Files
Gumini-1.5B-Base/trainer_state.json

7067 lines
149 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 250,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001,
"grad_norm": 1.578125,
"learning_rate": 0.0,
"loss": 2.5739,
"step": 1
},
{
"epoch": 0.002,
"grad_norm": 1.546875,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.5765,
"step": 2
},
{
"epoch": 0.003,
"grad_norm": 1.5859375,
"learning_rate": 4.000000000000001e-06,
"loss": 2.5582,
"step": 3
},
{
"epoch": 0.004,
"grad_norm": 1.421875,
"learning_rate": 6e-06,
"loss": 2.6072,
"step": 4
},
{
"epoch": 0.005,
"grad_norm": 1.28125,
"learning_rate": 8.000000000000001e-06,
"loss": 2.6148,
"step": 5
},
{
"epoch": 0.006,
"grad_norm": 0.90625,
"learning_rate": 1e-05,
"loss": 2.5762,
"step": 6
},
{
"epoch": 0.007,
"grad_norm": 0.72265625,
"learning_rate": 1.2e-05,
"loss": 2.5616,
"step": 7
},
{
"epoch": 0.008,
"grad_norm": 0.828125,
"learning_rate": 1.4000000000000001e-05,
"loss": 2.5081,
"step": 8
},
{
"epoch": 0.009,
"grad_norm": 1.046875,
"learning_rate": 1.6000000000000003e-05,
"loss": 2.5662,
"step": 9
},
{
"epoch": 0.01,
"grad_norm": 1.171875,
"learning_rate": 1.8e-05,
"loss": 2.5316,
"step": 10
},
{
"epoch": 0.011,
"grad_norm": 1.0078125,
"learning_rate": 2e-05,
"loss": 2.4746,
"step": 11
},
{
"epoch": 0.012,
"grad_norm": 0.703125,
"learning_rate": 2.2000000000000003e-05,
"loss": 2.5909,
"step": 12
},
{
"epoch": 0.013,
"grad_norm": 0.52734375,
"learning_rate": 2.4e-05,
"loss": 2.5583,
"step": 13
},
{
"epoch": 0.014,
"grad_norm": 0.62109375,
"learning_rate": 2.6000000000000002e-05,
"loss": 2.5573,
"step": 14
},
{
"epoch": 0.015,
"grad_norm": 0.78125,
"learning_rate": 2.8000000000000003e-05,
"loss": 2.5244,
"step": 15
},
{
"epoch": 0.016,
"grad_norm": 0.7890625,
"learning_rate": 3e-05,
"loss": 2.5049,
"step": 16
},
{
"epoch": 0.017,
"grad_norm": 0.69140625,
"learning_rate": 3.2000000000000005e-05,
"loss": 2.4377,
"step": 17
},
{
"epoch": 0.018,
"grad_norm": 0.5703125,
"learning_rate": 3.4000000000000007e-05,
"loss": 2.4077,
"step": 18
},
{
"epoch": 0.019,
"grad_norm": 0.484375,
"learning_rate": 3.6e-05,
"loss": 2.4239,
"step": 19
},
{
"epoch": 0.02,
"grad_norm": 0.62109375,
"learning_rate": 3.8e-05,
"loss": 2.4174,
"step": 20
},
{
"epoch": 0.021,
"grad_norm": 0.79296875,
"learning_rate": 4e-05,
"loss": 2.3856,
"step": 21
},
{
"epoch": 0.022,
"grad_norm": 0.66796875,
"learning_rate": 4.2e-05,
"loss": 2.3144,
"step": 22
},
{
"epoch": 0.023,
"grad_norm": 0.470703125,
"learning_rate": 4.4000000000000006e-05,
"loss": 2.3465,
"step": 23
},
{
"epoch": 0.024,
"grad_norm": 0.48046875,
"learning_rate": 4.600000000000001e-05,
"loss": 2.4592,
"step": 24
},
{
"epoch": 0.025,
"grad_norm": 0.59765625,
"learning_rate": 4.8e-05,
"loss": 2.4452,
"step": 25
},
{
"epoch": 0.026,
"grad_norm": 0.5546875,
"learning_rate": 5e-05,
"loss": 2.4379,
"step": 26
},
{
"epoch": 0.027,
"grad_norm": 0.482421875,
"learning_rate": 5.2000000000000004e-05,
"loss": 2.4406,
"step": 27
},
{
"epoch": 0.028,
"grad_norm": 0.443359375,
"learning_rate": 5.4000000000000005e-05,
"loss": 2.3942,
"step": 28
},
{
"epoch": 0.029,
"grad_norm": 0.4765625,
"learning_rate": 5.6000000000000006e-05,
"loss": 2.5056,
"step": 29
},
{
"epoch": 0.03,
"grad_norm": 0.5625,
"learning_rate": 5.8e-05,
"loss": 2.4823,
"step": 30
},
{
"epoch": 0.031,
"grad_norm": 0.490234375,
"learning_rate": 6e-05,
"loss": 2.4123,
"step": 31
},
{
"epoch": 0.032,
"grad_norm": 0.43359375,
"learning_rate": 6.2e-05,
"loss": 2.5221,
"step": 32
},
{
"epoch": 0.033,
"grad_norm": 0.48828125,
"learning_rate": 6.400000000000001e-05,
"loss": 2.4099,
"step": 33
},
{
"epoch": 0.034,
"grad_norm": 0.50390625,
"learning_rate": 6.6e-05,
"loss": 2.427,
"step": 34
},
{
"epoch": 0.035,
"grad_norm": 0.4453125,
"learning_rate": 6.800000000000001e-05,
"loss": 2.4107,
"step": 35
},
{
"epoch": 0.036,
"grad_norm": 0.44921875,
"learning_rate": 7e-05,
"loss": 2.5013,
"step": 36
},
{
"epoch": 0.037,
"grad_norm": 0.50390625,
"learning_rate": 7.2e-05,
"loss": 2.4065,
"step": 37
},
{
"epoch": 0.038,
"grad_norm": 0.4375,
"learning_rate": 7.4e-05,
"loss": 2.4337,
"step": 38
},
{
"epoch": 0.039,
"grad_norm": 0.4453125,
"learning_rate": 7.6e-05,
"loss": 2.4449,
"step": 39
},
{
"epoch": 0.04,
"grad_norm": 0.51171875,
"learning_rate": 7.800000000000001e-05,
"loss": 2.4418,
"step": 40
},
{
"epoch": 0.041,
"grad_norm": 0.474609375,
"learning_rate": 8e-05,
"loss": 2.3675,
"step": 41
},
{
"epoch": 0.042,
"grad_norm": 0.5625,
"learning_rate": 8.2e-05,
"loss": 2.4663,
"step": 42
},
{
"epoch": 0.043,
"grad_norm": 0.5390625,
"learning_rate": 8.4e-05,
"loss": 2.4475,
"step": 43
},
{
"epoch": 0.044,
"grad_norm": 0.48828125,
"learning_rate": 8.6e-05,
"loss": 2.4135,
"step": 44
},
{
"epoch": 0.045,
"grad_norm": 0.44140625,
"learning_rate": 8.800000000000001e-05,
"loss": 2.4406,
"step": 45
},
{
"epoch": 0.046,
"grad_norm": 0.78125,
"learning_rate": 9e-05,
"loss": 2.3966,
"step": 46
},
{
"epoch": 0.047,
"grad_norm": 0.5234375,
"learning_rate": 9.200000000000001e-05,
"loss": 2.4688,
"step": 47
},
{
"epoch": 0.048,
"grad_norm": 0.462890625,
"learning_rate": 9.4e-05,
"loss": 2.3773,
"step": 48
},
{
"epoch": 0.049,
"grad_norm": 0.51953125,
"learning_rate": 9.6e-05,
"loss": 2.4292,
"step": 49
},
{
"epoch": 0.05,
"grad_norm": 0.474609375,
"learning_rate": 9.8e-05,
"loss": 2.3661,
"step": 50
},
{
"epoch": 0.051,
"grad_norm": 0.50390625,
"learning_rate": 0.0001,
"loss": 2.4172,
"step": 51
},
{
"epoch": 0.052,
"grad_norm": 0.5,
"learning_rate": 0.00010200000000000001,
"loss": 2.3953,
"step": 52
},
{
"epoch": 0.053,
"grad_norm": 0.490234375,
"learning_rate": 0.00010400000000000001,
"loss": 2.3674,
"step": 53
},
{
"epoch": 0.054,
"grad_norm": 0.5390625,
"learning_rate": 0.00010600000000000002,
"loss": 2.4033,
"step": 54
},
{
"epoch": 0.055,
"grad_norm": 0.50390625,
"learning_rate": 0.00010800000000000001,
"loss": 2.3686,
"step": 55
},
{
"epoch": 0.056,
"grad_norm": 0.55078125,
"learning_rate": 0.00011000000000000002,
"loss": 2.353,
"step": 56
},
{
"epoch": 0.057,
"grad_norm": 0.55078125,
"learning_rate": 0.00011200000000000001,
"loss": 2.356,
"step": 57
},
{
"epoch": 0.058,
"grad_norm": 0.55859375,
"learning_rate": 0.00011399999999999999,
"loss": 2.3472,
"step": 58
},
{
"epoch": 0.059,
"grad_norm": 0.61328125,
"learning_rate": 0.000116,
"loss": 2.3558,
"step": 59
},
{
"epoch": 0.06,
"grad_norm": 0.66015625,
"learning_rate": 0.000118,
"loss": 2.3766,
"step": 60
},
{
"epoch": 0.061,
"grad_norm": 0.55859375,
"learning_rate": 0.00012,
"loss": 2.2852,
"step": 61
},
{
"epoch": 0.062,
"grad_norm": 0.62109375,
"learning_rate": 0.000122,
"loss": 2.3451,
"step": 62
},
{
"epoch": 0.063,
"grad_norm": 0.87890625,
"learning_rate": 0.000124,
"loss": 2.4868,
"step": 63
},
{
"epoch": 0.064,
"grad_norm": 0.640625,
"learning_rate": 0.000126,
"loss": 2.4392,
"step": 64
},
{
"epoch": 0.065,
"grad_norm": 0.6640625,
"learning_rate": 0.00012800000000000002,
"loss": 2.4329,
"step": 65
},
{
"epoch": 0.066,
"grad_norm": 0.6328125,
"learning_rate": 0.00013000000000000002,
"loss": 2.383,
"step": 66
},
{
"epoch": 0.067,
"grad_norm": 0.65234375,
"learning_rate": 0.000132,
"loss": 2.3924,
"step": 67
},
{
"epoch": 0.068,
"grad_norm": 0.65234375,
"learning_rate": 0.000134,
"loss": 2.44,
"step": 68
},
{
"epoch": 0.069,
"grad_norm": 0.62109375,
"learning_rate": 0.00013600000000000003,
"loss": 2.4202,
"step": 69
},
{
"epoch": 0.07,
"grad_norm": 0.55078125,
"learning_rate": 0.000138,
"loss": 2.3698,
"step": 70
},
{
"epoch": 0.071,
"grad_norm": 0.640625,
"learning_rate": 0.00014,
"loss": 2.3578,
"step": 71
},
{
"epoch": 0.072,
"grad_norm": 0.57421875,
"learning_rate": 0.000142,
"loss": 2.3932,
"step": 72
},
{
"epoch": 0.073,
"grad_norm": 0.5703125,
"learning_rate": 0.000144,
"loss": 2.4188,
"step": 73
},
{
"epoch": 0.074,
"grad_norm": 0.578125,
"learning_rate": 0.000146,
"loss": 2.3661,
"step": 74
},
{
"epoch": 0.075,
"grad_norm": 0.59375,
"learning_rate": 0.000148,
"loss": 2.369,
"step": 75
},
{
"epoch": 0.076,
"grad_norm": 0.64453125,
"learning_rate": 0.00015000000000000001,
"loss": 2.3451,
"step": 76
},
{
"epoch": 0.077,
"grad_norm": 0.671875,
"learning_rate": 0.000152,
"loss": 2.3752,
"step": 77
},
{
"epoch": 0.078,
"grad_norm": 0.671875,
"learning_rate": 0.000154,
"loss": 2.2468,
"step": 78
},
{
"epoch": 0.079,
"grad_norm": 0.70703125,
"learning_rate": 0.00015600000000000002,
"loss": 2.2885,
"step": 79
},
{
"epoch": 0.08,
"grad_norm": 0.69140625,
"learning_rate": 0.00015800000000000002,
"loss": 2.3265,
"step": 80
},
{
"epoch": 0.081,
"grad_norm": 0.70703125,
"learning_rate": 0.00016,
"loss": 2.3271,
"step": 81
},
{
"epoch": 0.082,
"grad_norm": 0.8515625,
"learning_rate": 0.000162,
"loss": 2.3673,
"step": 82
},
{
"epoch": 0.083,
"grad_norm": 1.3671875,
"learning_rate": 0.000164,
"loss": 2.3233,
"step": 83
},
{
"epoch": 0.084,
"grad_norm": 1.6171875,
"learning_rate": 0.000166,
"loss": 2.3503,
"step": 84
},
{
"epoch": 0.085,
"grad_norm": 0.85546875,
"learning_rate": 0.000168,
"loss": 2.3311,
"step": 85
},
{
"epoch": 0.086,
"grad_norm": 5.21875,
"learning_rate": 0.00017,
"loss": 2.4252,
"step": 86
},
{
"epoch": 0.087,
"grad_norm": 2.6875,
"learning_rate": 0.000172,
"loss": 2.4418,
"step": 87
},
{
"epoch": 0.088,
"grad_norm": 2.859375,
"learning_rate": 0.000174,
"loss": 2.4377,
"step": 88
},
{
"epoch": 0.089,
"grad_norm": 2.578125,
"learning_rate": 0.00017600000000000002,
"loss": 2.5101,
"step": 89
},
{
"epoch": 0.09,
"grad_norm": 1.2109375,
"learning_rate": 0.00017800000000000002,
"loss": 2.3661,
"step": 90
},
{
"epoch": 0.091,
"grad_norm": 3.03125,
"learning_rate": 0.00018,
"loss": 2.407,
"step": 91
},
{
"epoch": 0.092,
"grad_norm": 2.25,
"learning_rate": 0.000182,
"loss": 2.4116,
"step": 92
},
{
"epoch": 0.093,
"grad_norm": 2.1875,
"learning_rate": 0.00018400000000000003,
"loss": 2.4503,
"step": 93
},
{
"epoch": 0.094,
"grad_norm": 2.0625,
"learning_rate": 0.00018600000000000002,
"loss": 2.4242,
"step": 94
},
{
"epoch": 0.095,
"grad_norm": 1.1328125,
"learning_rate": 0.000188,
"loss": 2.404,
"step": 95
},
{
"epoch": 0.096,
"grad_norm": 2.140625,
"learning_rate": 0.00019,
"loss": 2.3241,
"step": 96
},
{
"epoch": 0.097,
"grad_norm": 1.4296875,
"learning_rate": 0.000192,
"loss": 2.3301,
"step": 97
},
{
"epoch": 0.098,
"grad_norm": 1.5,
"learning_rate": 0.000194,
"loss": 2.4137,
"step": 98
},
{
"epoch": 0.099,
"grad_norm": 1.8671875,
"learning_rate": 0.000196,
"loss": 2.3244,
"step": 99
},
{
"epoch": 0.1,
"grad_norm": 1.15625,
"learning_rate": 0.00019800000000000002,
"loss": 2.3515,
"step": 100
},
{
"epoch": 0.101,
"grad_norm": 1.421875,
"learning_rate": 0.0002,
"loss": 2.2329,
"step": 101
},
{
"epoch": 0.102,
"grad_norm": 1.140625,
"learning_rate": 0.00019999939076577905,
"loss": 2.3435,
"step": 102
},
{
"epoch": 0.103,
"grad_norm": 1.265625,
"learning_rate": 0.00019999756307053948,
"loss": 2.2787,
"step": 103
},
{
"epoch": 0.104,
"grad_norm": 1.09375,
"learning_rate": 0.00019999451693655123,
"loss": 2.3508,
"step": 104
},
{
"epoch": 0.105,
"grad_norm": 1.234375,
"learning_rate": 0.00019999025240093044,
"loss": 2.3212,
"step": 105
},
{
"epoch": 0.106,
"grad_norm": 1.15625,
"learning_rate": 0.00019998476951563915,
"loss": 2.2951,
"step": 106
},
{
"epoch": 0.107,
"grad_norm": 0.98828125,
"learning_rate": 0.00019997806834748456,
"loss": 2.3228,
"step": 107
},
{
"epoch": 0.108,
"grad_norm": 1.046875,
"learning_rate": 0.00019997014897811833,
"loss": 2.3505,
"step": 108
},
{
"epoch": 0.109,
"grad_norm": 1.015625,
"learning_rate": 0.00019996101150403543,
"loss": 2.2858,
"step": 109
},
{
"epoch": 0.11,
"grad_norm": 1.0,
"learning_rate": 0.00019995065603657316,
"loss": 2.3762,
"step": 110
},
{
"epoch": 0.111,
"grad_norm": 0.94921875,
"learning_rate": 0.0001999390827019096,
"loss": 2.3848,
"step": 111
},
{
"epoch": 0.112,
"grad_norm": 0.8671875,
"learning_rate": 0.0001999262916410621,
"loss": 2.2709,
"step": 112
},
{
"epoch": 0.113,
"grad_norm": 0.85546875,
"learning_rate": 0.00019991228300988585,
"loss": 2.3362,
"step": 113
},
{
"epoch": 0.114,
"grad_norm": 0.85546875,
"learning_rate": 0.00019989705697907149,
"loss": 2.3292,
"step": 114
},
{
"epoch": 0.115,
"grad_norm": 0.953125,
"learning_rate": 0.0001998806137341434,
"loss": 2.1826,
"step": 115
},
{
"epoch": 0.116,
"grad_norm": 0.73828125,
"learning_rate": 0.0001998629534754574,
"loss": 2.2754,
"step": 116
},
{
"epoch": 0.117,
"grad_norm": 0.921875,
"learning_rate": 0.00019984407641819812,
"loss": 2.3246,
"step": 117
},
{
"epoch": 0.118,
"grad_norm": 0.83203125,
"learning_rate": 0.00019982398279237655,
"loss": 2.3062,
"step": 118
},
{
"epoch": 0.119,
"grad_norm": 1.046875,
"learning_rate": 0.00019980267284282717,
"loss": 2.276,
"step": 119
},
{
"epoch": 0.12,
"grad_norm": 0.734375,
"learning_rate": 0.000199780146829205,
"loss": 2.3193,
"step": 120
},
{
"epoch": 0.121,
"grad_norm": 0.8828125,
"learning_rate": 0.00019975640502598244,
"loss": 2.3254,
"step": 121
},
{
"epoch": 0.122,
"grad_norm": 0.74609375,
"learning_rate": 0.00019973144772244582,
"loss": 2.2268,
"step": 122
},
{
"epoch": 0.123,
"grad_norm": 0.78125,
"learning_rate": 0.00019970527522269205,
"loss": 2.3131,
"step": 123
},
{
"epoch": 0.124,
"grad_norm": 0.79296875,
"learning_rate": 0.00019967788784562473,
"loss": 2.2756,
"step": 124
},
{
"epoch": 0.125,
"grad_norm": 1.4296875,
"learning_rate": 0.00019964928592495045,
"loss": 2.2906,
"step": 125
},
{
"epoch": 0.126,
"grad_norm": 0.86328125,
"learning_rate": 0.00019961946980917456,
"loss": 2.2905,
"step": 126
},
{
"epoch": 0.127,
"grad_norm": 0.78515625,
"learning_rate": 0.00019958843986159704,
"loss": 2.3722,
"step": 127
},
{
"epoch": 0.128,
"grad_norm": 0.9296875,
"learning_rate": 0.00019955619646030802,
"loss": 2.3203,
"step": 128
},
{
"epoch": 0.129,
"grad_norm": 0.84765625,
"learning_rate": 0.0001995227399981831,
"loss": 2.3034,
"step": 129
},
{
"epoch": 0.13,
"grad_norm": 0.75390625,
"learning_rate": 0.00019948807088287883,
"loss": 2.2203,
"step": 130
},
{
"epoch": 0.131,
"grad_norm": 0.9609375,
"learning_rate": 0.00019945218953682734,
"loss": 2.2994,
"step": 131
},
{
"epoch": 0.132,
"grad_norm": 0.79296875,
"learning_rate": 0.00019941509639723155,
"loss": 2.291,
"step": 132
},
{
"epoch": 0.133,
"grad_norm": 0.9921875,
"learning_rate": 0.00019937679191605963,
"loss": 2.2753,
"step": 133
},
{
"epoch": 0.134,
"grad_norm": 1.09375,
"learning_rate": 0.00019933727656003963,
"loss": 2.3568,
"step": 134
},
{
"epoch": 0.135,
"grad_norm": 0.84375,
"learning_rate": 0.0001992965508106537,
"loss": 2.3729,
"step": 135
},
{
"epoch": 0.136,
"grad_norm": 1.046875,
"learning_rate": 0.00019925461516413223,
"loss": 2.2552,
"step": 136
},
{
"epoch": 0.137,
"grad_norm": 0.86328125,
"learning_rate": 0.0001992114701314478,
"loss": 2.2658,
"step": 137
},
{
"epoch": 0.138,
"grad_norm": 0.92578125,
"learning_rate": 0.00019916711623830903,
"loss": 2.2669,
"step": 138
},
{
"epoch": 0.139,
"grad_norm": 0.984375,
"learning_rate": 0.00019912155402515417,
"loss": 2.2735,
"step": 139
},
{
"epoch": 0.14,
"grad_norm": 0.8203125,
"learning_rate": 0.00019907478404714436,
"loss": 2.2985,
"step": 140
},
{
"epoch": 0.141,
"grad_norm": 0.7890625,
"learning_rate": 0.00019902680687415705,
"loss": 2.2667,
"step": 141
},
{
"epoch": 0.142,
"grad_norm": 0.88671875,
"learning_rate": 0.0001989776230907789,
"loss": 2.2627,
"step": 142
},
{
"epoch": 0.143,
"grad_norm": 0.7734375,
"learning_rate": 0.00019892723329629887,
"loss": 2.2876,
"step": 143
},
{
"epoch": 0.144,
"grad_norm": 0.76953125,
"learning_rate": 0.0001988756381047006,
"loss": 2.2904,
"step": 144
},
{
"epoch": 0.145,
"grad_norm": 0.89453125,
"learning_rate": 0.0001988228381446553,
"loss": 2.2903,
"step": 145
},
{
"epoch": 0.146,
"grad_norm": 1.1484375,
"learning_rate": 0.00019876883405951377,
"loss": 2.2866,
"step": 146
},
{
"epoch": 0.147,
"grad_norm": 0.84375,
"learning_rate": 0.0001987136265072988,
"loss": 2.2743,
"step": 147
},
{
"epoch": 0.148,
"grad_norm": 1.046875,
"learning_rate": 0.00019865721616069696,
"loss": 2.2958,
"step": 148
},
{
"epoch": 0.149,
"grad_norm": 1.0390625,
"learning_rate": 0.0001985996037070505,
"loss": 2.2877,
"step": 149
},
{
"epoch": 0.15,
"grad_norm": 0.7890625,
"learning_rate": 0.00019854078984834903,
"loss": 2.3079,
"step": 150
},
{
"epoch": 0.151,
"grad_norm": 1.0625,
"learning_rate": 0.00019848077530122083,
"loss": 2.3422,
"step": 151
},
{
"epoch": 0.152,
"grad_norm": 1.0078125,
"learning_rate": 0.0001984195607969242,
"loss": 2.2979,
"step": 152
},
{
"epoch": 0.153,
"grad_norm": 0.76953125,
"learning_rate": 0.00019835714708133862,
"loss": 2.327,
"step": 153
},
{
"epoch": 0.154,
"grad_norm": 0.93359375,
"learning_rate": 0.00019829353491495545,
"loss": 2.2707,
"step": 154
},
{
"epoch": 0.155,
"grad_norm": 0.78515625,
"learning_rate": 0.0001982287250728689,
"loss": 2.2669,
"step": 155
},
{
"epoch": 0.156,
"grad_norm": 0.73046875,
"learning_rate": 0.00019816271834476642,
"loss": 2.2707,
"step": 156
},
{
"epoch": 0.157,
"grad_norm": 0.890625,
"learning_rate": 0.00019809551553491916,
"loss": 2.3557,
"step": 157
},
{
"epoch": 0.158,
"grad_norm": 0.83984375,
"learning_rate": 0.00019802711746217218,
"loss": 2.3086,
"step": 158
},
{
"epoch": 0.159,
"grad_norm": 0.828125,
"learning_rate": 0.0001979575249599344,
"loss": 2.2986,
"step": 159
},
{
"epoch": 0.16,
"grad_norm": 1.046875,
"learning_rate": 0.0001978867388761685,
"loss": 2.2621,
"step": 160
},
{
"epoch": 0.161,
"grad_norm": 0.91796875,
"learning_rate": 0.00019781476007338058,
"loss": 2.4438,
"step": 161
},
{
"epoch": 0.162,
"grad_norm": 1.203125,
"learning_rate": 0.0001977415894286096,
"loss": 2.3691,
"step": 162
},
{
"epoch": 0.163,
"grad_norm": 1.0,
"learning_rate": 0.0001976672278334168,
"loss": 2.2932,
"step": 163
},
{
"epoch": 0.164,
"grad_norm": 0.9765625,
"learning_rate": 0.00019759167619387476,
"loss": 2.2185,
"step": 164
},
{
"epoch": 0.165,
"grad_norm": 0.87890625,
"learning_rate": 0.00019751493543055632,
"loss": 2.3593,
"step": 165
},
{
"epoch": 0.166,
"grad_norm": 0.98828125,
"learning_rate": 0.00019743700647852354,
"loss": 2.2809,
"step": 166
},
{
"epoch": 0.167,
"grad_norm": 1.1015625,
"learning_rate": 0.00019735789028731604,
"loss": 2.2389,
"step": 167
},
{
"epoch": 0.168,
"grad_norm": 0.88671875,
"learning_rate": 0.00019727758782093967,
"loss": 2.2574,
"step": 168
},
{
"epoch": 0.169,
"grad_norm": 0.91796875,
"learning_rate": 0.00019719610005785465,
"loss": 2.2934,
"step": 169
},
{
"epoch": 0.17,
"grad_norm": 0.85546875,
"learning_rate": 0.00019711342799096361,
"loss": 2.3262,
"step": 170
},
{
"epoch": 0.171,
"grad_norm": 0.765625,
"learning_rate": 0.00019702957262759965,
"loss": 2.3175,
"step": 171
},
{
"epoch": 0.172,
"grad_norm": 0.75,
"learning_rate": 0.0001969445349895139,
"loss": 2.3418,
"step": 172
},
{
"epoch": 0.173,
"grad_norm": 0.85546875,
"learning_rate": 0.0001968583161128631,
"loss": 2.3432,
"step": 173
},
{
"epoch": 0.174,
"grad_norm": 0.78515625,
"learning_rate": 0.00019677091704819715,
"loss": 2.2738,
"step": 174
},
{
"epoch": 0.175,
"grad_norm": 0.76171875,
"learning_rate": 0.00019668233886044597,
"loss": 2.253,
"step": 175
},
{
"epoch": 0.176,
"grad_norm": 0.8671875,
"learning_rate": 0.00019659258262890683,
"loss": 2.2798,
"step": 176
},
{
"epoch": 0.177,
"grad_norm": 0.77734375,
"learning_rate": 0.00019650164944723115,
"loss": 2.3464,
"step": 177
},
{
"epoch": 0.178,
"grad_norm": 0.83203125,
"learning_rate": 0.00019640954042341103,
"loss": 2.2911,
"step": 178
},
{
"epoch": 0.179,
"grad_norm": 0.80859375,
"learning_rate": 0.00019631625667976583,
"loss": 2.2957,
"step": 179
},
{
"epoch": 0.18,
"grad_norm": 0.7734375,
"learning_rate": 0.00019622179935292855,
"loss": 2.3766,
"step": 180
},
{
"epoch": 0.181,
"grad_norm": 0.796875,
"learning_rate": 0.0001961261695938319,
"loss": 2.2897,
"step": 181
},
{
"epoch": 0.182,
"grad_norm": 0.80078125,
"learning_rate": 0.0001960293685676943,
"loss": 2.2706,
"step": 182
},
{
"epoch": 0.183,
"grad_norm": 0.8125,
"learning_rate": 0.00019593139745400576,
"loss": 2.328,
"step": 183
},
{
"epoch": 0.184,
"grad_norm": 1.3125,
"learning_rate": 0.00019583225744651333,
"loss": 2.4127,
"step": 184
},
{
"epoch": 0.185,
"grad_norm": 0.87890625,
"learning_rate": 0.00019573194975320673,
"loss": 2.3852,
"step": 185
},
{
"epoch": 0.186,
"grad_norm": 0.87109375,
"learning_rate": 0.00019563047559630357,
"loss": 2.3569,
"step": 186
},
{
"epoch": 0.187,
"grad_norm": 0.7578125,
"learning_rate": 0.00019552783621223436,
"loss": 2.3375,
"step": 187
},
{
"epoch": 0.188,
"grad_norm": 1.109375,
"learning_rate": 0.0001954240328516277,
"loss": 2.3259,
"step": 188
},
{
"epoch": 0.189,
"grad_norm": 0.87890625,
"learning_rate": 0.0001953190667792947,
"loss": 2.36,
"step": 189
},
{
"epoch": 0.19,
"grad_norm": 0.9921875,
"learning_rate": 0.00019521293927421388,
"loss": 2.2863,
"step": 190
},
{
"epoch": 0.191,
"grad_norm": 0.7265625,
"learning_rate": 0.00019510565162951537,
"loss": 2.3261,
"step": 191
},
{
"epoch": 0.192,
"grad_norm": 0.890625,
"learning_rate": 0.00019499720515246525,
"loss": 2.2758,
"step": 192
},
{
"epoch": 0.193,
"grad_norm": 0.7578125,
"learning_rate": 0.00019488760116444966,
"loss": 2.3762,
"step": 193
},
{
"epoch": 0.194,
"grad_norm": 0.83203125,
"learning_rate": 0.0001947768410009586,
"loss": 2.2936,
"step": 194
},
{
"epoch": 0.195,
"grad_norm": 0.73046875,
"learning_rate": 0.00019466492601156966,
"loss": 2.2921,
"step": 195
},
{
"epoch": 0.196,
"grad_norm": 0.79296875,
"learning_rate": 0.0001945518575599317,
"loss": 2.3079,
"step": 196
},
{
"epoch": 0.197,
"grad_norm": 0.7890625,
"learning_rate": 0.00019443763702374812,
"loss": 2.2715,
"step": 197
},
{
"epoch": 0.198,
"grad_norm": 0.73828125,
"learning_rate": 0.0001943222657947601,
"loss": 2.2216,
"step": 198
},
{
"epoch": 0.199,
"grad_norm": 0.8828125,
"learning_rate": 0.00019420574527872968,
"loss": 2.2707,
"step": 199
},
{
"epoch": 0.2,
"grad_norm": 0.78125,
"learning_rate": 0.00019408807689542257,
"loss": 2.3291,
"step": 200
},
{
"epoch": 0.201,
"grad_norm": 0.74609375,
"learning_rate": 0.00019396926207859084,
"loss": 2.2666,
"step": 201
},
{
"epoch": 0.202,
"grad_norm": 0.67578125,
"learning_rate": 0.0001938493022759556,
"loss": 2.3023,
"step": 202
},
{
"epoch": 0.203,
"grad_norm": 0.7890625,
"learning_rate": 0.00019372819894918915,
"loss": 2.2596,
"step": 203
},
{
"epoch": 0.204,
"grad_norm": 0.83203125,
"learning_rate": 0.00019360595357389735,
"loss": 2.4557,
"step": 204
},
{
"epoch": 0.205,
"grad_norm": 0.7578125,
"learning_rate": 0.00019348256763960145,
"loss": 2.3458,
"step": 205
},
{
"epoch": 0.206,
"grad_norm": 0.73046875,
"learning_rate": 0.00019335804264972018,
"loss": 2.2998,
"step": 206
},
{
"epoch": 0.207,
"grad_norm": 0.75,
"learning_rate": 0.00019323238012155123,
"loss": 2.2583,
"step": 207
},
{
"epoch": 0.208,
"grad_norm": 0.7265625,
"learning_rate": 0.00019310558158625285,
"loss": 2.2788,
"step": 208
},
{
"epoch": 0.209,
"grad_norm": 105.5,
"learning_rate": 0.00019297764858882514,
"loss": 2.2118,
"step": 209
},
{
"epoch": 0.21,
"grad_norm": 1.1875,
"learning_rate": 0.00019284858268809137,
"loss": 2.3434,
"step": 210
},
{
"epoch": 0.211,
"grad_norm": 0.7578125,
"learning_rate": 0.00019271838545667876,
"loss": 2.277,
"step": 211
},
{
"epoch": 0.212,
"grad_norm": 0.90625,
"learning_rate": 0.0001925870584809995,
"loss": 2.2519,
"step": 212
},
{
"epoch": 0.213,
"grad_norm": 0.828125,
"learning_rate": 0.00019245460336123134,
"loss": 2.3388,
"step": 213
},
{
"epoch": 0.214,
"grad_norm": 0.7890625,
"learning_rate": 0.00019232102171129811,
"loss": 2.2502,
"step": 214
},
{
"epoch": 0.215,
"grad_norm": 50.5,
"learning_rate": 0.00019218631515885006,
"loss": 2.3237,
"step": 215
},
{
"epoch": 0.216,
"grad_norm": 1.21875,
"learning_rate": 0.00019205048534524406,
"loss": 2.2732,
"step": 216
},
{
"epoch": 0.217,
"grad_norm": 0.8515625,
"learning_rate": 0.00019191353392552344,
"loss": 2.3619,
"step": 217
},
{
"epoch": 0.218,
"grad_norm": 1.03125,
"learning_rate": 0.00019177546256839812,
"loss": 2.349,
"step": 218
},
{
"epoch": 0.219,
"grad_norm": 0.8203125,
"learning_rate": 0.00019163627295622397,
"loss": 2.3792,
"step": 219
},
{
"epoch": 0.22,
"grad_norm": 0.765625,
"learning_rate": 0.0001914959667849825,
"loss": 2.2966,
"step": 220
},
{
"epoch": 0.221,
"grad_norm": 0.8359375,
"learning_rate": 0.0001913545457642601,
"loss": 2.3772,
"step": 221
},
{
"epoch": 0.222,
"grad_norm": 0.7734375,
"learning_rate": 0.0001912120116172273,
"loss": 2.3286,
"step": 222
},
{
"epoch": 0.223,
"grad_norm": 0.79296875,
"learning_rate": 0.00019106836608061772,
"loss": 2.3807,
"step": 223
},
{
"epoch": 0.224,
"grad_norm": 0.7421875,
"learning_rate": 0.00019092361090470688,
"loss": 2.2384,
"step": 224
},
{
"epoch": 0.225,
"grad_norm": 0.7109375,
"learning_rate": 0.00019077774785329087,
"loss": 2.3319,
"step": 225
},
{
"epoch": 0.226,
"grad_norm": 0.7578125,
"learning_rate": 0.000190630778703665,
"loss": 2.33,
"step": 226
},
{
"epoch": 0.227,
"grad_norm": 0.70703125,
"learning_rate": 0.00019048270524660196,
"loss": 2.319,
"step": 227
},
{
"epoch": 0.228,
"grad_norm": 0.75390625,
"learning_rate": 0.0001903335292863301,
"loss": 2.363,
"step": 228
},
{
"epoch": 0.229,
"grad_norm": 0.6796875,
"learning_rate": 0.0001901832526405114,
"loss": 2.2566,
"step": 229
},
{
"epoch": 0.23,
"grad_norm": 0.80078125,
"learning_rate": 0.00019003187714021938,
"loss": 2.0954,
"step": 230
},
{
"epoch": 0.231,
"grad_norm": 0.69140625,
"learning_rate": 0.0001898794046299167,
"loss": 2.1235,
"step": 231
},
{
"epoch": 0.232,
"grad_norm": 0.68359375,
"learning_rate": 0.00018972583696743285,
"loss": 2.1275,
"step": 232
},
{
"epoch": 0.233,
"grad_norm": 0.71875,
"learning_rate": 0.0001895711760239413,
"loss": 2.3279,
"step": 233
},
{
"epoch": 0.234,
"grad_norm": 0.73046875,
"learning_rate": 0.0001894154236839368,
"loss": 2.2539,
"step": 234
},
{
"epoch": 0.235,
"grad_norm": 0.953125,
"learning_rate": 0.00018925858184521256,
"loss": 2.4256,
"step": 235
},
{
"epoch": 0.236,
"grad_norm": 1.0,
"learning_rate": 0.0001891006524188368,
"loss": 2.3487,
"step": 236
},
{
"epoch": 0.237,
"grad_norm": 0.875,
"learning_rate": 0.00018894163732912977,
"loss": 2.2536,
"step": 237
},
{
"epoch": 0.238,
"grad_norm": 0.9140625,
"learning_rate": 0.00018878153851364013,
"loss": 2.272,
"step": 238
},
{
"epoch": 0.239,
"grad_norm": 0.796875,
"learning_rate": 0.00018862035792312147,
"loss": 2.3092,
"step": 239
},
{
"epoch": 0.24,
"grad_norm": 0.94921875,
"learning_rate": 0.0001884580975215084,
"loss": 2.2811,
"step": 240
},
{
"epoch": 0.241,
"grad_norm": 0.796875,
"learning_rate": 0.00018829475928589271,
"loss": 2.2895,
"step": 241
},
{
"epoch": 0.242,
"grad_norm": 0.84765625,
"learning_rate": 0.0001881303452064992,
"loss": 2.2619,
"step": 242
},
{
"epoch": 0.243,
"grad_norm": 0.81640625,
"learning_rate": 0.00018796485728666165,
"loss": 2.2598,
"step": 243
},
{
"epoch": 0.244,
"grad_norm": 0.87109375,
"learning_rate": 0.00018779829754279805,
"loss": 2.2788,
"step": 244
},
{
"epoch": 0.245,
"grad_norm": 0.9296875,
"learning_rate": 0.00018763066800438636,
"loss": 2.2685,
"step": 245
},
{
"epoch": 0.246,
"grad_norm": 0.94140625,
"learning_rate": 0.00018746197071393958,
"loss": 2.2358,
"step": 246
},
{
"epoch": 0.247,
"grad_norm": 0.85546875,
"learning_rate": 0.00018729220772698097,
"loss": 2.3087,
"step": 247
},
{
"epoch": 0.248,
"grad_norm": 0.96484375,
"learning_rate": 0.00018712138111201895,
"loss": 2.2697,
"step": 248
},
{
"epoch": 0.249,
"grad_norm": 0.859375,
"learning_rate": 0.0001869494929505219,
"loss": 2.2411,
"step": 249
},
{
"epoch": 0.25,
"grad_norm": 1.2109375,
"learning_rate": 0.00018677654533689287,
"loss": 2.3894,
"step": 250
},
{
"epoch": 0.25,
"eval_loss": 2.5482473373413086,
"eval_runtime": 23.7265,
"eval_samples_per_second": 21.073,
"eval_steps_per_second": 1.349,
"step": 250
},
{
"epoch": 0.251,
"grad_norm": 0.83203125,
"learning_rate": 0.00018660254037844388,
"loss": 2.3641,
"step": 251
},
{
"epoch": 0.252,
"grad_norm": 1.1484375,
"learning_rate": 0.0001864274801953705,
"loss": 2.3742,
"step": 252
},
{
"epoch": 0.253,
"grad_norm": 0.7734375,
"learning_rate": 0.00018625136692072575,
"loss": 2.3225,
"step": 253
},
{
"epoch": 0.254,
"grad_norm": 0.97265625,
"learning_rate": 0.0001860742027003944,
"loss": 2.2835,
"step": 254
},
{
"epoch": 0.255,
"grad_norm": 0.76953125,
"learning_rate": 0.00018589598969306645,
"loss": 2.3117,
"step": 255
},
{
"epoch": 0.256,
"grad_norm": 0.91015625,
"learning_rate": 0.00018571673007021123,
"loss": 2.2742,
"step": 256
},
{
"epoch": 0.257,
"grad_norm": 0.828125,
"learning_rate": 0.00018553642601605068,
"loss": 2.2516,
"step": 257
},
{
"epoch": 0.258,
"grad_norm": 0.87890625,
"learning_rate": 0.00018535507972753274,
"loss": 2.2713,
"step": 258
},
{
"epoch": 0.259,
"grad_norm": 0.80078125,
"learning_rate": 0.00018517269341430476,
"loss": 2.3012,
"step": 259
},
{
"epoch": 0.26,
"grad_norm": 0.80078125,
"learning_rate": 0.00018498926929868642,
"loss": 2.2785,
"step": 260
},
{
"epoch": 0.261,
"grad_norm": 0.75390625,
"learning_rate": 0.0001848048096156426,
"loss": 2.3399,
"step": 261
},
{
"epoch": 0.262,
"grad_norm": 0.78125,
"learning_rate": 0.00018461931661275643,
"loss": 2.3489,
"step": 262
},
{
"epoch": 0.263,
"grad_norm": 0.7421875,
"learning_rate": 0.00018443279255020152,
"loss": 2.3518,
"step": 263
},
{
"epoch": 0.264,
"grad_norm": 0.765625,
"learning_rate": 0.00018424523970071477,
"loss": 2.3159,
"step": 264
},
{
"epoch": 0.265,
"grad_norm": 0.859375,
"learning_rate": 0.00018405666034956844,
"loss": 2.3493,
"step": 265
},
{
"epoch": 0.266,
"grad_norm": 0.9453125,
"learning_rate": 0.00018386705679454242,
"loss": 2.4719,
"step": 266
},
{
"epoch": 0.267,
"grad_norm": 1.21875,
"learning_rate": 0.00018367643134589617,
"loss": 2.406,
"step": 267
},
{
"epoch": 0.268,
"grad_norm": 0.96875,
"learning_rate": 0.00018348478632634066,
"loss": 2.3473,
"step": 268
},
{
"epoch": 0.269,
"grad_norm": 0.7578125,
"learning_rate": 0.00018329212407100994,
"loss": 2.3684,
"step": 269
},
{
"epoch": 0.27,
"grad_norm": 0.9765625,
"learning_rate": 0.00018309844692743283,
"loss": 2.4001,
"step": 270
},
{
"epoch": 0.271,
"grad_norm": 1.078125,
"learning_rate": 0.00018290375725550417,
"loss": 2.3757,
"step": 271
},
{
"epoch": 0.272,
"grad_norm": 0.83984375,
"learning_rate": 0.00018270805742745617,
"loss": 2.307,
"step": 272
},
{
"epoch": 0.273,
"grad_norm": 0.828125,
"learning_rate": 0.00018251134982782952,
"loss": 2.3513,
"step": 273
},
{
"epoch": 0.274,
"grad_norm": 0.7109375,
"learning_rate": 0.0001823136368534442,
"loss": 2.2733,
"step": 274
},
{
"epoch": 0.275,
"grad_norm": 0.74609375,
"learning_rate": 0.00018211492091337042,
"loss": 2.3144,
"step": 275
},
{
"epoch": 0.276,
"grad_norm": 0.7109375,
"learning_rate": 0.0001819152044288992,
"loss": 2.348,
"step": 276
},
{
"epoch": 0.277,
"grad_norm": 0.64453125,
"learning_rate": 0.00018171448983351284,
"loss": 2.2922,
"step": 277
},
{
"epoch": 0.278,
"grad_norm": 0.66015625,
"learning_rate": 0.00018151277957285543,
"loss": 2.321,
"step": 278
},
{
"epoch": 0.279,
"grad_norm": 0.6640625,
"learning_rate": 0.00018131007610470276,
"loss": 2.2157,
"step": 279
},
{
"epoch": 0.28,
"grad_norm": 0.65234375,
"learning_rate": 0.00018110638189893267,
"loss": 2.2543,
"step": 280
},
{
"epoch": 0.281,
"grad_norm": 0.67578125,
"learning_rate": 0.00018090169943749476,
"loss": 2.2355,
"step": 281
},
{
"epoch": 0.282,
"grad_norm": 0.95703125,
"learning_rate": 0.00018069603121438022,
"loss": 2.4974,
"step": 282
},
{
"epoch": 0.283,
"grad_norm": 0.77734375,
"learning_rate": 0.0001804893797355914,
"loss": 2.3726,
"step": 283
},
{
"epoch": 0.284,
"grad_norm": 0.703125,
"learning_rate": 0.00018028174751911146,
"loss": 2.2914,
"step": 284
},
{
"epoch": 0.285,
"grad_norm": 0.625,
"learning_rate": 0.00018007313709487334,
"loss": 2.3303,
"step": 285
},
{
"epoch": 0.286,
"grad_norm": 0.734375,
"learning_rate": 0.00017986355100472928,
"loss": 2.3194,
"step": 286
},
{
"epoch": 0.287,
"grad_norm": 0.71484375,
"learning_rate": 0.00017965299180241963,
"loss": 2.294,
"step": 287
},
{
"epoch": 0.288,
"grad_norm": 0.60546875,
"learning_rate": 0.00017944146205354182,
"loss": 2.3698,
"step": 288
},
{
"epoch": 0.289,
"grad_norm": 0.7578125,
"learning_rate": 0.00017922896433551907,
"loss": 2.1954,
"step": 289
},
{
"epoch": 0.29,
"grad_norm": 0.66796875,
"learning_rate": 0.00017901550123756906,
"loss": 2.2302,
"step": 290
},
{
"epoch": 0.291,
"grad_norm": 0.921875,
"learning_rate": 0.00017880107536067218,
"loss": 2.3003,
"step": 291
},
{
"epoch": 0.292,
"grad_norm": 5.78125,
"learning_rate": 0.0001785856893175402,
"loss": 2.2813,
"step": 292
},
{
"epoch": 0.293,
"grad_norm": 1.0078125,
"learning_rate": 0.000178369345732584,
"loss": 2.2799,
"step": 293
},
{
"epoch": 0.294,
"grad_norm": 0.890625,
"learning_rate": 0.00017815204724188187,
"loss": 2.3409,
"step": 294
},
{
"epoch": 0.295,
"grad_norm": 0.7265625,
"learning_rate": 0.00017793379649314744,
"loss": 2.326,
"step": 295
},
{
"epoch": 0.296,
"grad_norm": 0.8828125,
"learning_rate": 0.0001777145961456971,
"loss": 2.3563,
"step": 296
},
{
"epoch": 0.297,
"grad_norm": 0.984375,
"learning_rate": 0.00017749444887041799,
"loss": 2.4839,
"step": 297
},
{
"epoch": 0.298,
"grad_norm": 0.8828125,
"learning_rate": 0.00017727335734973512,
"loss": 2.3034,
"step": 298
},
{
"epoch": 0.299,
"grad_norm": 0.765625,
"learning_rate": 0.00017705132427757895,
"loss": 2.3104,
"step": 299
},
{
"epoch": 0.3,
"grad_norm": 0.74609375,
"learning_rate": 0.00017682835235935236,
"loss": 2.3125,
"step": 300
},
{
"epoch": 0.301,
"grad_norm": 0.703125,
"learning_rate": 0.0001766044443118978,
"loss": 2.2883,
"step": 301
},
{
"epoch": 0.302,
"grad_norm": 0.6640625,
"learning_rate": 0.00017637960286346425,
"loss": 2.2707,
"step": 302
},
{
"epoch": 0.303,
"grad_norm": 0.71875,
"learning_rate": 0.0001761538307536737,
"loss": 2.285,
"step": 303
},
{
"epoch": 0.304,
"grad_norm": 0.671875,
"learning_rate": 0.00017592713073348807,
"loss": 2.2532,
"step": 304
},
{
"epoch": 0.305,
"grad_norm": 0.75390625,
"learning_rate": 0.00017569950556517566,
"loss": 2.2369,
"step": 305
},
{
"epoch": 0.306,
"grad_norm": 0.6875,
"learning_rate": 0.00017547095802227723,
"loss": 2.2466,
"step": 306
},
{
"epoch": 0.307,
"grad_norm": 0.6953125,
"learning_rate": 0.00017524149088957245,
"loss": 2.2617,
"step": 307
},
{
"epoch": 0.308,
"grad_norm": 0.75390625,
"learning_rate": 0.00017501110696304596,
"loss": 2.2783,
"step": 308
},
{
"epoch": 0.309,
"grad_norm": 0.7265625,
"learning_rate": 0.0001747798090498532,
"loss": 2.2501,
"step": 309
},
{
"epoch": 0.31,
"grad_norm": 0.76953125,
"learning_rate": 0.00017454759996828623,
"loss": 2.2314,
"step": 310
},
{
"epoch": 0.311,
"grad_norm": 1.1640625,
"learning_rate": 0.00017431448254773944,
"loss": 2.4266,
"step": 311
},
{
"epoch": 0.312,
"grad_norm": 0.81640625,
"learning_rate": 0.000174080459628675,
"loss": 2.2614,
"step": 312
},
{
"epoch": 0.313,
"grad_norm": 2.953125,
"learning_rate": 0.00017384553406258842,
"loss": 2.5082,
"step": 313
},
{
"epoch": 0.314,
"grad_norm": 1.0625,
"learning_rate": 0.00017360970871197346,
"loss": 2.3566,
"step": 314
},
{
"epoch": 0.315,
"grad_norm": 0.734375,
"learning_rate": 0.00017337298645028764,
"loss": 2.3117,
"step": 315
},
{
"epoch": 0.316,
"grad_norm": 0.96875,
"learning_rate": 0.00017313537016191706,
"loss": 2.2905,
"step": 316
},
{
"epoch": 0.317,
"grad_norm": 1.8828125,
"learning_rate": 0.00017289686274214118,
"loss": 2.311,
"step": 317
},
{
"epoch": 0.318,
"grad_norm": 1.9375,
"learning_rate": 0.0001726574670970976,
"loss": 2.2784,
"step": 318
},
{
"epoch": 0.319,
"grad_norm": 2.421875,
"learning_rate": 0.00017241718614374678,
"loss": 2.315,
"step": 319
},
{
"epoch": 0.32,
"grad_norm": 166.0,
"learning_rate": 0.00017217602280983623,
"loss": 2.2842,
"step": 320
},
{
"epoch": 0.321,
"grad_norm": 81.0,
"learning_rate": 0.0001719339800338651,
"loss": 2.3739,
"step": 321
},
{
"epoch": 0.322,
"grad_norm": 3.265625,
"learning_rate": 0.0001716910607650483,
"loss": 2.2612,
"step": 322
},
{
"epoch": 0.323,
"grad_norm": 6.46875,
"learning_rate": 0.00017144726796328034,
"loss": 2.3401,
"step": 323
},
{
"epoch": 0.324,
"grad_norm": 3.296875,
"learning_rate": 0.00017120260459909967,
"loss": 2.3751,
"step": 324
},
{
"epoch": 0.325,
"grad_norm": 1.0703125,
"learning_rate": 0.0001709570736536521,
"loss": 2.3854,
"step": 325
},
{
"epoch": 0.326,
"grad_norm": 0.74609375,
"learning_rate": 0.00017071067811865476,
"loss": 2.3424,
"step": 326
},
{
"epoch": 0.327,
"grad_norm": 0.77734375,
"learning_rate": 0.00017046342099635948,
"loss": 2.3893,
"step": 327
},
{
"epoch": 0.328,
"grad_norm": 0.69921875,
"learning_rate": 0.00017021530529951625,
"loss": 2.2686,
"step": 328
},
{
"epoch": 0.329,
"grad_norm": 0.69921875,
"learning_rate": 0.00016996633405133655,
"loss": 2.2848,
"step": 329
},
{
"epoch": 0.33,
"grad_norm": 0.66015625,
"learning_rate": 0.00016971651028545648,
"loss": 2.3436,
"step": 330
},
{
"epoch": 0.331,
"grad_norm": 0.68359375,
"learning_rate": 0.00016946583704589973,
"loss": 2.3179,
"step": 331
},
{
"epoch": 0.332,
"grad_norm": 0.67578125,
"learning_rate": 0.0001692143173870407,
"loss": 2.2736,
"step": 332
},
{
"epoch": 0.333,
"grad_norm": 0.65625,
"learning_rate": 0.000168961954373567,
"loss": 2.2638,
"step": 333
},
{
"epoch": 0.334,
"grad_norm": 2.421875,
"learning_rate": 0.0001687087510804423,
"loss": 2.2163,
"step": 334
},
{
"epoch": 0.335,
"grad_norm": 0.75390625,
"learning_rate": 0.00016845471059286887,
"loss": 2.2719,
"step": 335
},
{
"epoch": 0.336,
"grad_norm": 0.6953125,
"learning_rate": 0.00016819983600624986,
"loss": 2.2902,
"step": 336
},
{
"epoch": 0.337,
"grad_norm": 0.7109375,
"learning_rate": 0.00016794413042615168,
"loss": 2.2592,
"step": 337
},
{
"epoch": 0.338,
"grad_norm": 0.73828125,
"learning_rate": 0.00016768759696826608,
"loss": 2.3372,
"step": 338
},
{
"epoch": 0.339,
"grad_norm": 0.69140625,
"learning_rate": 0.00016743023875837233,
"loss": 2.2902,
"step": 339
},
{
"epoch": 0.34,
"grad_norm": 0.703125,
"learning_rate": 0.00016717205893229903,
"loss": 2.2491,
"step": 340
},
{
"epoch": 0.341,
"grad_norm": 0.73046875,
"learning_rate": 0.00016691306063588583,
"loss": 2.2438,
"step": 341
},
{
"epoch": 0.342,
"grad_norm": 0.6953125,
"learning_rate": 0.00016665324702494524,
"loss": 2.2902,
"step": 342
},
{
"epoch": 0.343,
"grad_norm": 0.7734375,
"learning_rate": 0.00016639262126522418,
"loss": 2.2529,
"step": 343
},
{
"epoch": 0.344,
"grad_norm": 0.76953125,
"learning_rate": 0.00016613118653236518,
"loss": 2.3238,
"step": 344
},
{
"epoch": 0.345,
"grad_norm": 0.74609375,
"learning_rate": 0.00016586894601186805,
"loss": 2.2791,
"step": 345
},
{
"epoch": 0.346,
"grad_norm": 0.80859375,
"learning_rate": 0.00016560590289905073,
"loss": 2.3074,
"step": 346
},
{
"epoch": 0.347,
"grad_norm": 0.7421875,
"learning_rate": 0.00016534206039901057,
"loss": 2.2004,
"step": 347
},
{
"epoch": 0.348,
"grad_norm": 0.7421875,
"learning_rate": 0.0001650774217265851,
"loss": 2.2595,
"step": 348
},
{
"epoch": 0.349,
"grad_norm": 0.68359375,
"learning_rate": 0.0001648119901063131,
"loss": 2.2285,
"step": 349
},
{
"epoch": 0.35,
"grad_norm": 0.765625,
"learning_rate": 0.00016454576877239507,
"loss": 2.2735,
"step": 350
},
{
"epoch": 0.351,
"grad_norm": 0.6953125,
"learning_rate": 0.00016427876096865394,
"loss": 2.2333,
"step": 351
},
{
"epoch": 0.352,
"grad_norm": 0.6875,
"learning_rate": 0.00016401096994849557,
"loss": 2.2694,
"step": 352
},
{
"epoch": 0.353,
"grad_norm": 0.79296875,
"learning_rate": 0.000163742398974869,
"loss": 2.2306,
"step": 353
},
{
"epoch": 0.354,
"grad_norm": 0.7578125,
"learning_rate": 0.00016347305132022677,
"loss": 2.2757,
"step": 354
},
{
"epoch": 0.355,
"grad_norm": 1.6953125,
"learning_rate": 0.0001632029302664851,
"loss": 2.3106,
"step": 355
},
{
"epoch": 0.356,
"grad_norm": 0.875,
"learning_rate": 0.00016293203910498376,
"loss": 2.2383,
"step": 356
},
{
"epoch": 0.357,
"grad_norm": 0.7734375,
"learning_rate": 0.00016266038113644607,
"loss": 2.2809,
"step": 357
},
{
"epoch": 0.358,
"grad_norm": 0.7265625,
"learning_rate": 0.00016238795967093864,
"loss": 2.235,
"step": 358
},
{
"epoch": 0.359,
"grad_norm": 0.79296875,
"learning_rate": 0.00016211477802783103,
"loss": 2.2461,
"step": 359
},
{
"epoch": 0.36,
"grad_norm": 0.703125,
"learning_rate": 0.0001618408395357554,
"loss": 2.3211,
"step": 360
},
{
"epoch": 0.361,
"grad_norm": 0.9765625,
"learning_rate": 0.0001615661475325658,
"loss": 2.2389,
"step": 361
},
{
"epoch": 0.362,
"grad_norm": 0.7421875,
"learning_rate": 0.00016129070536529766,
"loss": 2.3061,
"step": 362
},
{
"epoch": 0.363,
"grad_norm": 0.796875,
"learning_rate": 0.0001610145163901268,
"loss": 2.231,
"step": 363
},
{
"epoch": 0.364,
"grad_norm": 0.7890625,
"learning_rate": 0.00016073758397232868,
"loss": 2.2998,
"step": 364
},
{
"epoch": 0.365,
"grad_norm": 0.75,
"learning_rate": 0.0001604599114862375,
"loss": 2.3562,
"step": 365
},
{
"epoch": 0.366,
"grad_norm": 0.671875,
"learning_rate": 0.00016018150231520486,
"loss": 2.3519,
"step": 366
},
{
"epoch": 0.367,
"grad_norm": 0.65234375,
"learning_rate": 0.0001599023598515586,
"loss": 2.3189,
"step": 367
},
{
"epoch": 0.368,
"grad_norm": 0.66796875,
"learning_rate": 0.0001596224874965616,
"loss": 2.3226,
"step": 368
},
{
"epoch": 0.369,
"grad_norm": 0.66015625,
"learning_rate": 0.00015934188866037016,
"loss": 2.3127,
"step": 369
},
{
"epoch": 0.37,
"grad_norm": 0.6484375,
"learning_rate": 0.00015906056676199255,
"loss": 2.2712,
"step": 370
},
{
"epoch": 0.371,
"grad_norm": 0.6640625,
"learning_rate": 0.00015877852522924732,
"loss": 2.2647,
"step": 371
},
{
"epoch": 0.372,
"grad_norm": 0.65625,
"learning_rate": 0.00015849576749872157,
"loss": 2.3477,
"step": 372
},
{
"epoch": 0.373,
"grad_norm": 0.69140625,
"learning_rate": 0.00015821229701572896,
"loss": 2.3069,
"step": 373
},
{
"epoch": 0.374,
"grad_norm": 0.703125,
"learning_rate": 0.0001579281172342679,
"loss": 2.2372,
"step": 374
},
{
"epoch": 0.375,
"grad_norm": 20.75,
"learning_rate": 0.00015764323161697935,
"loss": 2.2838,
"step": 375
},
{
"epoch": 0.376,
"grad_norm": 36.75,
"learning_rate": 0.0001573576436351046,
"loss": 2.3038,
"step": 376
},
{
"epoch": 0.377,
"grad_norm": 1.0703125,
"learning_rate": 0.0001570713567684432,
"loss": 2.2962,
"step": 377
},
{
"epoch": 0.378,
"grad_norm": 0.80078125,
"learning_rate": 0.00015678437450531013,
"loss": 2.3049,
"step": 378
},
{
"epoch": 0.379,
"grad_norm": 0.7578125,
"learning_rate": 0.0001564967003424938,
"loss": 2.259,
"step": 379
},
{
"epoch": 0.38,
"grad_norm": 0.765625,
"learning_rate": 0.00015620833778521307,
"loss": 2.2974,
"step": 380
},
{
"epoch": 0.381,
"grad_norm": 0.796875,
"learning_rate": 0.0001559192903470747,
"loss": 2.394,
"step": 381
},
{
"epoch": 0.382,
"grad_norm": 0.75390625,
"learning_rate": 0.0001556295615500305,
"loss": 2.2193,
"step": 382
},
{
"epoch": 0.383,
"grad_norm": 0.7578125,
"learning_rate": 0.00015533915492433443,
"loss": 2.2806,
"step": 383
},
{
"epoch": 0.384,
"grad_norm": 0.77734375,
"learning_rate": 0.00015504807400849958,
"loss": 2.2132,
"step": 384
},
{
"epoch": 0.385,
"grad_norm": 0.66796875,
"learning_rate": 0.00015475632234925504,
"loss": 2.2832,
"step": 385
},
{
"epoch": 0.386,
"grad_norm": 0.76171875,
"learning_rate": 0.00015446390350150273,
"loss": 2.2467,
"step": 386
},
{
"epoch": 0.387,
"grad_norm": 0.67578125,
"learning_rate": 0.000154170821028274,
"loss": 2.3032,
"step": 387
},
{
"epoch": 0.388,
"grad_norm": 1.4765625,
"learning_rate": 0.0001538770785006863,
"loss": 2.2936,
"step": 388
},
{
"epoch": 0.389,
"grad_norm": 0.73828125,
"learning_rate": 0.00015358267949789966,
"loss": 2.2914,
"step": 389
},
{
"epoch": 0.39,
"grad_norm": 0.734375,
"learning_rate": 0.000153287627607073,
"loss": 2.2567,
"step": 390
},
{
"epoch": 0.391,
"grad_norm": 0.78125,
"learning_rate": 0.0001529919264233205,
"loss": 2.3406,
"step": 391
},
{
"epoch": 0.392,
"grad_norm": 0.72265625,
"learning_rate": 0.00015269557954966778,
"loss": 2.2966,
"step": 392
},
{
"epoch": 0.393,
"grad_norm": 0.703125,
"learning_rate": 0.00015239859059700794,
"loss": 2.2349,
"step": 393
},
{
"epoch": 0.394,
"grad_norm": 0.7109375,
"learning_rate": 0.00015210096318405767,
"loss": 2.3195,
"step": 394
},
{
"epoch": 0.395,
"grad_norm": 0.734375,
"learning_rate": 0.00015180270093731303,
"loss": 2.1866,
"step": 395
},
{
"epoch": 0.396,
"grad_norm": 1.5078125,
"learning_rate": 0.00015150380749100545,
"loss": 2.2624,
"step": 396
},
{
"epoch": 0.397,
"grad_norm": 0.765625,
"learning_rate": 0.00015120428648705717,
"loss": 2.2619,
"step": 397
},
{
"epoch": 0.398,
"grad_norm": 0.7734375,
"learning_rate": 0.00015090414157503714,
"loss": 2.341,
"step": 398
},
{
"epoch": 0.399,
"grad_norm": 0.69921875,
"learning_rate": 0.00015060337641211637,
"loss": 2.2264,
"step": 399
},
{
"epoch": 0.4,
"grad_norm": 0.703125,
"learning_rate": 0.00015030199466302353,
"loss": 2.3528,
"step": 400
},
{
"epoch": 0.401,
"grad_norm": 0.64453125,
"learning_rate": 0.00015000000000000001,
"loss": 2.2615,
"step": 401
},
{
"epoch": 0.402,
"grad_norm": 0.70703125,
"learning_rate": 0.00014969739610275556,
"loss": 2.2348,
"step": 402
},
{
"epoch": 0.403,
"grad_norm": 0.671875,
"learning_rate": 0.0001493941866584231,
"loss": 2.2387,
"step": 403
},
{
"epoch": 0.404,
"grad_norm": 0.71875,
"learning_rate": 0.00014909037536151409,
"loss": 2.3262,
"step": 404
},
{
"epoch": 0.405,
"grad_norm": 0.67578125,
"learning_rate": 0.0001487859659138733,
"loss": 2.2717,
"step": 405
},
{
"epoch": 0.406,
"grad_norm": 0.703125,
"learning_rate": 0.00014848096202463372,
"loss": 2.2678,
"step": 406
},
{
"epoch": 0.407,
"grad_norm": 0.88671875,
"learning_rate": 0.00014817536741017152,
"loss": 2.3754,
"step": 407
},
{
"epoch": 0.408,
"grad_norm": 0.6953125,
"learning_rate": 0.0001478691857940607,
"loss": 2.3504,
"step": 408
},
{
"epoch": 0.409,
"grad_norm": 0.71484375,
"learning_rate": 0.00014756242090702756,
"loss": 2.315,
"step": 409
},
{
"epoch": 0.41,
"grad_norm": 0.71484375,
"learning_rate": 0.00014725507648690543,
"loss": 2.2966,
"step": 410
},
{
"epoch": 0.411,
"grad_norm": 0.70703125,
"learning_rate": 0.00014694715627858908,
"loss": 2.3045,
"step": 411
},
{
"epoch": 0.412,
"grad_norm": 0.9296875,
"learning_rate": 0.00014663866403398913,
"loss": 2.3762,
"step": 412
},
{
"epoch": 0.413,
"grad_norm": 0.63671875,
"learning_rate": 0.00014632960351198618,
"loss": 2.3385,
"step": 413
},
{
"epoch": 0.414,
"grad_norm": 0.8203125,
"learning_rate": 0.00014601997847838518,
"loss": 2.2741,
"step": 414
},
{
"epoch": 0.415,
"grad_norm": 0.6328125,
"learning_rate": 0.00014570979270586945,
"loss": 2.2654,
"step": 415
},
{
"epoch": 0.416,
"grad_norm": 0.61328125,
"learning_rate": 0.00014539904997395468,
"loss": 2.2348,
"step": 416
},
{
"epoch": 0.417,
"grad_norm": 1.96875,
"learning_rate": 0.00014508775406894307,
"loss": 2.2532,
"step": 417
},
{
"epoch": 0.418,
"grad_norm": 0.796875,
"learning_rate": 0.00014477590878387696,
"loss": 2.3118,
"step": 418
},
{
"epoch": 0.419,
"grad_norm": 0.703125,
"learning_rate": 0.00014446351791849276,
"loss": 2.314,
"step": 419
},
{
"epoch": 0.42,
"grad_norm": 0.67578125,
"learning_rate": 0.00014415058527917452,
"loss": 2.2837,
"step": 420
},
{
"epoch": 0.421,
"grad_norm": 0.66015625,
"learning_rate": 0.00014383711467890774,
"loss": 2.2652,
"step": 421
},
{
"epoch": 0.422,
"grad_norm": 0.7421875,
"learning_rate": 0.00014352310993723277,
"loss": 2.402,
"step": 422
},
{
"epoch": 0.423,
"grad_norm": 0.65625,
"learning_rate": 0.00014320857488019824,
"loss": 2.287,
"step": 423
},
{
"epoch": 0.424,
"grad_norm": 0.703125,
"learning_rate": 0.0001428935133403146,
"loss": 2.3442,
"step": 424
},
{
"epoch": 0.425,
"grad_norm": 0.75,
"learning_rate": 0.00014257792915650728,
"loss": 2.3215,
"step": 425
},
{
"epoch": 0.426,
"grad_norm": 0.6953125,
"learning_rate": 0.00014226182617406996,
"loss": 2.3306,
"step": 426
},
{
"epoch": 0.427,
"grad_norm": 0.66796875,
"learning_rate": 0.00014194520824461771,
"loss": 2.2615,
"step": 427
},
{
"epoch": 0.428,
"grad_norm": 0.9609375,
"learning_rate": 0.00014162807922604012,
"loss": 2.438,
"step": 428
},
{
"epoch": 0.429,
"grad_norm": 0.734375,
"learning_rate": 0.0001413104429824542,
"loss": 2.3533,
"step": 429
},
{
"epoch": 0.43,
"grad_norm": 0.7734375,
"learning_rate": 0.00014099230338415728,
"loss": 2.2997,
"step": 430
},
{
"epoch": 0.431,
"grad_norm": 0.60546875,
"learning_rate": 0.00014067366430758004,
"loss": 2.3224,
"step": 431
},
{
"epoch": 0.432,
"grad_norm": 0.6484375,
"learning_rate": 0.00014035452963523902,
"loss": 2.3141,
"step": 432
},
{
"epoch": 0.433,
"grad_norm": 0.59765625,
"learning_rate": 0.00014003490325568954,
"loss": 2.2947,
"step": 433
},
{
"epoch": 0.434,
"grad_norm": 0.62109375,
"learning_rate": 0.00013971478906347806,
"loss": 2.2819,
"step": 434
},
{
"epoch": 0.435,
"grad_norm": 0.6171875,
"learning_rate": 0.00013939419095909512,
"loss": 2.2316,
"step": 435
},
{
"epoch": 0.436,
"grad_norm": 0.578125,
"learning_rate": 0.00013907311284892736,
"loss": 2.2626,
"step": 436
},
{
"epoch": 0.437,
"grad_norm": 0.63671875,
"learning_rate": 0.0001387515586452103,
"loss": 2.357,
"step": 437
},
{
"epoch": 0.438,
"grad_norm": 1.3828125,
"learning_rate": 0.00013842953226598037,
"loss": 2.4577,
"step": 438
},
{
"epoch": 0.439,
"grad_norm": 0.6015625,
"learning_rate": 0.00013810703763502744,
"loss": 2.2553,
"step": 439
},
{
"epoch": 0.44,
"grad_norm": 0.59765625,
"learning_rate": 0.00013778407868184672,
"loss": 2.3154,
"step": 440
},
{
"epoch": 0.441,
"grad_norm": 0.62109375,
"learning_rate": 0.00013746065934159123,
"loss": 2.2929,
"step": 441
},
{
"epoch": 0.442,
"grad_norm": 0.60546875,
"learning_rate": 0.00013713678355502351,
"loss": 2.3263,
"step": 442
},
{
"epoch": 0.443,
"grad_norm": 0.65625,
"learning_rate": 0.00013681245526846783,
"loss": 2.2654,
"step": 443
},
{
"epoch": 0.444,
"grad_norm": 0.765625,
"learning_rate": 0.00013648767843376196,
"loss": 2.3294,
"step": 444
},
{
"epoch": 0.445,
"grad_norm": 0.62890625,
"learning_rate": 0.00013616245700820922,
"loss": 2.3281,
"step": 445
},
{
"epoch": 0.446,
"grad_norm": 0.62890625,
"learning_rate": 0.00013583679495453,
"loss": 2.4152,
"step": 446
},
{
"epoch": 0.447,
"grad_norm": 0.6875,
"learning_rate": 0.0001355106962408137,
"loss": 2.3148,
"step": 447
},
{
"epoch": 0.448,
"grad_norm": 0.671875,
"learning_rate": 0.00013518416484047018,
"loss": 2.3279,
"step": 448
},
{
"epoch": 0.449,
"grad_norm": 0.609375,
"learning_rate": 0.00013485720473218154,
"loss": 2.4039,
"step": 449
},
{
"epoch": 0.45,
"grad_norm": 0.68359375,
"learning_rate": 0.00013452981989985348,
"loss": 2.3308,
"step": 450
},
{
"epoch": 0.451,
"grad_norm": 0.61328125,
"learning_rate": 0.00013420201433256689,
"loss": 2.363,
"step": 451
},
{
"epoch": 0.452,
"grad_norm": 0.6015625,
"learning_rate": 0.00013387379202452917,
"loss": 2.3299,
"step": 452
},
{
"epoch": 0.453,
"grad_norm": 0.578125,
"learning_rate": 0.00013354515697502553,
"loss": 2.342,
"step": 453
},
{
"epoch": 0.454,
"grad_norm": 0.57421875,
"learning_rate": 0.00013321611318837032,
"loss": 2.36,
"step": 454
},
{
"epoch": 0.455,
"grad_norm": 0.57421875,
"learning_rate": 0.00013288666467385833,
"loss": 2.2295,
"step": 455
},
{
"epoch": 0.456,
"grad_norm": 0.60546875,
"learning_rate": 0.00013255681544571568,
"loss": 2.2661,
"step": 456
},
{
"epoch": 0.457,
"grad_norm": 0.5625,
"learning_rate": 0.00013222656952305113,
"loss": 2.2833,
"step": 457
},
{
"epoch": 0.458,
"grad_norm": 0.58203125,
"learning_rate": 0.00013189593092980702,
"loss": 2.3106,
"step": 458
},
{
"epoch": 0.459,
"grad_norm": 0.83984375,
"learning_rate": 0.00013156490369471027,
"loss": 2.2704,
"step": 459
},
{
"epoch": 0.46,
"grad_norm": 0.609375,
"learning_rate": 0.00013123349185122327,
"loss": 2.3332,
"step": 460
},
{
"epoch": 0.461,
"grad_norm": 0.56640625,
"learning_rate": 0.00013090169943749476,
"loss": 2.2981,
"step": 461
},
{
"epoch": 0.462,
"grad_norm": 0.5859375,
"learning_rate": 0.00013056953049631057,
"loss": 2.309,
"step": 462
},
{
"epoch": 0.463,
"grad_norm": 0.5625,
"learning_rate": 0.00013023698907504446,
"loss": 2.3009,
"step": 463
},
{
"epoch": 0.464,
"grad_norm": 0.625,
"learning_rate": 0.00012990407922560868,
"loss": 2.3071,
"step": 464
},
{
"epoch": 0.465,
"grad_norm": 0.6171875,
"learning_rate": 0.00012957080500440468,
"loss": 2.2825,
"step": 465
},
{
"epoch": 0.466,
"grad_norm": 0.66796875,
"learning_rate": 0.00012923717047227368,
"loss": 2.3498,
"step": 466
},
{
"epoch": 0.467,
"grad_norm": 0.62890625,
"learning_rate": 0.00012890317969444716,
"loss": 2.282,
"step": 467
},
{
"epoch": 0.468,
"grad_norm": 0.63671875,
"learning_rate": 0.00012856883674049736,
"loss": 2.3279,
"step": 468
},
{
"epoch": 0.469,
"grad_norm": 0.98046875,
"learning_rate": 0.00012823414568428768,
"loss": 2.5406,
"step": 469
},
{
"epoch": 0.47,
"grad_norm": 0.69140625,
"learning_rate": 0.00012789911060392294,
"loss": 2.3485,
"step": 470
},
{
"epoch": 0.471,
"grad_norm": 0.85546875,
"learning_rate": 0.0001275637355816999,
"loss": 2.3823,
"step": 471
},
{
"epoch": 0.472,
"grad_norm": 0.703125,
"learning_rate": 0.00012722802470405744,
"loss": 2.3045,
"step": 472
},
{
"epoch": 0.473,
"grad_norm": 0.63671875,
"learning_rate": 0.00012689198206152657,
"loss": 2.2287,
"step": 473
},
{
"epoch": 0.474,
"grad_norm": 0.875,
"learning_rate": 0.00012655561174868088,
"loss": 2.3778,
"step": 474
},
{
"epoch": 0.475,
"grad_norm": 0.76171875,
"learning_rate": 0.00012621891786408648,
"loss": 2.3843,
"step": 475
},
{
"epoch": 0.476,
"grad_norm": 0.65625,
"learning_rate": 0.00012588190451025207,
"loss": 2.3916,
"step": 476
},
{
"epoch": 0.477,
"grad_norm": 0.62109375,
"learning_rate": 0.00012554457579357905,
"loss": 2.4068,
"step": 477
},
{
"epoch": 0.478,
"grad_norm": 0.69140625,
"learning_rate": 0.0001252069358243114,
"loss": 2.4231,
"step": 478
},
{
"epoch": 0.479,
"grad_norm": 0.58984375,
"learning_rate": 0.0001248689887164855,
"loss": 2.3278,
"step": 479
},
{
"epoch": 0.48,
"grad_norm": 1.0390625,
"learning_rate": 0.00012453073858788026,
"loss": 2.2753,
"step": 480
},
{
"epoch": 0.481,
"grad_norm": 0.703125,
"learning_rate": 0.00012419218955996676,
"loss": 2.3301,
"step": 481
},
{
"epoch": 0.482,
"grad_norm": 0.6328125,
"learning_rate": 0.0001238533457578581,
"loss": 2.3209,
"step": 482
},
{
"epoch": 0.483,
"grad_norm": 0.65625,
"learning_rate": 0.000123514211310259,
"loss": 2.3683,
"step": 483
},
{
"epoch": 0.484,
"grad_norm": 0.78515625,
"learning_rate": 0.00012317479034941573,
"loss": 2.3522,
"step": 484
},
{
"epoch": 0.485,
"grad_norm": 0.5859375,
"learning_rate": 0.00012283508701106557,
"loss": 2.3755,
"step": 485
},
{
"epoch": 0.486,
"grad_norm": 0.5625,
"learning_rate": 0.0001224951054343865,
"loss": 2.3384,
"step": 486
},
{
"epoch": 0.487,
"grad_norm": 0.64453125,
"learning_rate": 0.00012215484976194676,
"loss": 2.3361,
"step": 487
},
{
"epoch": 0.488,
"grad_norm": 0.578125,
"learning_rate": 0.00012181432413965428,
"loss": 2.3609,
"step": 488
},
{
"epoch": 0.489,
"grad_norm": 0.58203125,
"learning_rate": 0.00012147353271670634,
"loss": 2.2719,
"step": 489
},
{
"epoch": 0.49,
"grad_norm": 0.625,
"learning_rate": 0.00012113247964553888,
"loss": 2.3613,
"step": 490
},
{
"epoch": 0.491,
"grad_norm": 0.58203125,
"learning_rate": 0.00012079116908177593,
"loss": 2.3598,
"step": 491
},
{
"epoch": 0.492,
"grad_norm": 3.921875,
"learning_rate": 0.00012044960518417903,
"loss": 2.2725,
"step": 492
},
{
"epoch": 0.493,
"grad_norm": 0.65234375,
"learning_rate": 0.00012010779211459648,
"loss": 2.3553,
"step": 493
},
{
"epoch": 0.494,
"grad_norm": 0.6484375,
"learning_rate": 0.00011976573403791262,
"loss": 2.3235,
"step": 494
},
{
"epoch": 0.495,
"grad_norm": 0.56640625,
"learning_rate": 0.0001194234351219972,
"loss": 2.3259,
"step": 495
},
{
"epoch": 0.496,
"grad_norm": 0.578125,
"learning_rate": 0.00011908089953765449,
"loss": 2.3102,
"step": 496
},
{
"epoch": 0.497,
"grad_norm": 0.58203125,
"learning_rate": 0.00011873813145857249,
"loss": 2.3712,
"step": 497
},
{
"epoch": 0.498,
"grad_norm": 0.55078125,
"learning_rate": 0.00011839513506127203,
"loss": 2.3372,
"step": 498
},
{
"epoch": 0.499,
"grad_norm": 0.55859375,
"learning_rate": 0.00011805191452505602,
"loss": 2.2306,
"step": 499
},
{
"epoch": 0.5,
"grad_norm": 0.6796875,
"learning_rate": 0.00011770847403195834,
"loss": 2.3247,
"step": 500
},
{
"epoch": 0.5,
"eval_loss": 2.4592103958129883,
"eval_runtime": 23.6853,
"eval_samples_per_second": 21.11,
"eval_steps_per_second": 1.351,
"step": 500
},
{
"epoch": 0.501,
"grad_norm": 0.5390625,
"learning_rate": 0.00011736481776669306,
"loss": 2.3064,
"step": 501
},
{
"epoch": 0.502,
"grad_norm": 0.53125,
"learning_rate": 0.00011702094991660326,
"loss": 2.3575,
"step": 502
},
{
"epoch": 0.503,
"grad_norm": 0.57421875,
"learning_rate": 0.00011667687467161024,
"loss": 2.3675,
"step": 503
},
{
"epoch": 0.504,
"grad_norm": 0.55859375,
"learning_rate": 0.00011633259622416224,
"loss": 2.2809,
"step": 504
},
{
"epoch": 0.505,
"grad_norm": 0.61328125,
"learning_rate": 0.0001159881187691835,
"loss": 2.3624,
"step": 505
},
{
"epoch": 0.506,
"grad_norm": 0.57421875,
"learning_rate": 0.0001156434465040231,
"loss": 2.3088,
"step": 506
},
{
"epoch": 0.507,
"grad_norm": 0.5625,
"learning_rate": 0.00011529858362840382,
"loss": 2.3574,
"step": 507
},
{
"epoch": 0.508,
"grad_norm": 0.55078125,
"learning_rate": 0.00011495353434437098,
"loss": 2.3009,
"step": 508
},
{
"epoch": 0.509,
"grad_norm": 0.55859375,
"learning_rate": 0.00011460830285624118,
"loss": 2.3158,
"step": 509
},
{
"epoch": 0.51,
"grad_norm": 0.59765625,
"learning_rate": 0.00011426289337055119,
"loss": 2.3213,
"step": 510
},
{
"epoch": 0.511,
"grad_norm": 0.57421875,
"learning_rate": 0.00011391731009600654,
"loss": 2.3374,
"step": 511
},
{
"epoch": 0.512,
"grad_norm": 0.546875,
"learning_rate": 0.00011357155724343045,
"loss": 2.3362,
"step": 512
},
{
"epoch": 0.513,
"grad_norm": 0.57421875,
"learning_rate": 0.00011322563902571226,
"loss": 2.3183,
"step": 513
},
{
"epoch": 0.514,
"grad_norm": 0.55078125,
"learning_rate": 0.0001128795596577563,
"loss": 2.3554,
"step": 514
},
{
"epoch": 0.515,
"grad_norm": 0.5390625,
"learning_rate": 0.00011253332335643043,
"loss": 2.3667,
"step": 515
},
{
"epoch": 0.516,
"grad_norm": 0.55859375,
"learning_rate": 0.00011218693434051475,
"loss": 2.3427,
"step": 516
},
{
"epoch": 0.517,
"grad_norm": 0.55078125,
"learning_rate": 0.00011184039683065013,
"loss": 2.2795,
"step": 517
},
{
"epoch": 0.518,
"grad_norm": 0.55859375,
"learning_rate": 0.00011149371504928668,
"loss": 2.3214,
"step": 518
},
{
"epoch": 0.519,
"grad_norm": 0.5234375,
"learning_rate": 0.00011114689322063255,
"loss": 2.3325,
"step": 519
},
{
"epoch": 0.52,
"grad_norm": 0.52734375,
"learning_rate": 0.0001107999355706023,
"loss": 2.3662,
"step": 520
},
{
"epoch": 0.521,
"grad_norm": 0.78515625,
"learning_rate": 0.00011045284632676536,
"loss": 2.3307,
"step": 521
},
{
"epoch": 0.522,
"grad_norm": 0.51171875,
"learning_rate": 0.00011010562971829463,
"loss": 2.3141,
"step": 522
},
{
"epoch": 0.523,
"grad_norm": 0.52734375,
"learning_rate": 0.00010975828997591495,
"loss": 2.3271,
"step": 523
},
{
"epoch": 0.524,
"grad_norm": 0.515625,
"learning_rate": 0.00010941083133185146,
"loss": 2.2998,
"step": 524
},
{
"epoch": 0.525,
"grad_norm": 0.53515625,
"learning_rate": 0.00010906325801977804,
"loss": 2.3527,
"step": 525
},
{
"epoch": 0.526,
"grad_norm": 0.546875,
"learning_rate": 0.00010871557427476583,
"loss": 2.338,
"step": 526
},
{
"epoch": 0.527,
"grad_norm": 0.5234375,
"learning_rate": 0.00010836778433323158,
"loss": 2.3353,
"step": 527
},
{
"epoch": 0.528,
"grad_norm": 0.5078125,
"learning_rate": 0.00010801989243288589,
"loss": 2.3122,
"step": 528
},
{
"epoch": 0.529,
"grad_norm": 0.50390625,
"learning_rate": 0.00010767190281268187,
"loss": 2.3208,
"step": 529
},
{
"epoch": 0.53,
"grad_norm": 0.51171875,
"learning_rate": 0.00010732381971276318,
"loss": 2.317,
"step": 530
},
{
"epoch": 0.531,
"grad_norm": 0.5625,
"learning_rate": 0.00010697564737441252,
"loss": 2.3857,
"step": 531
},
{
"epoch": 0.532,
"grad_norm": 0.5078125,
"learning_rate": 0.00010662739004000005,
"loss": 2.3655,
"step": 532
},
{
"epoch": 0.533,
"grad_norm": 0.478515625,
"learning_rate": 0.00010627905195293135,
"loss": 2.3453,
"step": 533
},
{
"epoch": 0.534,
"grad_norm": 0.474609375,
"learning_rate": 0.00010593063735759618,
"loss": 2.4349,
"step": 534
},
{
"epoch": 0.535,
"grad_norm": 0.50390625,
"learning_rate": 0.00010558215049931638,
"loss": 2.4106,
"step": 535
},
{
"epoch": 0.536,
"grad_norm": 0.51953125,
"learning_rate": 0.0001052335956242944,
"loss": 2.3444,
"step": 536
},
{
"epoch": 0.537,
"grad_norm": 0.4609375,
"learning_rate": 0.00010488497697956135,
"loss": 2.387,
"step": 537
},
{
"epoch": 0.538,
"grad_norm": 0.48828125,
"learning_rate": 0.00010453629881292538,
"loss": 2.3259,
"step": 538
},
{
"epoch": 0.539,
"grad_norm": 0.484375,
"learning_rate": 0.00010418756537291996,
"loss": 2.3679,
"step": 539
},
{
"epoch": 0.54,
"grad_norm": 0.484375,
"learning_rate": 0.00010383878090875201,
"loss": 2.3782,
"step": 540
},
{
"epoch": 0.541,
"grad_norm": 0.470703125,
"learning_rate": 0.00010348994967025012,
"loss": 2.3477,
"step": 541
},
{
"epoch": 0.542,
"grad_norm": 0.9375,
"learning_rate": 0.00010314107590781284,
"loss": 2.3404,
"step": 542
},
{
"epoch": 0.543,
"grad_norm": 0.4921875,
"learning_rate": 0.0001027921638723569,
"loss": 2.3693,
"step": 543
},
{
"epoch": 0.544,
"grad_norm": 0.443359375,
"learning_rate": 0.00010244321781526533,
"loss": 2.3677,
"step": 544
},
{
"epoch": 0.545,
"grad_norm": 0.51171875,
"learning_rate": 0.0001020942419883357,
"loss": 2.3876,
"step": 545
},
{
"epoch": 0.546,
"grad_norm": 0.62890625,
"learning_rate": 0.00010174524064372837,
"loss": 2.4314,
"step": 546
},
{
"epoch": 0.547,
"grad_norm": 0.79296875,
"learning_rate": 0.00010139621803391455,
"loss": 2.5126,
"step": 547
},
{
"epoch": 0.548,
"grad_norm": 0.46875,
"learning_rate": 0.00010104717841162458,
"loss": 2.3566,
"step": 548
},
{
"epoch": 0.549,
"grad_norm": 0.515625,
"learning_rate": 0.00010069812602979615,
"loss": 2.4065,
"step": 549
},
{
"epoch": 0.55,
"grad_norm": 1.9765625,
"learning_rate": 0.00010034906514152238,
"loss": 2.3664,
"step": 550
},
{
"epoch": 0.551,
"grad_norm": 0.486328125,
"learning_rate": 0.0001,
"loss": 2.4679,
"step": 551
},
{
"epoch": 0.552,
"grad_norm": 0.474609375,
"learning_rate": 9.965093485847767e-05,
"loss": 2.3533,
"step": 552
},
{
"epoch": 0.553,
"grad_norm": 0.44140625,
"learning_rate": 9.930187397020386e-05,
"loss": 2.4311,
"step": 553
},
{
"epoch": 0.554,
"grad_norm": 0.478515625,
"learning_rate": 9.895282158837545e-05,
"loss": 2.4134,
"step": 554
},
{
"epoch": 0.555,
"grad_norm": 0.443359375,
"learning_rate": 9.860378196608549e-05,
"loss": 2.3801,
"step": 555
},
{
"epoch": 0.556,
"grad_norm": 0.44140625,
"learning_rate": 9.825475935627165e-05,
"loss": 2.3335,
"step": 556
},
{
"epoch": 0.557,
"grad_norm": 0.455078125,
"learning_rate": 9.790575801166432e-05,
"loss": 2.3869,
"step": 557
},
{
"epoch": 0.558,
"grad_norm": 0.453125,
"learning_rate": 9.755678218473469e-05,
"loss": 2.4241,
"step": 558
},
{
"epoch": 0.559,
"grad_norm": 0.466796875,
"learning_rate": 9.720783612764314e-05,
"loss": 2.3583,
"step": 559
},
{
"epoch": 0.56,
"grad_norm": 0.4453125,
"learning_rate": 9.685892409218717e-05,
"loss": 2.3754,
"step": 560
},
{
"epoch": 0.561,
"grad_norm": 0.43359375,
"learning_rate": 9.651005032974994e-05,
"loss": 2.3859,
"step": 561
},
{
"epoch": 0.562,
"grad_norm": 0.478515625,
"learning_rate": 9.616121909124801e-05,
"loss": 2.3954,
"step": 562
},
{
"epoch": 0.563,
"grad_norm": 1.0546875,
"learning_rate": 9.581243462708006e-05,
"loss": 2.3813,
"step": 563
},
{
"epoch": 0.564,
"grad_norm": 0.427734375,
"learning_rate": 9.546370118707463e-05,
"loss": 2.3673,
"step": 564
},
{
"epoch": 0.565,
"grad_norm": 0.439453125,
"learning_rate": 9.511502302043868e-05,
"loss": 2.4001,
"step": 565
},
{
"epoch": 0.566,
"grad_norm": 0.4375,
"learning_rate": 9.476640437570562e-05,
"loss": 2.3565,
"step": 566
},
{
"epoch": 0.567,
"grad_norm": 0.478515625,
"learning_rate": 9.441784950068362e-05,
"loss": 2.4309,
"step": 567
},
{
"epoch": 0.568,
"grad_norm": 0.44921875,
"learning_rate": 9.406936264240386e-05,
"loss": 2.4603,
"step": 568
},
{
"epoch": 0.569,
"grad_norm": 0.42578125,
"learning_rate": 9.372094804706867e-05,
"loss": 2.4104,
"step": 569
},
{
"epoch": 0.57,
"grad_norm": 0.43359375,
"learning_rate": 9.337260996000002e-05,
"loss": 2.3554,
"step": 570
},
{
"epoch": 0.571,
"grad_norm": 0.4453125,
"learning_rate": 9.302435262558747e-05,
"loss": 2.3554,
"step": 571
},
{
"epoch": 0.572,
"grad_norm": 0.439453125,
"learning_rate": 9.267618028723686e-05,
"loss": 2.443,
"step": 572
},
{
"epoch": 0.573,
"grad_norm": 0.41796875,
"learning_rate": 9.232809718731814e-05,
"loss": 2.4322,
"step": 573
},
{
"epoch": 0.574,
"grad_norm": 0.412109375,
"learning_rate": 9.198010756711412e-05,
"loss": 2.3285,
"step": 574
},
{
"epoch": 0.575,
"grad_norm": 0.416015625,
"learning_rate": 9.163221566676847e-05,
"loss": 2.3862,
"step": 575
},
{
"epoch": 0.576,
"grad_norm": 0.427734375,
"learning_rate": 9.128442572523417e-05,
"loss": 2.4323,
"step": 576
},
{
"epoch": 0.577,
"grad_norm": 0.435546875,
"learning_rate": 9.093674198022201e-05,
"loss": 2.4212,
"step": 577
},
{
"epoch": 0.578,
"grad_norm": 0.44140625,
"learning_rate": 9.058916866814858e-05,
"loss": 2.4744,
"step": 578
},
{
"epoch": 0.579,
"grad_norm": 0.5078125,
"learning_rate": 9.024171002408506e-05,
"loss": 2.4637,
"step": 579
},
{
"epoch": 0.58,
"grad_norm": 0.43359375,
"learning_rate": 8.989437028170537e-05,
"loss": 2.3661,
"step": 580
},
{
"epoch": 0.581,
"grad_norm": 0.466796875,
"learning_rate": 8.954715367323468e-05,
"loss": 2.4102,
"step": 581
},
{
"epoch": 0.582,
"grad_norm": 0.427734375,
"learning_rate": 8.920006442939772e-05,
"loss": 2.4211,
"step": 582
},
{
"epoch": 0.583,
"grad_norm": 0.439453125,
"learning_rate": 8.885310677936746e-05,
"loss": 2.4349,
"step": 583
},
{
"epoch": 0.584,
"grad_norm": 1.296875,
"learning_rate": 8.850628495071336e-05,
"loss": 2.4152,
"step": 584
},
{
"epoch": 0.585,
"grad_norm": 0.470703125,
"learning_rate": 8.81596031693499e-05,
"loss": 2.4262,
"step": 585
},
{
"epoch": 0.586,
"grad_norm": 0.42578125,
"learning_rate": 8.781306565948528e-05,
"loss": 2.4667,
"step": 586
},
{
"epoch": 0.587,
"grad_norm": 0.416015625,
"learning_rate": 8.746667664356956e-05,
"loss": 2.4461,
"step": 587
},
{
"epoch": 0.588,
"grad_norm": 0.44140625,
"learning_rate": 8.712044034224374e-05,
"loss": 2.4536,
"step": 588
},
{
"epoch": 0.589,
"grad_norm": 0.443359375,
"learning_rate": 8.677436097428775e-05,
"loss": 2.3793,
"step": 589
},
{
"epoch": 0.59,
"grad_norm": 0.408203125,
"learning_rate": 8.642844275656957e-05,
"loss": 2.3652,
"step": 590
},
{
"epoch": 0.591,
"grad_norm": 0.44140625,
"learning_rate": 8.608268990399349e-05,
"loss": 2.3829,
"step": 591
},
{
"epoch": 0.592,
"grad_norm": 0.435546875,
"learning_rate": 8.573710662944885e-05,
"loss": 2.4227,
"step": 592
},
{
"epoch": 0.593,
"grad_norm": 0.41796875,
"learning_rate": 8.539169714375885e-05,
"loss": 2.4205,
"step": 593
},
{
"epoch": 0.594,
"grad_norm": 0.4375,
"learning_rate": 8.504646565562906e-05,
"loss": 2.3441,
"step": 594
},
{
"epoch": 0.595,
"grad_norm": 0.416015625,
"learning_rate": 8.47014163715962e-05,
"loss": 2.4742,
"step": 595
},
{
"epoch": 0.596,
"grad_norm": 0.408203125,
"learning_rate": 8.435655349597689e-05,
"loss": 2.4074,
"step": 596
},
{
"epoch": 0.597,
"grad_norm": 0.431640625,
"learning_rate": 8.401188123081653e-05,
"loss": 2.3733,
"step": 597
},
{
"epoch": 0.598,
"grad_norm": 0.400390625,
"learning_rate": 8.366740377583781e-05,
"loss": 2.3856,
"step": 598
},
{
"epoch": 0.599,
"grad_norm": 0.41796875,
"learning_rate": 8.332312532838978e-05,
"loss": 2.4065,
"step": 599
},
{
"epoch": 0.6,
"grad_norm": 0.408203125,
"learning_rate": 8.297905008339677e-05,
"loss": 2.4104,
"step": 600
},
{
"epoch": 0.601,
"grad_norm": 0.4140625,
"learning_rate": 8.263518223330697e-05,
"loss": 2.4794,
"step": 601
},
{
"epoch": 0.602,
"grad_norm": 0.71484375,
"learning_rate": 8.229152596804168e-05,
"loss": 2.3402,
"step": 602
},
{
"epoch": 0.603,
"grad_norm": 0.400390625,
"learning_rate": 8.194808547494401e-05,
"loss": 2.4263,
"step": 603
},
{
"epoch": 0.604,
"grad_norm": 0.4140625,
"learning_rate": 8.160486493872798e-05,
"loss": 2.4263,
"step": 604
},
{
"epoch": 0.605,
"grad_norm": 1.2265625,
"learning_rate": 8.126186854142752e-05,
"loss": 2.3963,
"step": 605
},
{
"epoch": 0.606,
"grad_norm": 0.408203125,
"learning_rate": 8.091910046234552e-05,
"loss": 2.3798,
"step": 606
},
{
"epoch": 0.607,
"grad_norm": 0.396484375,
"learning_rate": 8.057656487800282e-05,
"loss": 2.4172,
"step": 607
},
{
"epoch": 0.608,
"grad_norm": 0.470703125,
"learning_rate": 8.023426596208739e-05,
"loss": 2.3866,
"step": 608
},
{
"epoch": 0.609,
"grad_norm": 0.419921875,
"learning_rate": 7.989220788540355e-05,
"loss": 2.3881,
"step": 609
},
{
"epoch": 0.61,
"grad_norm": 0.435546875,
"learning_rate": 7.955039481582097e-05,
"loss": 2.3612,
"step": 610
},
{
"epoch": 0.611,
"grad_norm": 0.384765625,
"learning_rate": 7.920883091822408e-05,
"loss": 2.3408,
"step": 611
},
{
"epoch": 0.612,
"grad_norm": 0.384765625,
"learning_rate": 7.886752035446114e-05,
"loss": 2.4884,
"step": 612
},
{
"epoch": 0.613,
"grad_norm": 0.39453125,
"learning_rate": 7.852646728329368e-05,
"loss": 2.4139,
"step": 613
},
{
"epoch": 0.614,
"grad_norm": 0.40625,
"learning_rate": 7.818567586034577e-05,
"loss": 2.4098,
"step": 614
},
{
"epoch": 0.615,
"grad_norm": 0.3984375,
"learning_rate": 7.784515023805328e-05,
"loss": 2.4564,
"step": 615
},
{
"epoch": 0.616,
"grad_norm": 0.400390625,
"learning_rate": 7.750489456561352e-05,
"loss": 2.4217,
"step": 616
},
{
"epoch": 0.617,
"grad_norm": 0.4140625,
"learning_rate": 7.716491298893442e-05,
"loss": 2.3497,
"step": 617
},
{
"epoch": 0.618,
"grad_norm": 0.40234375,
"learning_rate": 7.682520965058428e-05,
"loss": 2.3541,
"step": 618
},
{
"epoch": 0.619,
"grad_norm": 0.421875,
"learning_rate": 7.6485788689741e-05,
"loss": 2.3899,
"step": 619
},
{
"epoch": 0.62,
"grad_norm": 0.392578125,
"learning_rate": 7.614665424214193e-05,
"loss": 2.3754,
"step": 620
},
{
"epoch": 0.621,
"grad_norm": 0.376953125,
"learning_rate": 7.580781044003324e-05,
"loss": 2.3817,
"step": 621
},
{
"epoch": 0.622,
"grad_norm": 0.3984375,
"learning_rate": 7.546926141211974e-05,
"loss": 2.4627,
"step": 622
},
{
"epoch": 0.623,
"grad_norm": 0.4140625,
"learning_rate": 7.513101128351454e-05,
"loss": 2.4366,
"step": 623
},
{
"epoch": 0.624,
"grad_norm": 0.390625,
"learning_rate": 7.479306417568864e-05,
"loss": 2.5163,
"step": 624
},
{
"epoch": 0.625,
"grad_norm": 20.375,
"learning_rate": 7.445542420642097e-05,
"loss": 2.4236,
"step": 625
},
{
"epoch": 0.626,
"grad_norm": 0.37890625,
"learning_rate": 7.411809548974792e-05,
"loss": 2.4004,
"step": 626
},
{
"epoch": 0.627,
"grad_norm": 0.388671875,
"learning_rate": 7.378108213591355e-05,
"loss": 2.3935,
"step": 627
},
{
"epoch": 0.628,
"grad_norm": 0.408203125,
"learning_rate": 7.344438825131911e-05,
"loss": 2.4508,
"step": 628
},
{
"epoch": 0.629,
"grad_norm": 0.37890625,
"learning_rate": 7.310801793847344e-05,
"loss": 2.4123,
"step": 629
},
{
"epoch": 0.63,
"grad_norm": 0.38671875,
"learning_rate": 7.277197529594257e-05,
"loss": 2.3306,
"step": 630
},
{
"epoch": 0.631,
"grad_norm": 0.40234375,
"learning_rate": 7.243626441830009e-05,
"loss": 2.3514,
"step": 631
},
{
"epoch": 0.632,
"grad_norm": 0.39453125,
"learning_rate": 7.210088939607708e-05,
"loss": 2.4246,
"step": 632
},
{
"epoch": 0.633,
"grad_norm": 0.39453125,
"learning_rate": 7.176585431571235e-05,
"loss": 2.4561,
"step": 633
},
{
"epoch": 0.634,
"grad_norm": 0.390625,
"learning_rate": 7.143116325950265e-05,
"loss": 2.4402,
"step": 634
},
{
"epoch": 0.635,
"grad_norm": 0.373046875,
"learning_rate": 7.109682030555283e-05,
"loss": 2.4464,
"step": 635
},
{
"epoch": 0.636,
"grad_norm": 0.37890625,
"learning_rate": 7.076282952772633e-05,
"loss": 2.4168,
"step": 636
},
{
"epoch": 0.637,
"grad_norm": 0.369140625,
"learning_rate": 7.042919499559537e-05,
"loss": 2.4331,
"step": 637
},
{
"epoch": 0.638,
"grad_norm": 0.3828125,
"learning_rate": 7.009592077439134e-05,
"loss": 2.393,
"step": 638
},
{
"epoch": 0.639,
"grad_norm": 0.380859375,
"learning_rate": 6.976301092495556e-05,
"loss": 2.4643,
"step": 639
},
{
"epoch": 0.64,
"grad_norm": 0.404296875,
"learning_rate": 6.943046950368944e-05,
"loss": 2.3915,
"step": 640
},
{
"epoch": 0.641,
"grad_norm": 0.388671875,
"learning_rate": 6.909830056250527e-05,
"loss": 2.3675,
"step": 641
},
{
"epoch": 0.642,
"grad_norm": 0.369140625,
"learning_rate": 6.876650814877674e-05,
"loss": 2.4692,
"step": 642
},
{
"epoch": 0.643,
"grad_norm": 0.37890625,
"learning_rate": 6.843509630528977e-05,
"loss": 2.4107,
"step": 643
},
{
"epoch": 0.644,
"grad_norm": 0.369140625,
"learning_rate": 6.8104069070193e-05,
"loss": 2.356,
"step": 644
},
{
"epoch": 0.645,
"grad_norm": 0.3828125,
"learning_rate": 6.77734304769489e-05,
"loss": 2.3564,
"step": 645
},
{
"epoch": 0.646,
"grad_norm": 0.875,
"learning_rate": 6.744318455428436e-05,
"loss": 2.3626,
"step": 646
},
{
"epoch": 0.647,
"grad_norm": 0.38671875,
"learning_rate": 6.711333532614168e-05,
"loss": 2.4191,
"step": 647
},
{
"epoch": 0.648,
"grad_norm": 0.390625,
"learning_rate": 6.67838868116297e-05,
"loss": 2.4287,
"step": 648
},
{
"epoch": 0.649,
"grad_norm": 0.375,
"learning_rate": 6.64548430249745e-05,
"loss": 2.4516,
"step": 649
},
{
"epoch": 0.65,
"grad_norm": 0.3984375,
"learning_rate": 6.612620797547087e-05,
"loss": 2.461,
"step": 650
},
{
"epoch": 0.651,
"grad_norm": 0.38671875,
"learning_rate": 6.579798566743314e-05,
"loss": 2.4782,
"step": 651
},
{
"epoch": 0.652,
"grad_norm": 0.3828125,
"learning_rate": 6.547018010014654e-05,
"loss": 2.384,
"step": 652
},
{
"epoch": 0.653,
"grad_norm": 0.408203125,
"learning_rate": 6.51427952678185e-05,
"loss": 2.486,
"step": 653
},
{
"epoch": 0.654,
"grad_norm": 0.388671875,
"learning_rate": 6.481583515952983e-05,
"loss": 2.38,
"step": 654
},
{
"epoch": 0.655,
"grad_norm": 0.38671875,
"learning_rate": 6.448930375918631e-05,
"loss": 2.4087,
"step": 655
},
{
"epoch": 0.656,
"grad_norm": 0.373046875,
"learning_rate": 6.416320504546997e-05,
"loss": 2.3902,
"step": 656
},
{
"epoch": 0.657,
"grad_norm": 0.392578125,
"learning_rate": 6.383754299179079e-05,
"loss": 2.4243,
"step": 657
},
{
"epoch": 0.658,
"grad_norm": 0.38671875,
"learning_rate": 6.351232156623803e-05,
"loss": 2.4313,
"step": 658
},
{
"epoch": 0.659,
"grad_norm": 0.39453125,
"learning_rate": 6.318754473153221e-05,
"loss": 2.4931,
"step": 659
},
{
"epoch": 0.66,
"grad_norm": 0.3671875,
"learning_rate": 6.286321644497655e-05,
"loss": 2.4406,
"step": 660
},
{
"epoch": 0.661,
"grad_norm": 0.37109375,
"learning_rate": 6.25393406584088e-05,
"loss": 2.357,
"step": 661
},
{
"epoch": 0.662,
"grad_norm": 0.369140625,
"learning_rate": 6.22159213181533e-05,
"loss": 2.4506,
"step": 662
},
{
"epoch": 0.663,
"grad_norm": 0.388671875,
"learning_rate": 6.18929623649726e-05,
"loss": 2.3841,
"step": 663
},
{
"epoch": 0.664,
"grad_norm": 0.3671875,
"learning_rate": 6.157046773401964e-05,
"loss": 2.4144,
"step": 664
},
{
"epoch": 0.665,
"grad_norm": 0.37109375,
"learning_rate": 6.12484413547897e-05,
"loss": 2.46,
"step": 665
},
{
"epoch": 0.666,
"grad_norm": 0.3828125,
"learning_rate": 6.092688715107264e-05,
"loss": 2.4667,
"step": 666
},
{
"epoch": 0.667,
"grad_norm": 0.62109375,
"learning_rate": 6.0605809040904894e-05,
"loss": 2.4011,
"step": 667
},
{
"epoch": 0.668,
"grad_norm": 0.384765625,
"learning_rate": 6.0285210936521955e-05,
"loss": 2.4369,
"step": 668
},
{
"epoch": 0.669,
"grad_norm": 0.375,
"learning_rate": 5.9965096744310526e-05,
"loss": 2.4847,
"step": 669
},
{
"epoch": 0.67,
"grad_norm": 0.369140625,
"learning_rate": 5.964547036476099e-05,
"loss": 2.3905,
"step": 670
},
{
"epoch": 0.671,
"grad_norm": 0.384765625,
"learning_rate": 5.9326335692419995e-05,
"loss": 2.4256,
"step": 671
},
{
"epoch": 0.672,
"grad_norm": 0.375,
"learning_rate": 5.900769661584272e-05,
"loss": 2.3908,
"step": 672
},
{
"epoch": 0.673,
"grad_norm": 0.37109375,
"learning_rate": 5.868955701754584e-05,
"loss": 2.4197,
"step": 673
},
{
"epoch": 0.674,
"grad_norm": 0.36328125,
"learning_rate": 5.83719207739599e-05,
"loss": 2.3769,
"step": 674
},
{
"epoch": 0.675,
"grad_norm": 0.357421875,
"learning_rate": 5.805479175538229e-05,
"loss": 2.4318,
"step": 675
},
{
"epoch": 0.676,
"grad_norm": 0.3671875,
"learning_rate": 5.773817382593008e-05,
"loss": 2.3903,
"step": 676
},
{
"epoch": 0.677,
"grad_norm": 0.365234375,
"learning_rate": 5.7422070843492734e-05,
"loss": 2.4176,
"step": 677
},
{
"epoch": 0.678,
"grad_norm": 0.365234375,
"learning_rate": 5.710648665968543e-05,
"loss": 2.3889,
"step": 678
},
{
"epoch": 0.679,
"grad_norm": 0.37109375,
"learning_rate": 5.679142511980175e-05,
"loss": 2.4607,
"step": 679
},
{
"epoch": 0.68,
"grad_norm": 0.384765625,
"learning_rate": 5.647689006276726e-05,
"loss": 2.5372,
"step": 680
},
{
"epoch": 0.681,
"grad_norm": 0.384765625,
"learning_rate": 5.616288532109225e-05,
"loss": 2.5151,
"step": 681
},
{
"epoch": 0.682,
"grad_norm": 0.384765625,
"learning_rate": 5.584941472082549e-05,
"loss": 2.4688,
"step": 682
},
{
"epoch": 0.683,
"grad_norm": 0.392578125,
"learning_rate": 5.553648208150728e-05,
"loss": 2.4693,
"step": 683
},
{
"epoch": 0.684,
"grad_norm": 0.376953125,
"learning_rate": 5.522409121612304e-05,
"loss": 2.4525,
"step": 684
},
{
"epoch": 0.685,
"grad_norm": 0.373046875,
"learning_rate": 5.491224593105695e-05,
"loss": 2.4313,
"step": 685
},
{
"epoch": 0.686,
"grad_norm": 0.40234375,
"learning_rate": 5.4600950026045326e-05,
"loss": 2.4354,
"step": 686
},
{
"epoch": 0.687,
"grad_norm": 0.375,
"learning_rate": 5.4290207294130615e-05,
"loss": 2.4878,
"step": 687
},
{
"epoch": 0.688,
"grad_norm": 0.7890625,
"learning_rate": 5.398002152161484e-05,
"loss": 2.4165,
"step": 688
},
{
"epoch": 0.689,
"grad_norm": 0.390625,
"learning_rate": 5.3670396488013854e-05,
"loss": 2.5034,
"step": 689
},
{
"epoch": 0.69,
"grad_norm": 0.3828125,
"learning_rate": 5.33613359660109e-05,
"loss": 2.4738,
"step": 690
},
{
"epoch": 0.691,
"grad_norm": 0.38671875,
"learning_rate": 5.305284372141095e-05,
"loss": 2.4331,
"step": 691
},
{
"epoch": 0.692,
"grad_norm": 0.41796875,
"learning_rate": 5.274492351309461e-05,
"loss": 2.4987,
"step": 692
},
{
"epoch": 0.693,
"grad_norm": 0.361328125,
"learning_rate": 5.243757909297247e-05,
"loss": 2.4656,
"step": 693
},
{
"epoch": 0.694,
"grad_norm": 0.3671875,
"learning_rate": 5.213081420593933e-05,
"loss": 2.5297,
"step": 694
},
{
"epoch": 0.695,
"grad_norm": 0.412109375,
"learning_rate": 5.182463258982846e-05,
"loss": 2.4327,
"step": 695
},
{
"epoch": 0.696,
"grad_norm": 0.392578125,
"learning_rate": 5.15190379753663e-05,
"loss": 2.4265,
"step": 696
},
{
"epoch": 0.697,
"grad_norm": 0.376953125,
"learning_rate": 5.121403408612672e-05,
"loss": 2.4845,
"step": 697
},
{
"epoch": 0.698,
"grad_norm": 0.38671875,
"learning_rate": 5.090962463848592e-05,
"loss": 2.4739,
"step": 698
},
{
"epoch": 0.699,
"grad_norm": 0.375,
"learning_rate": 5.0605813341576924e-05,
"loss": 2.5051,
"step": 699
},
{
"epoch": 0.7,
"grad_norm": 0.39453125,
"learning_rate": 5.0302603897244474e-05,
"loss": 2.4842,
"step": 700
},
{
"epoch": 0.701,
"grad_norm": 0.38671875,
"learning_rate": 5.000000000000002e-05,
"loss": 2.4546,
"step": 701
},
{
"epoch": 0.702,
"grad_norm": 0.38671875,
"learning_rate": 4.969800533697649e-05,
"loss": 2.5519,
"step": 702
},
{
"epoch": 0.703,
"grad_norm": 0.3828125,
"learning_rate": 4.939662358788364e-05,
"loss": 2.5172,
"step": 703
},
{
"epoch": 0.704,
"grad_norm": 0.392578125,
"learning_rate": 4.909585842496287e-05,
"loss": 2.4724,
"step": 704
},
{
"epoch": 0.705,
"grad_norm": 0.384765625,
"learning_rate": 4.8795713512942865e-05,
"loss": 2.3813,
"step": 705
},
{
"epoch": 0.706,
"grad_norm": 0.3671875,
"learning_rate": 4.8496192508994576e-05,
"loss": 2.529,
"step": 706
},
{
"epoch": 0.707,
"grad_norm": 0.357421875,
"learning_rate": 4.8197299062686995e-05,
"loss": 2.5233,
"step": 707
},
{
"epoch": 0.708,
"grad_norm": 0.365234375,
"learning_rate": 4.78990368159424e-05,
"loss": 2.4315,
"step": 708
},
{
"epoch": 0.709,
"grad_norm": 3.15625,
"learning_rate": 4.7601409402992106e-05,
"loss": 2.3629,
"step": 709
},
{
"epoch": 0.71,
"grad_norm": 0.6015625,
"learning_rate": 4.7304420450332244e-05,
"loss": 2.5188,
"step": 710
},
{
"epoch": 0.711,
"grad_norm": 0.359375,
"learning_rate": 4.700807357667952e-05,
"loss": 2.4445,
"step": 711
},
{
"epoch": 0.712,
"grad_norm": 0.37109375,
"learning_rate": 4.6712372392927e-05,
"loss": 2.4884,
"step": 712
},
{
"epoch": 0.713,
"grad_norm": 0.400390625,
"learning_rate": 4.6417320502100316e-05,
"loss": 2.4909,
"step": 713
},
{
"epoch": 0.714,
"grad_norm": 0.36328125,
"learning_rate": 4.612292149931369e-05,
"loss": 2.497,
"step": 714
},
{
"epoch": 0.715,
"grad_norm": 0.357421875,
"learning_rate": 4.582917897172603e-05,
"loss": 2.4811,
"step": 715
},
{
"epoch": 0.716,
"grad_norm": 0.361328125,
"learning_rate": 4.5536096498497295e-05,
"loss": 2.4532,
"step": 716
},
{
"epoch": 0.717,
"grad_norm": 0.375,
"learning_rate": 4.524367765074499e-05,
"loss": 2.4833,
"step": 717
},
{
"epoch": 0.718,
"grad_norm": 0.357421875,
"learning_rate": 4.495192599150044e-05,
"loss": 2.426,
"step": 718
},
{
"epoch": 0.719,
"grad_norm": 0.37109375,
"learning_rate": 4.46608450756656e-05,
"loss": 2.4621,
"step": 719
},
{
"epoch": 0.72,
"grad_norm": 0.490234375,
"learning_rate": 4.437043844996952e-05,
"loss": 2.4483,
"step": 720
},
{
"epoch": 0.721,
"grad_norm": 0.359375,
"learning_rate": 4.4080709652925336e-05,
"loss": 2.45,
"step": 721
},
{
"epoch": 0.722,
"grad_norm": 0.345703125,
"learning_rate": 4.379166221478697e-05,
"loss": 2.4183,
"step": 722
},
{
"epoch": 0.723,
"grad_norm": 0.35546875,
"learning_rate": 4.350329965750621e-05,
"loss": 2.4723,
"step": 723
},
{
"epoch": 0.724,
"grad_norm": 0.375,
"learning_rate": 4.32156254946899e-05,
"loss": 2.418,
"step": 724
},
{
"epoch": 0.725,
"grad_norm": 0.392578125,
"learning_rate": 4.2928643231556844e-05,
"loss": 2.4797,
"step": 725
},
{
"epoch": 0.726,
"grad_norm": 0.361328125,
"learning_rate": 4.264235636489542e-05,
"loss": 2.5008,
"step": 726
},
{
"epoch": 0.727,
"grad_norm": 0.373046875,
"learning_rate": 4.235676838302068e-05,
"loss": 2.5085,
"step": 727
},
{
"epoch": 0.728,
"grad_norm": 0.380859375,
"learning_rate": 4.207188276573214e-05,
"loss": 2.4161,
"step": 728
},
{
"epoch": 0.729,
"grad_norm": 0.365234375,
"learning_rate": 4.1787702984271074e-05,
"loss": 2.5006,
"step": 729
},
{
"epoch": 0.73,
"grad_norm": 2.65625,
"learning_rate": 4.150423250127845e-05,
"loss": 2.4602,
"step": 730
},
{
"epoch": 0.731,
"grad_norm": 0.357421875,
"learning_rate": 4.12214747707527e-05,
"loss": 2.4451,
"step": 731
},
{
"epoch": 0.732,
"grad_norm": 0.35546875,
"learning_rate": 4.093943323800745e-05,
"loss": 2.469,
"step": 732
},
{
"epoch": 0.733,
"grad_norm": 0.359375,
"learning_rate": 4.065811133962987e-05,
"loss": 2.4601,
"step": 733
},
{
"epoch": 0.734,
"grad_norm": 0.361328125,
"learning_rate": 4.037751250343841e-05,
"loss": 2.3847,
"step": 734
},
{
"epoch": 0.735,
"grad_norm": 0.369140625,
"learning_rate": 4.009764014844143e-05,
"loss": 2.5012,
"step": 735
},
{
"epoch": 0.736,
"grad_norm": 0.369140625,
"learning_rate": 3.981849768479517e-05,
"loss": 2.4749,
"step": 736
},
{
"epoch": 0.737,
"grad_norm": 0.359375,
"learning_rate": 3.954008851376252e-05,
"loss": 2.3949,
"step": 737
},
{
"epoch": 0.738,
"grad_norm": 0.345703125,
"learning_rate": 3.9262416027671356e-05,
"loss": 2.4543,
"step": 738
},
{
"epoch": 0.739,
"grad_norm": 0.35546875,
"learning_rate": 3.8985483609873244e-05,
"loss": 2.491,
"step": 739
},
{
"epoch": 0.74,
"grad_norm": 0.36328125,
"learning_rate": 3.8709294634702376e-05,
"loss": 2.4547,
"step": 740
},
{
"epoch": 0.741,
"grad_norm": 0.359375,
"learning_rate": 3.843385246743417e-05,
"loss": 2.4588,
"step": 741
},
{
"epoch": 0.742,
"grad_norm": 0.37109375,
"learning_rate": 3.8159160464244606e-05,
"loss": 2.542,
"step": 742
},
{
"epoch": 0.743,
"grad_norm": 0.357421875,
"learning_rate": 3.788522197216897e-05,
"loss": 2.4396,
"step": 743
},
{
"epoch": 0.744,
"grad_norm": 0.34765625,
"learning_rate": 3.7612040329061405e-05,
"loss": 2.4729,
"step": 744
},
{
"epoch": 0.745,
"grad_norm": 0.36328125,
"learning_rate": 3.733961886355398e-05,
"loss": 2.4302,
"step": 745
},
{
"epoch": 0.746,
"grad_norm": 0.349609375,
"learning_rate": 3.7067960895016275e-05,
"loss": 2.3982,
"step": 746
},
{
"epoch": 0.747,
"grad_norm": 0.361328125,
"learning_rate": 3.679706973351491e-05,
"loss": 2.425,
"step": 747
},
{
"epoch": 0.748,
"grad_norm": 0.365234375,
"learning_rate": 3.6526948679773257e-05,
"loss": 2.5,
"step": 748
},
{
"epoch": 0.749,
"grad_norm": 0.3671875,
"learning_rate": 3.6257601025131026e-05,
"loss": 2.4297,
"step": 749
},
{
"epoch": 0.75,
"grad_norm": 0.921875,
"learning_rate": 3.5989030051504434e-05,
"loss": 2.5088,
"step": 750
},
{
"epoch": 0.75,
"eval_loss": 2.3882558345794678,
"eval_runtime": 23.5861,
"eval_samples_per_second": 21.199,
"eval_steps_per_second": 1.357,
"step": 750
},
{
"epoch": 0.751,
"grad_norm": 0.3515625,
"learning_rate": 3.5721239031346066e-05,
"loss": 2.523,
"step": 751
},
{
"epoch": 0.752,
"grad_norm": 0.3359375,
"learning_rate": 3.545423122760493e-05,
"loss": 2.4527,
"step": 752
},
{
"epoch": 0.753,
"grad_norm": 0.359375,
"learning_rate": 3.518800989368691e-05,
"loss": 2.4095,
"step": 753
},
{
"epoch": 0.754,
"grad_norm": 0.359375,
"learning_rate": 3.492257827341492e-05,
"loss": 2.4962,
"step": 754
},
{
"epoch": 0.755,
"grad_norm": 0.34765625,
"learning_rate": 3.465793960098945e-05,
"loss": 2.4922,
"step": 755
},
{
"epoch": 0.756,
"grad_norm": 0.37109375,
"learning_rate": 3.439409710094929e-05,
"loss": 2.5168,
"step": 756
},
{
"epoch": 0.757,
"grad_norm": 0.37109375,
"learning_rate": 3.413105398813195e-05,
"loss": 2.4681,
"step": 757
},
{
"epoch": 0.758,
"grad_norm": 0.33984375,
"learning_rate": 3.386881346763483e-05,
"loss": 2.386,
"step": 758
},
{
"epoch": 0.759,
"grad_norm": 0.361328125,
"learning_rate": 3.360737873477584e-05,
"loss": 2.5006,
"step": 759
},
{
"epoch": 0.76,
"grad_norm": 0.359375,
"learning_rate": 3.334675297505476e-05,
"loss": 2.4503,
"step": 760
},
{
"epoch": 0.761,
"grad_norm": 0.3671875,
"learning_rate": 3.308693936411421e-05,
"loss": 2.5063,
"step": 761
},
{
"epoch": 0.762,
"grad_norm": 0.35546875,
"learning_rate": 3.2827941067700996e-05,
"loss": 2.5382,
"step": 762
},
{
"epoch": 0.763,
"grad_norm": 0.40234375,
"learning_rate": 3.2569761241627696e-05,
"loss": 2.5106,
"step": 763
},
{
"epoch": 0.764,
"grad_norm": 0.361328125,
"learning_rate": 3.231240303173394e-05,
"loss": 2.4896,
"step": 764
},
{
"epoch": 0.765,
"grad_norm": 0.345703125,
"learning_rate": 3.205586957384838e-05,
"loss": 2.4733,
"step": 765
},
{
"epoch": 0.766,
"grad_norm": 0.353515625,
"learning_rate": 3.1800163993750166e-05,
"loss": 2.4365,
"step": 766
},
{
"epoch": 0.767,
"grad_norm": 0.35546875,
"learning_rate": 3.154528940713113e-05,
"loss": 2.4427,
"step": 767
},
{
"epoch": 0.768,
"grad_norm": 0.361328125,
"learning_rate": 3.129124891955771e-05,
"loss": 2.4716,
"step": 768
},
{
"epoch": 0.769,
"grad_norm": 0.349609375,
"learning_rate": 3.103804562643302e-05,
"loss": 2.4051,
"step": 769
},
{
"epoch": 0.77,
"grad_norm": 0.349609375,
"learning_rate": 3.078568261295933e-05,
"loss": 2.4672,
"step": 770
},
{
"epoch": 0.771,
"grad_norm": 1.6875,
"learning_rate": 3.053416295410026e-05,
"loss": 2.4865,
"step": 771
},
{
"epoch": 0.772,
"grad_norm": 0.3671875,
"learning_rate": 3.0283489714543556e-05,
"loss": 2.4112,
"step": 772
},
{
"epoch": 0.773,
"grad_norm": 0.35546875,
"learning_rate": 3.0033665948663448e-05,
"loss": 2.4911,
"step": 773
},
{
"epoch": 0.774,
"grad_norm": 0.376953125,
"learning_rate": 2.9784694700483762e-05,
"loss": 2.4914,
"step": 774
},
{
"epoch": 0.775,
"grad_norm": 0.361328125,
"learning_rate": 2.953657900364053e-05,
"loss": 2.4886,
"step": 775
},
{
"epoch": 0.776,
"grad_norm": 0.369140625,
"learning_rate": 2.9289321881345254e-05,
"loss": 2.4868,
"step": 776
},
{
"epoch": 0.777,
"grad_norm": 0.35546875,
"learning_rate": 2.904292634634793e-05,
"loss": 2.447,
"step": 777
},
{
"epoch": 0.778,
"grad_norm": 0.349609375,
"learning_rate": 2.879739540090036e-05,
"loss": 2.4644,
"step": 778
},
{
"epoch": 0.779,
"grad_norm": 0.35546875,
"learning_rate": 2.8552732036719687e-05,
"loss": 2.3846,
"step": 779
},
{
"epoch": 0.78,
"grad_norm": 0.353515625,
"learning_rate": 2.8308939234951726e-05,
"loss": 2.4676,
"step": 780
},
{
"epoch": 0.781,
"grad_norm": 0.3359375,
"learning_rate": 2.8066019966134904e-05,
"loss": 2.4526,
"step": 781
},
{
"epoch": 0.782,
"grad_norm": 0.3515625,
"learning_rate": 2.7823977190163786e-05,
"loss": 2.5223,
"step": 782
},
{
"epoch": 0.783,
"grad_norm": 0.33984375,
"learning_rate": 2.7582813856253275e-05,
"loss": 2.4876,
"step": 783
},
{
"epoch": 0.784,
"grad_norm": 0.337890625,
"learning_rate": 2.734253290290242e-05,
"loss": 2.4798,
"step": 784
},
{
"epoch": 0.785,
"grad_norm": 0.357421875,
"learning_rate": 2.7103137257858868e-05,
"loss": 2.4792,
"step": 785
},
{
"epoch": 0.786,
"grad_norm": 0.35546875,
"learning_rate": 2.6864629838082956e-05,
"loss": 2.4283,
"step": 786
},
{
"epoch": 0.787,
"grad_norm": 0.35546875,
"learning_rate": 2.6627013549712355e-05,
"loss": 2.4006,
"step": 787
},
{
"epoch": 0.788,
"grad_norm": 0.345703125,
"learning_rate": 2.639029128802657e-05,
"loss": 2.4266,
"step": 788
},
{
"epoch": 0.789,
"grad_norm": 0.341796875,
"learning_rate": 2.615446593741161e-05,
"loss": 2.4849,
"step": 789
},
{
"epoch": 0.79,
"grad_norm": 0.3515625,
"learning_rate": 2.5919540371325e-05,
"loss": 2.4477,
"step": 790
},
{
"epoch": 0.791,
"grad_norm": 0.341796875,
"learning_rate": 2.5685517452260567e-05,
"loss": 2.3939,
"step": 791
},
{
"epoch": 0.792,
"grad_norm": 1.265625,
"learning_rate": 2.5452400031713785e-05,
"loss": 2.4166,
"step": 792
},
{
"epoch": 0.793,
"grad_norm": 0.34765625,
"learning_rate": 2.5220190950146827e-05,
"loss": 2.4355,
"step": 793
},
{
"epoch": 0.794,
"grad_norm": 0.357421875,
"learning_rate": 2.4988893036954043e-05,
"loss": 2.4066,
"step": 794
},
{
"epoch": 0.795,
"grad_norm": 0.3515625,
"learning_rate": 2.4758509110427575e-05,
"loss": 2.4974,
"step": 795
},
{
"epoch": 0.796,
"grad_norm": 0.36328125,
"learning_rate": 2.45290419777228e-05,
"loss": 2.5034,
"step": 796
},
{
"epoch": 0.797,
"grad_norm": 0.353515625,
"learning_rate": 2.4300494434824373e-05,
"loss": 2.4447,
"step": 797
},
{
"epoch": 0.798,
"grad_norm": 0.36328125,
"learning_rate": 2.407286926651192e-05,
"loss": 2.516,
"step": 798
},
{
"epoch": 0.799,
"grad_norm": 0.34375,
"learning_rate": 2.3846169246326343e-05,
"loss": 2.4922,
"step": 799
},
{
"epoch": 0.8,
"grad_norm": 0.353515625,
"learning_rate": 2.362039713653581e-05,
"loss": 2.517,
"step": 800
},
{
"epoch": 0.801,
"grad_norm": 0.361328125,
"learning_rate": 2.339555568810221e-05,
"loss": 2.491,
"step": 801
},
{
"epoch": 0.802,
"grad_norm": 0.345703125,
"learning_rate": 2.3171647640647687e-05,
"loss": 2.431,
"step": 802
},
{
"epoch": 0.803,
"grad_norm": 0.34765625,
"learning_rate": 2.2948675722421086e-05,
"loss": 2.463,
"step": 803
},
{
"epoch": 0.804,
"grad_norm": 0.369140625,
"learning_rate": 2.2726642650264895e-05,
"loss": 2.4783,
"step": 804
},
{
"epoch": 0.805,
"grad_norm": 0.337890625,
"learning_rate": 2.2505551129582047e-05,
"loss": 2.4011,
"step": 805
},
{
"epoch": 0.806,
"grad_norm": 0.3515625,
"learning_rate": 2.2285403854302912e-05,
"loss": 2.4602,
"step": 806
},
{
"epoch": 0.807,
"grad_norm": 0.337890625,
"learning_rate": 2.2066203506852566e-05,
"loss": 2.4072,
"step": 807
},
{
"epoch": 0.808,
"grad_norm": 0.375,
"learning_rate": 2.1847952758118117e-05,
"loss": 2.4557,
"step": 808
},
{
"epoch": 0.809,
"grad_norm": 1.15625,
"learning_rate": 2.163065426741603e-05,
"loss": 2.4827,
"step": 809
},
{
"epoch": 0.81,
"grad_norm": 0.3359375,
"learning_rate": 2.1414310682459802e-05,
"loss": 2.4986,
"step": 810
},
{
"epoch": 0.811,
"grad_norm": 0.34765625,
"learning_rate": 2.119892463932781e-05,
"loss": 2.4235,
"step": 811
},
{
"epoch": 0.812,
"grad_norm": 0.373046875,
"learning_rate": 2.098449876243096e-05,
"loss": 2.433,
"step": 812
},
{
"epoch": 0.813,
"grad_norm": 2.125,
"learning_rate": 2.0771035664480942e-05,
"loss": 2.4635,
"step": 813
},
{
"epoch": 0.814,
"grad_norm": 0.3515625,
"learning_rate": 2.0558537946458177e-05,
"loss": 2.3816,
"step": 814
},
{
"epoch": 0.815,
"grad_norm": 0.34375,
"learning_rate": 2.0347008197580374e-05,
"loss": 2.3673,
"step": 815
},
{
"epoch": 0.816,
"grad_norm": 0.34375,
"learning_rate": 2.013644899527074e-05,
"loss": 2.4821,
"step": 816
},
{
"epoch": 0.817,
"grad_norm": 0.357421875,
"learning_rate": 1.9926862905126665e-05,
"loss": 2.4915,
"step": 817
},
{
"epoch": 0.818,
"grad_norm": 0.341796875,
"learning_rate": 1.9718252480888566e-05,
"loss": 2.5261,
"step": 818
},
{
"epoch": 0.819,
"grad_norm": 0.359375,
"learning_rate": 1.9510620264408596e-05,
"loss": 2.388,
"step": 819
},
{
"epoch": 0.82,
"grad_norm": 0.349609375,
"learning_rate": 1.930396878561983e-05,
"loss": 2.4732,
"step": 820
},
{
"epoch": 0.821,
"grad_norm": 0.353515625,
"learning_rate": 1.9098300562505266e-05,
"loss": 2.4026,
"step": 821
},
{
"epoch": 0.822,
"grad_norm": 0.341796875,
"learning_rate": 1.8893618101067355e-05,
"loss": 2.5404,
"step": 822
},
{
"epoch": 0.823,
"grad_norm": 0.36328125,
"learning_rate": 1.8689923895297245e-05,
"loss": 2.4978,
"step": 823
},
{
"epoch": 0.824,
"grad_norm": 0.390625,
"learning_rate": 1.848722042714457e-05,
"loss": 2.4942,
"step": 824
},
{
"epoch": 0.825,
"grad_norm": 0.34375,
"learning_rate": 1.8285510166487152e-05,
"loss": 2.4354,
"step": 825
},
{
"epoch": 0.826,
"grad_norm": 0.34765625,
"learning_rate": 1.808479557110081e-05,
"loss": 2.4651,
"step": 826
},
{
"epoch": 0.827,
"grad_norm": 0.337890625,
"learning_rate": 1.78850790866296e-05,
"loss": 2.4525,
"step": 827
},
{
"epoch": 0.828,
"grad_norm": 0.357421875,
"learning_rate": 1.7686363146555805e-05,
"loss": 2.5286,
"step": 828
},
{
"epoch": 0.829,
"grad_norm": 0.35546875,
"learning_rate": 1.7488650172170496e-05,
"loss": 2.4464,
"step": 829
},
{
"epoch": 0.83,
"grad_norm": 0.36328125,
"learning_rate": 1.7291942572543807e-05,
"loss": 2.4798,
"step": 830
},
{
"epoch": 0.831,
"grad_norm": 0.35546875,
"learning_rate": 1.7096242744495837e-05,
"loss": 2.4694,
"step": 831
},
{
"epoch": 0.832,
"grad_norm": 0.359375,
"learning_rate": 1.690155307256719e-05,
"loss": 2.4659,
"step": 832
},
{
"epoch": 0.833,
"grad_norm": 0.357421875,
"learning_rate": 1.6707875928990058e-05,
"loss": 2.4666,
"step": 833
},
{
"epoch": 0.834,
"grad_norm": 3.8125,
"learning_rate": 1.6515213673659357e-05,
"loss": 2.4691,
"step": 834
},
{
"epoch": 0.835,
"grad_norm": 0.3515625,
"learning_rate": 1.632356865410384e-05,
"loss": 2.4684,
"step": 835
},
{
"epoch": 0.836,
"grad_norm": 0.353515625,
"learning_rate": 1.6132943205457606e-05,
"loss": 2.5592,
"step": 836
},
{
"epoch": 0.837,
"grad_norm": 0.341796875,
"learning_rate": 1.5943339650431576e-05,
"loss": 2.4339,
"step": 837
},
{
"epoch": 0.838,
"grad_norm": 0.34765625,
"learning_rate": 1.5754760299285252e-05,
"loss": 2.5299,
"step": 838
},
{
"epoch": 0.839,
"grad_norm": 0.34375,
"learning_rate": 1.5567207449798515e-05,
"loss": 2.4162,
"step": 839
},
{
"epoch": 0.84,
"grad_norm": 0.353515625,
"learning_rate": 1.538068338724361e-05,
"loss": 2.4302,
"step": 840
},
{
"epoch": 0.841,
"grad_norm": 0.349609375,
"learning_rate": 1.5195190384357404e-05,
"loss": 2.491,
"step": 841
},
{
"epoch": 0.842,
"grad_norm": 0.384765625,
"learning_rate": 1.5010730701313625e-05,
"loss": 2.4681,
"step": 842
},
{
"epoch": 0.843,
"grad_norm": 0.34765625,
"learning_rate": 1.4827306585695234e-05,
"loss": 2.503,
"step": 843
},
{
"epoch": 0.844,
"grad_norm": 0.34375,
"learning_rate": 1.4644920272467244e-05,
"loss": 2.4634,
"step": 844
},
{
"epoch": 0.845,
"grad_norm": 0.345703125,
"learning_rate": 1.4463573983949341e-05,
"loss": 2.4758,
"step": 845
},
{
"epoch": 0.846,
"grad_norm": 0.3671875,
"learning_rate": 1.4283269929788779e-05,
"loss": 2.4392,
"step": 846
},
{
"epoch": 0.847,
"grad_norm": 0.33984375,
"learning_rate": 1.4104010306933557e-05,
"loss": 2.4774,
"step": 847
},
{
"epoch": 0.848,
"grad_norm": 0.345703125,
"learning_rate": 1.3925797299605647e-05,
"loss": 2.4524,
"step": 848
},
{
"epoch": 0.849,
"grad_norm": 0.34375,
"learning_rate": 1.3748633079274253e-05,
"loss": 2.4578,
"step": 849
},
{
"epoch": 0.85,
"grad_norm": 0.349609375,
"learning_rate": 1.3572519804629536e-05,
"loss": 2.4707,
"step": 850
},
{
"epoch": 0.851,
"grad_norm": 0.353515625,
"learning_rate": 1.339745962155613e-05,
"loss": 2.4744,
"step": 851
},
{
"epoch": 0.852,
"grad_norm": 0.349609375,
"learning_rate": 1.3223454663107172e-05,
"loss": 2.4554,
"step": 852
},
{
"epoch": 0.853,
"grad_norm": 0.359375,
"learning_rate": 1.30505070494781e-05,
"loss": 2.4564,
"step": 853
},
{
"epoch": 0.854,
"grad_norm": 0.349609375,
"learning_rate": 1.2878618887981064e-05,
"loss": 2.5088,
"step": 854
},
{
"epoch": 0.855,
"grad_norm": 0.51171875,
"learning_rate": 1.2707792273019048e-05,
"loss": 2.502,
"step": 855
},
{
"epoch": 0.856,
"grad_norm": 0.35546875,
"learning_rate": 1.2538029286060426e-05,
"loss": 2.4917,
"step": 856
},
{
"epoch": 0.857,
"grad_norm": 0.34375,
"learning_rate": 1.2369331995613665e-05,
"loss": 2.4712,
"step": 857
},
{
"epoch": 0.858,
"grad_norm": 0.34375,
"learning_rate": 1.2201702457201947e-05,
"loss": 2.4626,
"step": 858
},
{
"epoch": 0.859,
"grad_norm": 0.341796875,
"learning_rate": 1.2035142713338366e-05,
"loss": 2.5001,
"step": 859
},
{
"epoch": 0.86,
"grad_norm": 0.3359375,
"learning_rate": 1.1869654793500784e-05,
"loss": 2.4413,
"step": 860
},
{
"epoch": 0.861,
"grad_norm": 0.3359375,
"learning_rate": 1.1705240714107302e-05,
"loss": 2.4324,
"step": 861
},
{
"epoch": 0.862,
"grad_norm": 0.349609375,
"learning_rate": 1.1541902478491606e-05,
"loss": 2.4573,
"step": 862
},
{
"epoch": 0.863,
"grad_norm": 0.359375,
"learning_rate": 1.1379642076878527e-05,
"loss": 2.4696,
"step": 863
},
{
"epoch": 0.864,
"grad_norm": 0.35546875,
"learning_rate": 1.1218461486359877e-05,
"loss": 2.5201,
"step": 864
},
{
"epoch": 0.865,
"grad_norm": 0.361328125,
"learning_rate": 1.1058362670870249e-05,
"loss": 2.4874,
"step": 865
},
{
"epoch": 0.866,
"grad_norm": 0.35546875,
"learning_rate": 1.0899347581163221e-05,
"loss": 2.4856,
"step": 866
},
{
"epoch": 0.867,
"grad_norm": 0.33984375,
"learning_rate": 1.0741418154787442e-05,
"loss": 2.45,
"step": 867
},
{
"epoch": 0.868,
"grad_norm": 0.359375,
"learning_rate": 1.0584576316063188e-05,
"loss": 2.5455,
"step": 868
},
{
"epoch": 0.869,
"grad_norm": 0.34765625,
"learning_rate": 1.042882397605871e-05,
"loss": 2.4775,
"step": 869
},
{
"epoch": 0.87,
"grad_norm": 0.349609375,
"learning_rate": 1.0274163032567163e-05,
"loss": 2.4642,
"step": 870
},
{
"epoch": 0.871,
"grad_norm": 0.359375,
"learning_rate": 1.0120595370083318e-05,
"loss": 2.4814,
"step": 871
},
{
"epoch": 0.872,
"grad_norm": 0.353515625,
"learning_rate": 9.968122859780648e-06,
"loss": 2.4842,
"step": 872
},
{
"epoch": 0.873,
"grad_norm": 0.361328125,
"learning_rate": 9.816747359488632e-06,
"loss": 2.4489,
"step": 873
},
{
"epoch": 0.874,
"grad_norm": 0.365234375,
"learning_rate": 9.666470713669918e-06,
"loss": 2.4983,
"step": 874
},
{
"epoch": 0.875,
"grad_norm": 0.8046875,
"learning_rate": 9.517294753398064e-06,
"loss": 2.383,
"step": 875
},
{
"epoch": 0.876,
"grad_norm": 0.349609375,
"learning_rate": 9.369221296335006e-06,
"loss": 2.5233,
"step": 876
},
{
"epoch": 0.877,
"grad_norm": 0.3515625,
"learning_rate": 9.222252146709142e-06,
"loss": 2.4413,
"step": 877
},
{
"epoch": 0.878,
"grad_norm": 0.33984375,
"learning_rate": 9.076389095293148e-06,
"loss": 2.5355,
"step": 878
},
{
"epoch": 0.879,
"grad_norm": 0.353515625,
"learning_rate": 8.931633919382298e-06,
"loss": 2.4576,
"step": 879
},
{
"epoch": 0.88,
"grad_norm": 0.3359375,
"learning_rate": 8.787988382772705e-06,
"loss": 2.5046,
"step": 880
},
{
"epoch": 0.881,
"grad_norm": 0.359375,
"learning_rate": 8.645454235739903e-06,
"loss": 2.4818,
"step": 881
},
{
"epoch": 0.882,
"grad_norm": 0.341796875,
"learning_rate": 8.504033215017527e-06,
"loss": 2.5373,
"step": 882
},
{
"epoch": 0.883,
"grad_norm": 0.333984375,
"learning_rate": 8.363727043776038e-06,
"loss": 2.4601,
"step": 883
},
{
"epoch": 0.884,
"grad_norm": 0.33984375,
"learning_rate": 8.224537431601886e-06,
"loss": 2.5439,
"step": 884
},
{
"epoch": 0.885,
"grad_norm": 0.33984375,
"learning_rate": 8.086466074476563e-06,
"loss": 2.4206,
"step": 885
},
{
"epoch": 0.886,
"grad_norm": 0.34375,
"learning_rate": 7.949514654755962e-06,
"loss": 2.4462,
"step": 886
},
{
"epoch": 0.887,
"grad_norm": 0.345703125,
"learning_rate": 7.81368484114996e-06,
"loss": 2.4666,
"step": 887
},
{
"epoch": 0.888,
"grad_norm": 0.3671875,
"learning_rate": 7.67897828870191e-06,
"loss": 2.4672,
"step": 888
},
{
"epoch": 0.889,
"grad_norm": 0.35546875,
"learning_rate": 7.545396638768698e-06,
"loss": 2.5082,
"step": 889
},
{
"epoch": 0.89,
"grad_norm": 0.341796875,
"learning_rate": 7.412941519000527e-06,
"loss": 2.4445,
"step": 890
},
{
"epoch": 0.891,
"grad_norm": 0.345703125,
"learning_rate": 7.281614543321269e-06,
"loss": 2.4783,
"step": 891
},
{
"epoch": 0.892,
"grad_norm": 0.345703125,
"learning_rate": 7.151417311908648e-06,
"loss": 2.4691,
"step": 892
},
{
"epoch": 0.893,
"grad_norm": 0.34765625,
"learning_rate": 7.022351411174866e-06,
"loss": 2.5065,
"step": 893
},
{
"epoch": 0.894,
"grad_norm": 0.353515625,
"learning_rate": 6.894418413747183e-06,
"loss": 2.4405,
"step": 894
},
{
"epoch": 0.895,
"grad_norm": 0.34765625,
"learning_rate": 6.767619878448783e-06,
"loss": 2.4714,
"step": 895
},
{
"epoch": 0.896,
"grad_norm": 1.0625,
"learning_rate": 6.6419573502798374e-06,
"loss": 2.4901,
"step": 896
},
{
"epoch": 0.897,
"grad_norm": 0.341796875,
"learning_rate": 6.517432360398556e-06,
"loss": 2.5167,
"step": 897
},
{
"epoch": 0.898,
"grad_norm": 0.359375,
"learning_rate": 6.394046426102674e-06,
"loss": 2.4734,
"step": 898
},
{
"epoch": 0.899,
"grad_norm": 0.353515625,
"learning_rate": 6.2718010508108545e-06,
"loss": 2.4281,
"step": 899
},
{
"epoch": 0.9,
"grad_norm": 0.3515625,
"learning_rate": 6.1506977240444074e-06,
"loss": 2.4806,
"step": 900
},
{
"epoch": 0.901,
"grad_norm": 0.34375,
"learning_rate": 6.030737921409169e-06,
"loss": 2.4104,
"step": 901
},
{
"epoch": 0.902,
"grad_norm": 0.349609375,
"learning_rate": 5.911923104577455e-06,
"loss": 2.4529,
"step": 902
},
{
"epoch": 0.903,
"grad_norm": 0.36328125,
"learning_rate": 5.7942547212703315e-06,
"loss": 2.5531,
"step": 903
},
{
"epoch": 0.904,
"grad_norm": 0.34375,
"learning_rate": 5.6777342052399045e-06,
"loss": 2.4894,
"step": 904
},
{
"epoch": 0.905,
"grad_norm": 0.345703125,
"learning_rate": 5.562362976251901e-06,
"loss": 2.4596,
"step": 905
},
{
"epoch": 0.906,
"grad_norm": 0.34765625,
"learning_rate": 5.448142440068316e-06,
"loss": 2.5246,
"step": 906
},
{
"epoch": 0.907,
"grad_norm": 0.3359375,
"learning_rate": 5.335073988430372e-06,
"loss": 2.521,
"step": 907
},
{
"epoch": 0.908,
"grad_norm": 0.3515625,
"learning_rate": 5.223158999041444e-06,
"loss": 2.4798,
"step": 908
},
{
"epoch": 0.909,
"grad_norm": 0.357421875,
"learning_rate": 5.1123988355503475e-06,
"loss": 2.5506,
"step": 909
},
{
"epoch": 0.91,
"grad_norm": 0.357421875,
"learning_rate": 5.002794847534764e-06,
"loss": 2.4723,
"step": 910
},
{
"epoch": 0.911,
"grad_norm": 0.333984375,
"learning_rate": 4.8943483704846475e-06,
"loss": 2.5085,
"step": 911
},
{
"epoch": 0.912,
"grad_norm": 0.37109375,
"learning_rate": 4.7870607257861415e-06,
"loss": 2.4041,
"step": 912
},
{
"epoch": 0.913,
"grad_norm": 0.345703125,
"learning_rate": 4.680933220705308e-06,
"loss": 2.4633,
"step": 913
},
{
"epoch": 0.914,
"grad_norm": 0.3359375,
"learning_rate": 4.575967148372317e-06,
"loss": 2.528,
"step": 914
},
{
"epoch": 0.915,
"grad_norm": 0.333984375,
"learning_rate": 4.4721637877656375e-06,
"loss": 2.4561,
"step": 915
},
{
"epoch": 0.916,
"grad_norm": 0.416015625,
"learning_rate": 4.369524403696457e-06,
"loss": 2.4552,
"step": 916
},
{
"epoch": 0.917,
"grad_norm": 0.470703125,
"learning_rate": 4.268050246793276e-06,
"loss": 2.4971,
"step": 917
},
{
"epoch": 0.918,
"grad_norm": 0.341796875,
"learning_rate": 4.167742553486675e-06,
"loss": 2.4616,
"step": 918
},
{
"epoch": 0.919,
"grad_norm": 0.3515625,
"learning_rate": 4.068602545994249e-06,
"loss": 2.4429,
"step": 919
},
{
"epoch": 0.92,
"grad_norm": 0.337890625,
"learning_rate": 3.970631432305694e-06,
"loss": 2.4485,
"step": 920
},
{
"epoch": 0.921,
"grad_norm": 0.3359375,
"learning_rate": 3.873830406168111e-06,
"loss": 2.5458,
"step": 921
},
{
"epoch": 0.922,
"grad_norm": 0.34375,
"learning_rate": 3.7782006470714616e-06,
"loss": 2.5266,
"step": 922
},
{
"epoch": 0.923,
"grad_norm": 0.337890625,
"learning_rate": 3.68374332023419e-06,
"loss": 2.479,
"step": 923
},
{
"epoch": 0.924,
"grad_norm": 0.33203125,
"learning_rate": 3.5904595765890005e-06,
"loss": 2.4356,
"step": 924
},
{
"epoch": 0.925,
"grad_norm": 0.3515625,
"learning_rate": 3.4983505527688586e-06,
"loss": 2.4789,
"step": 925
},
{
"epoch": 0.926,
"grad_norm": 0.341796875,
"learning_rate": 3.40741737109318e-06,
"loss": 2.4874,
"step": 926
},
{
"epoch": 0.927,
"grad_norm": 0.361328125,
"learning_rate": 3.3176611395540626e-06,
"loss": 2.5194,
"step": 927
},
{
"epoch": 0.928,
"grad_norm": 0.349609375,
"learning_rate": 3.2290829518028862e-06,
"loss": 2.4987,
"step": 928
},
{
"epoch": 0.929,
"grad_norm": 0.34765625,
"learning_rate": 3.1416838871368924e-06,
"loss": 2.4636,
"step": 929
},
{
"epoch": 0.93,
"grad_norm": 0.345703125,
"learning_rate": 3.0554650104861136e-06,
"loss": 2.4595,
"step": 930
},
{
"epoch": 0.931,
"grad_norm": 0.357421875,
"learning_rate": 2.970427372400353e-06,
"loss": 2.4153,
"step": 931
},
{
"epoch": 0.932,
"grad_norm": 0.349609375,
"learning_rate": 2.8865720090364034e-06,
"loss": 2.5477,
"step": 932
},
{
"epoch": 0.933,
"grad_norm": 0.345703125,
"learning_rate": 2.8038999421453826e-06,
"loss": 2.4451,
"step": 933
},
{
"epoch": 0.934,
"grad_norm": 0.328125,
"learning_rate": 2.7224121790603517e-06,
"loss": 2.4521,
"step": 934
},
{
"epoch": 0.935,
"grad_norm": 0.34765625,
"learning_rate": 2.6421097126839712e-06,
"loss": 2.4565,
"step": 935
},
{
"epoch": 0.936,
"grad_norm": 0.341796875,
"learning_rate": 2.5629935214764865e-06,
"loss": 2.4708,
"step": 936
},
{
"epoch": 0.937,
"grad_norm": 0.361328125,
"learning_rate": 2.4850645694436736e-06,
"loss": 2.4826,
"step": 937
},
{
"epoch": 0.938,
"grad_norm": 2.703125,
"learning_rate": 2.4083238061252567e-06,
"loss": 2.4774,
"step": 938
},
{
"epoch": 0.939,
"grad_norm": 0.375,
"learning_rate": 2.332772166583208e-06,
"loss": 2.4619,
"step": 939
},
{
"epoch": 0.94,
"grad_norm": 0.345703125,
"learning_rate": 2.2584105713904125e-06,
"loss": 2.5096,
"step": 940
},
{
"epoch": 0.941,
"grad_norm": 0.3515625,
"learning_rate": 2.1852399266194314e-06,
"loss": 2.4763,
"step": 941
},
{
"epoch": 0.942,
"grad_norm": 0.359375,
"learning_rate": 2.1132611238315003e-06,
"loss": 2.5113,
"step": 942
},
{
"epoch": 0.943,
"grad_norm": 0.33203125,
"learning_rate": 2.0424750400655947e-06,
"loss": 2.4475,
"step": 943
},
{
"epoch": 0.944,
"grad_norm": 0.36328125,
"learning_rate": 1.9728825378278246e-06,
"loss": 2.4407,
"step": 944
},
{
"epoch": 0.945,
"grad_norm": 0.34375,
"learning_rate": 1.904484465080847e-06,
"loss": 2.4143,
"step": 945
},
{
"epoch": 0.946,
"grad_norm": 0.359375,
"learning_rate": 1.8372816552336026e-06,
"loss": 2.4565,
"step": 946
},
{
"epoch": 0.947,
"grad_norm": 0.33984375,
"learning_rate": 1.771274927131139e-06,
"loss": 2.4452,
"step": 947
},
{
"epoch": 0.948,
"grad_norm": 0.349609375,
"learning_rate": 1.706465085044584e-06,
"loss": 2.4039,
"step": 948
},
{
"epoch": 0.949,
"grad_norm": 0.34375,
"learning_rate": 1.6428529186614195e-06,
"loss": 2.525,
"step": 949
},
{
"epoch": 0.95,
"grad_norm": 0.337890625,
"learning_rate": 1.580439203075812e-06,
"loss": 2.4337,
"step": 950
},
{
"epoch": 0.951,
"grad_norm": 0.359375,
"learning_rate": 1.5192246987791981e-06,
"loss": 2.4567,
"step": 951
},
{
"epoch": 0.952,
"grad_norm": 0.412109375,
"learning_rate": 1.4592101516509914e-06,
"loss": 2.4604,
"step": 952
},
{
"epoch": 0.953,
"grad_norm": 0.341796875,
"learning_rate": 1.400396292949513e-06,
"loss": 2.4341,
"step": 953
},
{
"epoch": 0.954,
"grad_norm": 0.34765625,
"learning_rate": 1.3427838393030633e-06,
"loss": 2.4539,
"step": 954
},
{
"epoch": 0.955,
"grad_norm": 0.34765625,
"learning_rate": 1.2863734927012095e-06,
"loss": 2.4511,
"step": 955
},
{
"epoch": 0.956,
"grad_norm": 0.349609375,
"learning_rate": 1.231165940486234e-06,
"loss": 2.4771,
"step": 956
},
{
"epoch": 0.957,
"grad_norm": 0.357421875,
"learning_rate": 1.1771618553447216e-06,
"loss": 2.51,
"step": 957
},
{
"epoch": 0.958,
"grad_norm": 0.353515625,
"learning_rate": 1.1243618952994195e-06,
"loss": 2.458,
"step": 958
},
{
"epoch": 0.959,
"grad_norm": 1.0078125,
"learning_rate": 1.0727667037011668e-06,
"loss": 2.4529,
"step": 959
},
{
"epoch": 0.96,
"grad_norm": 0.3515625,
"learning_rate": 1.0223769092211012e-06,
"loss": 2.5278,
"step": 960
},
{
"epoch": 0.961,
"grad_norm": 0.341796875,
"learning_rate": 9.731931258429638e-07,
"loss": 2.4614,
"step": 961
},
{
"epoch": 0.962,
"grad_norm": 0.37109375,
"learning_rate": 9.252159528556403e-07,
"loss": 2.486,
"step": 962
},
{
"epoch": 0.963,
"grad_norm": 0.34375,
"learning_rate": 8.784459748458318e-07,
"loss": 2.4673,
"step": 963
},
{
"epoch": 0.964,
"grad_norm": 0.349609375,
"learning_rate": 8.328837616909613e-07,
"loss": 2.523,
"step": 964
},
{
"epoch": 0.965,
"grad_norm": 0.333984375,
"learning_rate": 7.885298685522235e-07,
"loss": 2.4779,
"step": 965
},
{
"epoch": 0.966,
"grad_norm": 0.345703125,
"learning_rate": 7.453848358678017e-07,
"loss": 2.4554,
"step": 966
},
{
"epoch": 0.967,
"grad_norm": 0.345703125,
"learning_rate": 7.034491893463058e-07,
"loss": 2.4431,
"step": 967
},
{
"epoch": 0.968,
"grad_norm": 0.341796875,
"learning_rate": 6.627234399603555e-07,
"loss": 2.446,
"step": 968
},
{
"epoch": 0.969,
"grad_norm": 0.3515625,
"learning_rate": 6.232080839403631e-07,
"loss": 2.4309,
"step": 969
},
{
"epoch": 0.97,
"grad_norm": 0.341796875,
"learning_rate": 5.849036027684606e-07,
"loss": 2.4665,
"step": 970
},
{
"epoch": 0.971,
"grad_norm": 0.333984375,
"learning_rate": 5.478104631726711e-07,
"loss": 2.4196,
"step": 971
},
{
"epoch": 0.972,
"grad_norm": 0.3671875,
"learning_rate": 5.119291171211793e-07,
"loss": 2.5172,
"step": 972
},
{
"epoch": 0.973,
"grad_norm": 0.33984375,
"learning_rate": 4.772600018168816e-07,
"loss": 2.4503,
"step": 973
},
{
"epoch": 0.974,
"grad_norm": 0.3515625,
"learning_rate": 4.438035396920004e-07,
"loss": 2.5282,
"step": 974
},
{
"epoch": 0.975,
"grad_norm": 0.3515625,
"learning_rate": 4.115601384029666e-07,
"loss": 2.537,
"step": 975
},
{
"epoch": 0.976,
"grad_norm": 0.349609375,
"learning_rate": 3.805301908254455e-07,
"loss": 2.4862,
"step": 976
},
{
"epoch": 0.977,
"grad_norm": 0.345703125,
"learning_rate": 3.50714075049563e-07,
"loss": 2.5061,
"step": 977
},
{
"epoch": 0.978,
"grad_norm": 0.357421875,
"learning_rate": 3.2211215437528694e-07,
"loss": 2.5166,
"step": 978
},
{
"epoch": 0.979,
"grad_norm": 0.365234375,
"learning_rate": 2.947247773079753e-07,
"loss": 2.4817,
"step": 979
},
{
"epoch": 0.98,
"grad_norm": 0.91015625,
"learning_rate": 2.685522775541904e-07,
"loss": 2.4994,
"step": 980
},
{
"epoch": 0.981,
"grad_norm": 0.349609375,
"learning_rate": 2.4359497401758024e-07,
"loss": 2.4299,
"step": 981
},
{
"epoch": 0.982,
"grad_norm": 0.330078125,
"learning_rate": 2.1985317079500356e-07,
"loss": 2.5098,
"step": 982
},
{
"epoch": 0.983,
"grad_norm": 0.3515625,
"learning_rate": 1.973271571728441e-07,
"loss": 2.4577,
"step": 983
},
{
"epoch": 0.984,
"grad_norm": 0.345703125,
"learning_rate": 1.7601720762346897e-07,
"loss": 2.4706,
"step": 984
},
{
"epoch": 0.985,
"grad_norm": 0.35546875,
"learning_rate": 1.5592358180189782e-07,
"loss": 2.4612,
"step": 985
},
{
"epoch": 0.986,
"grad_norm": 0.34765625,
"learning_rate": 1.3704652454261668e-07,
"loss": 2.4683,
"step": 986
},
{
"epoch": 0.987,
"grad_norm": 0.34765625,
"learning_rate": 1.193862658566025e-07,
"loss": 2.4788,
"step": 987
},
{
"epoch": 0.988,
"grad_norm": 0.337890625,
"learning_rate": 1.0294302092853647e-07,
"loss": 2.4645,
"step": 988
},
{
"epoch": 0.989,
"grad_norm": 0.35546875,
"learning_rate": 8.771699011416168e-08,
"loss": 2.4886,
"step": 989
},
{
"epoch": 0.99,
"grad_norm": 0.345703125,
"learning_rate": 7.370835893788508e-08,
"loss": 2.4874,
"step": 990
},
{
"epoch": 0.991,
"grad_norm": 0.3515625,
"learning_rate": 6.09172980904238e-08,
"loss": 2.5821,
"step": 991
},
{
"epoch": 0.992,
"grad_norm": 0.341796875,
"learning_rate": 4.934396342684e-08,
"loss": 2.5404,
"step": 992
},
{
"epoch": 0.993,
"grad_norm": 0.375,
"learning_rate": 3.898849596456478e-08,
"loss": 2.4836,
"step": 993
},
{
"epoch": 0.994,
"grad_norm": 0.361328125,
"learning_rate": 2.985102188168831e-08,
"loss": 2.5449,
"step": 994
},
{
"epoch": 0.995,
"grad_norm": 0.349609375,
"learning_rate": 2.193165251545004e-08,
"loss": 2.4132,
"step": 995
},
{
"epoch": 0.996,
"grad_norm": 0.345703125,
"learning_rate": 1.5230484360873044e-08,
"loss": 2.476,
"step": 996
},
{
"epoch": 0.997,
"grad_norm": 0.3359375,
"learning_rate": 9.747599069576119e-09,
"loss": 2.4409,
"step": 997
},
{
"epoch": 0.998,
"grad_norm": 0.34375,
"learning_rate": 5.483063448785686e-09,
"loss": 2.4674,
"step": 998
},
{
"epoch": 0.999,
"grad_norm": 0.345703125,
"learning_rate": 2.4369294605253166e-09,
"loss": 2.5205,
"step": 999
},
{
"epoch": 1.0,
"grad_norm": 0.71484375,
"learning_rate": 6.092342209607083e-10,
"loss": 2.5571,
"step": 1000
},
{
"epoch": 1.0,
"eval_loss": 2.3852813243865967,
"eval_runtime": 23.571,
"eval_samples_per_second": 21.212,
"eval_steps_per_second": 1.358,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.81912770904064e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}