": 1.3087,
+      "step": 446
+    },
+    {
+      "epoch": 2.257575757575758,
+      "grad_norm": 0.3312906048392972,
+      "learning_rate": 4.73787496928523e-06,
+      "loss": 1.3784,
+      "step": 447
+    },
+    {
+      "epoch": 2.2626262626262625,
+      "grad_norm": 0.3471224148442727,
+      "learning_rate": 4.677202472207706e-06,
+      "loss": 1.3469,
+      "step": 448
+    },
+    {
+      "epoch": 2.2676767676767677,
+      "grad_norm": 0.33785032384606045,
+      "learning_rate": 4.616849128283658e-06,
+      "loss": 1.3877,
+      "step": 449
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 0.34074480809570873,
+      "learning_rate": 4.556816803476261e-06,
+      "loss": 1.3675,
+      "step": 450
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "eval_loss": 1.5548166036605835,
+      "eval_runtime": 230.2364,
+      "eval_samples_per_second": 6.215,
+      "eval_steps_per_second": 0.777,
+      "step": 450
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 0.32624573777354227,
+      "learning_rate": 4.497107353823638e-06,
+      "loss": 1.3828,
+      "step": 451
+    },
+    {
+      "epoch": 2.282828282828283,
+      "grad_norm": 0.3369163688557624,
+      "learning_rate": 4.437722625381481e-06,
+      "loss": 1.3295,
+      "step": 452
+    },
+    {
+      "epoch": 2.287878787878788,
+      "grad_norm": 0.3355454228013588,
+      "learning_rate": 4.378664454165983e-06,
+      "loss": 1.3544,
+      "step": 453
+    },
+    {
+      "epoch": 2.292929292929293,
+      "grad_norm": 0.348354298966465,
+      "learning_rate": 4.319934666097055e-06,
+      "loss": 1.3835,
+      "step": 454
+    },
+    {
+      "epoch": 2.297979797979798,
+      "grad_norm": 0.3305159849956163,
+      "learning_rate": 4.261535076941888e-06,
+      "loss": 1.2923,
+      "step": 455
+    },
+    {
+      "epoch": 2.303030303030303,
+      "grad_norm": 0.3240680403742593,
+      "learning_rate": 4.20346749225881e-06,
+      "loss": 1.3421,
+      "step": 456
+    },
+    {
+      "epoch": 2.308080808080808,
+      "grad_norm": 0.35833405093271425,
+      "learning_rate": 4.145733707341457e-06,
+      "loss": 1.4357,
+      "step": 457
+    },
+    {
+      "epoch": 2.313131313131313,
+      "grad_norm": 0.3454656348650401,
+      "learning_rate": 4.08833550716329e-06,
+      "loss": 1.3921,
+      "step": 458
+    },
+    {
+      "epoch": 2.3181818181818183,
+      "grad_norm": 0.32930254998399916,
+      "learning_rate": 4.031274666322372e-06,
+      "loss": 1.3532,
+      "step": 459
+    },
+    {
+      "epoch": 2.323232323232323,
+      "grad_norm": 0.35468933367330896,
+      "learning_rate": 3.9745529489865325e-06,
+      "loss": 1.3409,
+      "step": 460
+    },
+    {
+      "epoch": 2.3282828282828283,
+      "grad_norm": 0.3407135461627606,
+      "learning_rate": 3.918172108838819e-06,
+      "loss": 1.3504,
+      "step": 461
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.3174290975875319,
+      "learning_rate": 3.862133889023264e-06,
+      "loss": 1.2801,
+      "step": 462
+    },
+    {
+      "epoch": 2.3383838383838382,
+      "grad_norm": 0.33257974936544793,
+      "learning_rate": 3.806440022090991e-06,
+      "loss": 1.3476,
+      "step": 463
+    },
+    {
+      "epoch": 2.3434343434343434,
+      "grad_norm": 0.34433935520925646,
+      "learning_rate": 3.7510922299466818e-06,
+      "loss": 1.4109,
+      "step": 464
+    },
+    {
+      "epoch": 2.3484848484848486,
+      "grad_norm": 0.3329914479718482,
+      "learning_rate": 3.696092223795293e-06,
+      "loss": 1.3107,
+      "step": 465
+    },
+    {
+      "epoch": 2.3535353535353534,
+      "grad_norm": 0.34511424002581903,
+      "learning_rate": 3.641441704089195e-06,
+      "loss": 1.344,
+      "step": 466
+    },
+    {
+      "epoch": 2.3585858585858586,
+      "grad_norm": 0.3225246247472432,
+      "learning_rate": 3.5871423604755576e-06,
+      "loss": 1.379,
+      "step": 467
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 0.3314079832160016,
+      "learning_rate": 3.53319587174414e-06,
+      "loss": 1.3605,
+      "step": 468
+    },
+    {
+      "epoch": 2.3686868686868685,
+      "grad_norm": 0.33155417660439823,
+      "learning_rate": 3.4796039057753703e-06,
+      "loss": 1.3788,
+      "step": 469
+    },
+    {
+      "epoch": 2.3737373737373737,
+      "grad_norm": 0.3217080508265444,
+      "learning_rate": 3.4263681194887926e-06,
+      "loss": 1.2229,
+      "step": 470
+    },
+    {
+      "epoch": 2.378787878787879,
+      "grad_norm": 0.32856611753986226,
+      "learning_rate": 3.3734901587918234e-06,
+      "loss": 1.3069,
+      "step": 471
+    },
+    {
+      "epoch": 2.3838383838383836,
+      "grad_norm": 0.33453934067183627,
+      "learning_rate": 3.320971658528889e-06,
+      "loss": 1.3685,
+      "step": 472
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 0.3195781522323383,
+      "learning_rate": 3.2688142424308464e-06,
+      "loss": 1.3335,
+      "step": 473
+    },
+    {
+      "epoch": 2.393939393939394,
+      "grad_norm": 0.33384250654885655,
+      "learning_rate": 3.217019523064825e-06,
+      "loss": 1.3149,
+      "step": 474
+    },
+    {
+      "epoch": 2.398989898989899,
+      "grad_norm": 0.3417470523103563,
+      "learning_rate": 3.1655891017843152e-06,
+      "loss": 1.4034,
+      "step": 475
+    },
+    {
+      "epoch": 2.404040404040404,
+      "grad_norm": 0.3192940536310456,
+      "learning_rate": 3.1145245686797107e-06,
+      "loss": 1.3257,
+      "step": 476
+    },
+    {
+      "epoch": 2.409090909090909,
+      "grad_norm": 0.33117280629763984,
+      "learning_rate": 3.063827502529125e-06,
+      "loss": 1.3705,
+      "step": 477
+    },
+    {
+      "epoch": 2.4141414141414144,
+      "grad_norm": 0.31577233152801576,
+      "learning_rate": 3.0134994707495776e-06,
+      "loss": 1.3051,
+      "step": 478
+    },
+    {
+      "epoch": 2.419191919191919,
+      "grad_norm": 0.32359860321254813,
+      "learning_rate": 2.96354202934853e-06,
+      "loss": 1.3105,
+      "step": 479
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "grad_norm": 0.3323495077715317,
+      "learning_rate": 2.913956722875803e-06,
+      "loss": 1.3609,
+      "step": 480
+    },
+    {
+      "epoch": 2.429292929292929,
+      "grad_norm": 0.34559349440125164,
+      "learning_rate": 2.86474508437579e-06,
+      "loss": 1.3739,
+      "step": 481
+    },
+    {
+      "epoch": 2.4343434343434343,
+      "grad_norm": 0.3229886400387147,
+      "learning_rate": 2.8159086353400787e-06,
+      "loss": 1.2993,
+      "step": 482
+    },
+    {
+      "epoch": 2.4393939393939394,
+      "grad_norm": 0.3151626607633294,
+      "learning_rate": 2.7674488856604106e-06,
+      "loss": 1.2994,
+      "step": 483
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.33940231127513143,
+      "learning_rate": 2.7193673335819893e-06,
+      "loss": 1.3887,
+      "step": 484
+    },
+    {
+      "epoch": 2.4494949494949494,
+      "grad_norm": 0.3314650305858189,
+      "learning_rate": 2.6716654656571766e-06,
+      "loss": 1.2696,
+      "step": 485
+    },
+    {
+      "epoch": 2.4545454545454546,
+      "grad_norm": 0.3283054013523294,
+      "learning_rate": 2.6243447566995075e-06,
+      "loss": 1.3248,
+      "step": 486
+    },
+    {
+      "epoch": 2.45959595959596,
+      "grad_norm": 0.3294385291650391,
+      "learning_rate": 2.5774066697381173e-06,
+      "loss": 1.3744,
+      "step": 487
+    },
+    {
+      "epoch": 2.4646464646464645,
+      "grad_norm": 0.32701883167697327,
+      "learning_rate": 2.530852655972487e-06,
+      "loss": 1.4138,
+      "step": 488
+    },
+    {
+      "epoch": 2.4696969696969697,
+      "grad_norm": 0.334931690168335,
+      "learning_rate": 2.4846841547275916e-06,
+      "loss": 1.3611,
+      "step": 489
+    },
+    {
+      "epoch": 2.474747474747475,
+      "grad_norm": 0.33546755336530826,
+      "learning_rate": 2.4389025934094045e-06,
+      "loss": 1.268,
+      "step": 490
+    },
+    {
+      "epoch": 2.4797979797979797,
+      "grad_norm": 0.31726122931536993,
+      "learning_rate": 2.39350938746074e-06,
+      "loss": 1.3288,
+      "step": 491
+    },
+    {
+      "epoch": 2.484848484848485,
+      "grad_norm": 0.3287627054478962,
+      "learning_rate": 2.348505940317516e-06,
+      "loss": 1.4282,
+      "step": 492
+    },
+    {
+      "epoch": 2.48989898989899,
+      "grad_norm": 0.32239418628280586,
+      "learning_rate": 2.3038936433653617e-06,
+      "loss": 1.3585,
+      "step": 493
+    },
+    {
+      "epoch": 2.494949494949495,
+      "grad_norm": 0.31867368014934067,
+      "learning_rate": 2.259673875896585e-06,
+      "loss": 1.279,
+      "step": 494
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.34355794868397255,
+      "learning_rate": 2.2158480050675407e-06,
+      "loss": 1.3742,
+      "step": 495
+    },
+    {
+      "epoch": 2.505050505050505,
+      "grad_norm": 0.32224758819615007,
+      "learning_rate": 2.1724173858563546e-06,
+      "loss": 1.2032,
+      "step": 496
+    },
+    {
+      "epoch": 2.51010101010101,
+      "grad_norm": 0.338860108565893,
+      "learning_rate": 2.129383361021038e-06,
+      "loss": 1.3843,
+      "step": 497
+    },
+    {
+      "epoch": 2.515151515151515,
+      "grad_norm": 0.3358553477580223,
+      "learning_rate": 2.0867472610579753e-06,
+      "loss": 1.3089,
+      "step": 498
+    },
+    {
+      "epoch": 2.5202020202020203,
+      "grad_norm": 0.3343623935360769,
+      "learning_rate": 2.044510404160774e-06,
+      "loss": 1.3487,
+      "step": 499
+    },
+    {
+      "epoch": 2.525252525252525,
+      "grad_norm": 0.3546274911997211,
+      "learning_rate": 2.002674096179525e-06,
+      "loss": 1.3862,
+      "step": 500
+    },
+    {
+      "epoch": 2.525252525252525,
+      "eval_loss": 1.5534101724624634,
+      "eval_runtime": 230.0945,
+      "eval_samples_per_second": 6.219,
+      "eval_steps_per_second": 0.778,
+      "step": 500
+    },
+    {
+      "epoch": 2.5303030303030303,
+      "grad_norm": 0.33054033498676905,
+      "learning_rate": 1.961239630580419e-06,
+      "loss": 1.3379,
+      "step": 501
+    },
+    {
+      "epoch": 2.5353535353535355,
+      "grad_norm": 0.3243578841256009,
+      "learning_rate": 1.9202082884057663e-06,
+      "loss": 1.361,
+      "step": 502
+    },
+    {
+      "epoch": 2.5404040404040407,
+      "grad_norm": 0.3453871510381904,
+      "learning_rate": 1.8795813382343757e-06,
+      "loss": 1.3128,
+      "step": 503
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 0.3240177636684765,
+      "learning_rate": 1.8393600361423534e-06,
+      "loss": 1.2967,
+      "step": 504
+    },
+    {
+      "epoch": 2.5505050505050506,
+      "grad_norm": 0.32198961074160476,
+      "learning_rate": 1.7995456256642467e-06,
+      "loss": 1.3792,
+      "step": 505
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 0.3323091802115782,
+      "learning_rate": 1.7601393377546189e-06,
+      "loss": 1.3401,
+      "step": 506
+    },
+    {
+      "epoch": 2.5606060606060606,
+      "grad_norm": 0.33782378441397226,
+      "learning_rate": 1.7211423907499729e-06,
+      "loss": 1.39,
+      "step": 507
+    },
+    {
+      "epoch": 2.5656565656565657,
+      "grad_norm": 0.3173070512019593,
+      "learning_rate": 1.6825559903310827e-06,
+      "loss": 1.3419,
+      "step": 508
+    },
+    {
+      "epoch": 2.570707070707071,
+      "grad_norm": 0.328610068013649,
+      "learning_rate": 1.6443813294857452e-06,
+      "loss": 1.3739,
+      "step": 509
+    },
+    {
+      "epoch": 2.5757575757575757,
+      "grad_norm": 0.3275374111063948,
+      "learning_rate": 1.6066195884718631e-06,
+      "loss": 1.2777,
+      "step": 510
+    },
+    {
+      "epoch": 2.580808080808081,
+      "grad_norm": 0.35141551632003326,
+      "learning_rate": 1.5692719347809686e-06,
+      "loss": 1.3861,
+      "step": 511
+    },
+    {
+      "epoch": 2.5858585858585856,
+      "grad_norm": 0.31546480930066206,
+      "learning_rate": 1.5323395231021325e-06,
+      "loss": 1.2477,
+      "step": 512
+    },
+    {
+      "epoch": 2.590909090909091,
+      "grad_norm": 0.33265355608600455,
+      "learning_rate": 1.4958234952862553e-06,
+      "loss": 1.4034,
+      "step": 513
+    },
+    {
+      "epoch": 2.595959595959596,
+      "grad_norm": 0.35507158066405115,
+      "learning_rate": 1.459724980310767e-06,
+      "loss": 1.3902,
+      "step": 514
+    },
+    {
+      "epoch": 2.601010101010101,
+      "grad_norm": 0.3295175897866719,
+      "learning_rate": 1.4240450942447213e-06,
+      "loss": 1.373,
+      "step": 515
+    },
+    {
+      "epoch": 2.606060606060606,
+      "grad_norm": 0.32252779319703984,
+      "learning_rate": 1.388784940214292e-06,
+      "loss": 1.3497,
+      "step": 516
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 0.3402791161335901,
+      "learning_rate": 1.3539456083686736e-06,
+      "loss": 1.3999,
+      "step": 517
+    },
+    {
+      "epoch": 2.616161616161616,
+      "grad_norm": 0.3279145522242951,
+      "learning_rate": 1.3195281758463624e-06,
+      "loss": 1.3533,
+      "step": 518
+    },
+    {
+      "epoch": 2.621212121212121,
+      "grad_norm": 0.32392571532268805,
+      "learning_rate": 1.2855337067418576e-06,
+      "loss": 1.3212,
+      "step": 519
+    },
+    {
+      "epoch": 2.6262626262626263,
+      "grad_norm": 0.325724061463982,
+      "learning_rate": 1.2519632520727825e-06,
+      "loss": 1.3995,
+      "step": 520
+    },
+    {
+      "epoch": 2.6313131313131315,
+      "grad_norm": 0.33044089457213144,
+      "learning_rate": 1.2188178497473513e-06,
+      "loss": 1.3527,
+      "step": 521
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 0.3249288237340594,
+      "learning_rate": 1.1860985245323237e-06,
+      "loss": 1.4036,
+      "step": 522
+    },
+    {
+      "epoch": 2.6414141414141414,
+      "grad_norm": 0.3316367696024126,
+      "learning_rate": 1.1538062880212868e-06,
+      "loss": 1.4117,
+      "step": 523
+    },
+    {
+      "epoch": 2.6464646464646466,
+      "grad_norm": 0.330308972754837,
+      "learning_rate": 1.1219421386033958e-06,
+      "loss": 1.3664,
+      "step": 524
+    },
+    {
+      "epoch": 2.6515151515151514,
+      "grad_norm": 0.32101693883494814,
+      "learning_rate": 1.090507061432507e-06,
+      "loss": 1.3609,
+      "step": 525
+    },
+    {
+      "epoch": 2.6565656565656566,
+      "grad_norm": 0.31721696490080514,
+      "learning_rate": 1.059502028396714e-06,
+      "loss": 1.4163,
+      "step": 526
+    },
+    {
+      "epoch": 2.6616161616161618,
+      "grad_norm": 0.33100041836983457,
+      "learning_rate": 1.028927998088302e-06,
+      "loss": 1.373,
+      "step": 527
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.3087065560652621,
+      "learning_rate": 9.987859157741063e-07,
+      "loss": 1.3462,
+      "step": 528
+    },
+    {
+      "epoch": 2.6717171717171717,
+      "grad_norm": 0.3228221395167405,
+      "learning_rate": 9.690767133662976e-07,
+      "loss": 1.2952,
+      "step": 529
+    },
+    {
+      "epoch": 2.676767676767677,
+      "grad_norm": 0.34478777480046996,
+      "learning_rate": 9.398013093935604e-07,
+      "loss": 1.3755,
+      "step": 530
+    },
+    {
+      "epoch": 2.6818181818181817,
+      "grad_norm": 0.32420291241316956,
+      "learning_rate": 9.10960608972698e-07,
+      "loss": 1.3832,
+      "step": 531
+    },
+    {
+      "epoch": 2.686868686868687,
+      "grad_norm": 0.33022446961420726,
+      "learning_rate": 8.825555037806493e-07,
+      "loss": 1.347,
+      "step": 532
+    },
+    {
+      "epoch": 2.691919191919192,
+      "grad_norm": 0.33958181683593025,
+      "learning_rate": 8.545868720269145e-07,
+      "loss": 1.3724,
+      "step": 533
+    },
+    {
+      "epoch": 2.6969696969696972,
+      "grad_norm": 0.33725010658515825,
+      "learning_rate": 8.270555784264167e-07,
+      "loss": 1.2772,
+      "step": 534
+    },
+    {
+      "epoch": 2.702020202020202,
+      "grad_norm": 0.3515586496897235,
+      "learning_rate": 7.999624741727479e-07,
+      "loss": 1.4157,
+      "step": 535
+    },
+    {
+      "epoch": 2.707070707070707,
+      "grad_norm": 0.3315645750045216,
+      "learning_rate": 7.733083969118759e-07,
+      "loss": 1.3699,
+      "step": 536
+    },
+    {
+      "epoch": 2.712121212121212,
+      "grad_norm": 0.3265957901814612,
+      "learning_rate": 7.470941707162226e-07,
+      "loss": 1.3797,
+      "step": 537
+    },
+    {
+      "epoch": 2.717171717171717,
+      "grad_norm": 0.3209125002172067,
+      "learning_rate": 7.213206060592064e-07,
+      "loss": 1.3694,
+      "step": 538
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 0.32513518265436614,
+      "learning_rate": 6.959884997901705e-07,
+      "loss": 1.2818,
+      "step": 539
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.31979488074104667,
+      "learning_rate": 6.710986351097493e-07,
+      "loss": 1.2585,
+      "step": 540
+    },
+    {
+      "epoch": 2.7323232323232323,
+      "grad_norm": 0.323915220461446,
+      "learning_rate": 6.466517815456607e-07,
+      "loss": 1.2766,
+      "step": 541
+    },
+    {
+      "epoch": 2.7373737373737375,
+      "grad_norm": 0.3272707694127074,
+      "learning_rate": 6.226486949289079e-07,
+      "loss": 1.2811,
+      "step": 542
+    },
+    {
+      "epoch": 2.742424242424242,
+      "grad_norm": 0.32467673338245945,
+      "learning_rate": 5.990901173704083e-07,
+      "loss": 1.3697,
+      "step": 543
+    },
+    {
+      "epoch": 2.7474747474747474,
+      "grad_norm": 0.331433447678859,
+      "learning_rate": 5.759767772380648e-07,
+      "loss": 1.4064,
+      "step": 544
+    },
+    {
+      "epoch": 2.7525252525252526,
+      "grad_norm": 0.3439032539794961,
+      "learning_rate": 5.533093891342262e-07,
+      "loss": 1.3781,
+      "step": 545
+    },
+    {
+      "epoch": 2.757575757575758,
+      "grad_norm": 0.3223680031666151,
+      "learning_rate": 5.310886538736037e-07,
+      "loss": 1.3459,
+      "step": 546
+    },
+    {
+      "epoch": 2.7626262626262625,
+      "grad_norm": 0.3219629931268615,
+      "learning_rate": 5.093152584616101e-07,
+      "loss": 1.3199,
+      "step": 547
+    },
+    {
+      "epoch": 2.7676767676767677,
+      "grad_norm": 0.30885242518385303,
+      "learning_rate": 4.879898760731028e-07,
+      "loss": 1.3183,
+      "step": 548
+    },
+    {
+      "epoch": 2.7727272727272725,
+      "grad_norm": 0.3671027111648377,
+      "learning_rate": 4.6711316603159084e-07,
+      "loss": 1.3361,
+      "step": 549
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.31733325019140657,
+      "learning_rate": 4.46685773788833e-07,
+      "loss": 1.3552,
+      "step": 550
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "eval_loss": 1.5522303581237793,
+      "eval_runtime": 230.1843,
+      "eval_samples_per_second": 6.217,
+      "eval_steps_per_second": 0.778,
+      "step": 550
+    },
+    {
+      "epoch": 2.782828282828283,
+      "grad_norm": 0.3200707645533683,
+      "learning_rate": 4.2670833090489193e-07,
+      "loss": 1.2946,
+      "step": 551
+    },
+    {
+      "epoch": 2.787878787878788,
+      "grad_norm": 0.331980900223843,
+      "learning_rate": 4.0718145502860627e-07,
+      "loss": 1.3345,
+      "step": 552
+    },
+    {
+      "epoch": 2.792929292929293,
+      "grad_norm": 0.3282199495755412,
+      "learning_rate": 3.8810574987849126e-07,
+      "loss": 1.3024,
+      "step": 553
+    },
+    {
+      "epoch": 2.797979797979798,
+      "grad_norm": 0.3175787008542941,
+      "learning_rate": 3.6948180522408006e-07,
+      "loss": 1.3115,
+      "step": 554
+    },
+    {
+      "epoch": 2.8030303030303028,
+      "grad_norm": 0.3179435007333598,
+      "learning_rate": 3.513101968676802e-07,
+      "loss": 1.3098,
+      "step": 555
+    },
+    {
+      "epoch": 2.808080808080808,
+      "grad_norm": 0.33169394245798245,
+      "learning_rate": 3.3359148662658103e-07,
+      "loss": 1.3372,
+      "step": 556
+    },
+    {
+      "epoch": 2.813131313131313,
+      "grad_norm": 0.33115578745917224,
+      "learning_rate": 3.163262223156793e-07,
+      "loss": 1.3867,
+      "step": 557
+    },
+    {
+      "epoch": 2.8181818181818183,
+      "grad_norm": 0.31264293668678106,
+      "learning_rate": 2.9951493773053953e-07,
+      "loss": 1.3085,
+      "step": 558
+    },
+    {
+      "epoch": 2.823232323232323,
+      "grad_norm": 0.3401717880353158,
+      "learning_rate": 2.831581526308935e-07,
+      "loss": 1.3873,
+      "step": 559
+    },
+    {
+      "epoch": 2.8282828282828283,
+      "grad_norm": 0.33495812017470433,
+      "learning_rate": 2.6725637272457527e-07,
+      "loss": 1.4253,
+      "step": 560
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.33801260984820947,
+      "learning_rate": 2.518100896518766e-07,
+      "loss": 1.3547,
+      "step": 561
+    },
+    {
+      "epoch": 2.8383838383838382,
+      "grad_norm": 0.3348385881354705,
+      "learning_rate": 2.3681978097035263e-07,
+      "loss": 1.4229,
+      "step": 562
+    },
+    {
+      "epoch": 2.8434343434343434,
+      "grad_norm": 0.31719634542323716,
+      "learning_rate": 2.222859101400554e-07,
+      "loss": 1.2818,
+      "step": 563
+    },
+    {
+      "epoch": 2.8484848484848486,
+      "grad_norm": 0.3183857279402652,
+      "learning_rate": 2.0820892650920686e-07,
+      "loss": 1.3834,
+      "step": 564
+    },
+    {
+      "epoch": 2.8535353535353534,
+      "grad_norm": 0.37637626910255345,
+      "learning_rate": 1.9458926530030675e-07,
+      "loss": 1.3882,
+      "step": 565
+    },
+    {
+      "epoch": 2.8585858585858586,
+      "grad_norm": 0.30878369912934317,
+      "learning_rate": 1.8142734759666823e-07,
+      "loss": 1.2498,
+      "step": 566
+    },
+    {
+      "epoch": 2.8636363636363638,
+      "grad_norm": 0.328503940896062,
+      "learning_rate": 1.6872358032941014e-07,
+      "loss": 1.3391,
+      "step": 567
+    },
+    {
+      "epoch": 2.8686868686868685,
+      "grad_norm": 0.3185916122047658,
+      "learning_rate": 1.5647835626487194e-07,
+      "loss": 1.3697,
+      "step": 568
+    },
+    {
+      "epoch": 2.8737373737373737,
+      "grad_norm": 0.333090152305754,
+      "learning_rate": 1.4469205399246844e-07,
+      "loss": 1.2437,
+      "step": 569
+    },
+    {
+      "epoch": 2.878787878787879,
+      "grad_norm": 0.31288406908685207,
+      "learning_rate": 1.3336503791298582e-07,
+      "loss": 1.2924,
+      "step": 570
+    },
+    {
+      "epoch": 2.883838383838384,
+      "grad_norm": 0.3096586778222418,
+      "learning_rate": 1.2249765822731574e-07,
+      "loss": 1.3853,
+      "step": 571
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.3310184954270413,
+      "learning_rate": 1.1209025092562719e-07,
+      "loss": 1.3591,
+      "step": 572
+    },
+    {
+      "epoch": 2.893939393939394,
+      "grad_norm": 0.312664849919579,
+      "learning_rate": 1.0214313777698325e-07,
+      "loss": 1.3503,
+      "step": 573
+    },
+    {
+      "epoch": 2.898989898989899,
+      "grad_norm": 0.3232002984438332,
+      "learning_rate": 9.265662631938399e-08,
+      "loss": 1.323,
+      "step": 574
+    },
+    {
+      "epoch": 2.904040404040404,
+      "grad_norm": 0.32862385769715274,
+      "learning_rate": 8.363100985026406e-08,
+      "loss": 1.3899,
+      "step": 575
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.31778651266022473,
+      "learning_rate": 7.506656741743001e-08,
+      "loss": 1.3727,
+      "step": 576
+    },
+    {
+      "epoch": 2.9141414141414144,
+      "grad_norm": 0.31945025164654717,
+      "learning_rate": 6.696356381041712e-08,
+      "loss": 1.3538,
+      "step": 577
+    },
+    {
+      "epoch": 2.919191919191919,
+      "grad_norm": 0.32448621680923295,
+      "learning_rate": 5.932224955231935e-08,
+      "loss": 1.2681,
+      "step": 578
+    },
+    {
+      "epoch": 2.9242424242424243,
+      "grad_norm": 0.3255830375698755,
+      "learning_rate": 5.214286089203546e-08,
+      "loss": 1.2082,
+      "step": 579
+    },
+    {
+      "epoch": 2.929292929292929,
+      "grad_norm": 0.3230864932191129,
+      "learning_rate": 4.542561979696491e-08,
+      "loss": 1.3078,
+      "step": 580
+    },
+    {
+      "epoch": 2.9343434343434343,
+      "grad_norm": 0.32044440900847904,
+      "learning_rate": 3.917073394614667e-08,
+      "loss": 1.2221,
+      "step": 581
+    },
+    {
+      "epoch": 2.9393939393939394,
+      "grad_norm": 0.33325632897727475,
+      "learning_rate": 3.337839672384102e-08,
+      "loss": 1.4102,
+      "step": 582
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 0.3376927329733297,
+      "learning_rate": 2.8048787213544315e-08,
+      "loss": 1.3118,
+      "step": 583
+    },
+    {
+      "epoch": 2.9494949494949494,
+      "grad_norm": 0.33092086824129646,
+      "learning_rate": 2.3182070192460104e-08,
+      "loss": 1.3261,
+      "step": 584
+    },
+    {
+      "epoch": 2.9545454545454546,
+      "grad_norm": 0.3317204876226306,
+      "learning_rate": 1.8778396126393206e-08,
+      "loss": 1.2283,
+      "step": 585
+    },
+    {
+      "epoch": 2.9595959595959593,
+      "grad_norm": 0.32504734717078215,
+      "learning_rate": 1.483790116511341e-08,
+      "loss": 1.3745,
+      "step": 586
+    },
+    {
+      "epoch": 2.9646464646464645,
+      "grad_norm": 0.3417117505937822,
+      "learning_rate": 1.1360707138130532e-08,
+      "loss": 1.3584,
+      "step": 587
+    },
+    {
+      "epoch": 2.9696969696969697,
+      "grad_norm": 0.3322967272336344,
+      "learning_rate": 8.346921550939079e-09,
+      "loss": 1.3358,
+      "step": 588
+    },
+    {
+      "epoch": 2.974747474747475,
+      "grad_norm": 0.32207157482687426,
+      "learning_rate": 5.796637581689246e-09,
+      "loss": 1.3833,
+      "step": 589
+    },
+    {
+      "epoch": 2.9797979797979797,
+      "grad_norm": 0.3127525864356668,
+      "learning_rate": 3.709934078307553e-09,
+      "loss": 1.3835,
+      "step": 590
+    },
+    {
+      "epoch": 2.984848484848485,
+      "grad_norm": 0.32045284467637863,
+      "learning_rate": 2.0868755560538023e-09,
+      "loss": 1.3423,
+      "step": 591
+    },
+    {
+      "epoch": 2.98989898989899,
+      "grad_norm": 0.3197154903249698,
+      "learning_rate": 9.275121955393262e-10,
+      "loss": 1.2623,
+      "step": 592
+    },
+    {
+      "epoch": 2.994949494949495,
+      "grad_norm": 0.33061919914451243,
+      "learning_rate": 2.318798411599099e-10,
+      "loss": 1.2716,
+      "step": 593
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3257190208394002,
+      "learning_rate": 0.0,
+      "loss": 1.4233,
+      "step": 594
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 594,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5937910249441198e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}