robometer-4b-full-threading-d0 / trainer_state.json
chomeed's picture
Full fine-tune of Robometer-4B on mimicgen threading_d0 (best ckpt, step 500)
8e0d74a verified
Raw
History Blame Contribute Delete
80.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.025,
"eval_steps": 100,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 5e-05,
"grad_norm": 122.16051483154297,
"learning_rate": 0.0,
"loss": 1.8961,
"step": 1
},
{
"epoch": 0.0001,
"grad_norm": 68.64327239990234,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.7336,
"step": 2
},
{
"epoch": 0.00015,
"grad_norm": 34.362892150878906,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.5657,
"step": 3
},
{
"epoch": 0.0002,
"grad_norm": 151.97335815429688,
"learning_rate": 6.000000000000001e-07,
"loss": 1.3111,
"step": 4
},
{
"epoch": 0.00025,
"grad_norm": 94.8849105834961,
"learning_rate": 8.000000000000001e-07,
"loss": 1.1928,
"step": 5
},
{
"epoch": 0.0003,
"grad_norm": 68.24564361572266,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.7754,
"step": 6
},
{
"epoch": 0.00035,
"grad_norm": 106.96430969238281,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.3711,
"step": 7
},
{
"epoch": 0.0004,
"grad_norm": 51.82062911987305,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.6345,
"step": 8
},
{
"epoch": 0.00045,
"grad_norm": 106.7470703125,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.4213,
"step": 9
},
{
"epoch": 0.0005,
"grad_norm": 63.274295806884766,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.612,
"step": 10
},
{
"epoch": 0.00055,
"grad_norm": 88.51773071289062,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.951,
"step": 11
},
{
"epoch": 0.0006,
"grad_norm": 120.4432373046875,
"learning_rate": 2.2e-06,
"loss": 1.3452,
"step": 12
},
{
"epoch": 0.00065,
"grad_norm": 68.23249053955078,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.594,
"step": 13
},
{
"epoch": 0.0007,
"grad_norm": 113.59532165527344,
"learning_rate": 2.6e-06,
"loss": 1.4108,
"step": 14
},
{
"epoch": 0.00075,
"grad_norm": 78.14495086669922,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.217,
"step": 15
},
{
"epoch": 0.0008,
"grad_norm": 71.44093322753906,
"learning_rate": 3e-06,
"loss": 0.9976,
"step": 16
},
{
"epoch": 0.00085,
"grad_norm": 109.85679626464844,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.9354,
"step": 17
},
{
"epoch": 0.0009,
"grad_norm": 166.82241821289062,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.7291,
"step": 18
},
{
"epoch": 0.00095,
"grad_norm": 132.03396606445312,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.8239,
"step": 19
},
{
"epoch": 0.001,
"grad_norm": 117.53083801269531,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.138,
"step": 20
},
{
"epoch": 0.00105,
"grad_norm": 131.66677856445312,
"learning_rate": 4.000000000000001e-06,
"loss": 1.407,
"step": 21
},
{
"epoch": 0.0011,
"grad_norm": 65.3564453125,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.3293,
"step": 22
},
{
"epoch": 0.00115,
"grad_norm": 61.042724609375,
"learning_rate": 4.4e-06,
"loss": 0.4492,
"step": 23
},
{
"epoch": 0.0012,
"grad_norm": 21.75531578063965,
"learning_rate": 4.600000000000001e-06,
"loss": 0.254,
"step": 24
},
{
"epoch": 0.00125,
"grad_norm": 100.22628021240234,
"learning_rate": 4.800000000000001e-06,
"loss": 1.13,
"step": 25
},
{
"epoch": 0.0013,
"grad_norm": 39.203861236572266,
"learning_rate": 5e-06,
"loss": 0.3655,
"step": 26
},
{
"epoch": 0.00135,
"grad_norm": 73.43659210205078,
"learning_rate": 5.2e-06,
"loss": 0.3976,
"step": 27
},
{
"epoch": 0.0014,
"grad_norm": 16.97100067138672,
"learning_rate": 5.400000000000001e-06,
"loss": 0.1936,
"step": 28
},
{
"epoch": 0.00145,
"grad_norm": 19.03237533569336,
"learning_rate": 5.600000000000001e-06,
"loss": 0.2026,
"step": 29
},
{
"epoch": 0.0015,
"grad_norm": 37.67095184326172,
"learning_rate": 5.8e-06,
"loss": 0.3037,
"step": 30
},
{
"epoch": 0.00155,
"grad_norm": 50.00007629394531,
"learning_rate": 6e-06,
"loss": 0.5524,
"step": 31
},
{
"epoch": 0.0016,
"grad_norm": 103.01881408691406,
"learning_rate": 6.200000000000001e-06,
"loss": 0.4411,
"step": 32
},
{
"epoch": 0.00165,
"grad_norm": 72.95085906982422,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.5604,
"step": 33
},
{
"epoch": 0.0017,
"grad_norm": 82.70050811767578,
"learning_rate": 6.600000000000001e-06,
"loss": 0.5243,
"step": 34
},
{
"epoch": 0.00175,
"grad_norm": 44.8000602722168,
"learning_rate": 6.800000000000001e-06,
"loss": 0.4454,
"step": 35
},
{
"epoch": 0.0018,
"grad_norm": 74.06373596191406,
"learning_rate": 7e-06,
"loss": 0.8086,
"step": 36
},
{
"epoch": 0.00185,
"grad_norm": 26.03671646118164,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.2051,
"step": 37
},
{
"epoch": 0.0019,
"grad_norm": 51.14949035644531,
"learning_rate": 7.4e-06,
"loss": 0.8955,
"step": 38
},
{
"epoch": 0.00195,
"grad_norm": 43.66708755493164,
"learning_rate": 7.600000000000001e-06,
"loss": 0.9189,
"step": 39
},
{
"epoch": 0.002,
"grad_norm": 9.864336967468262,
"learning_rate": 7.800000000000002e-06,
"loss": 0.1676,
"step": 40
},
{
"epoch": 0.00205,
"grad_norm": 73.50960540771484,
"learning_rate": 8.000000000000001e-06,
"loss": 0.4905,
"step": 41
},
{
"epoch": 0.0021,
"grad_norm": 21.008272171020508,
"learning_rate": 8.2e-06,
"loss": 0.1899,
"step": 42
},
{
"epoch": 0.00215,
"grad_norm": 33.150264739990234,
"learning_rate": 8.400000000000001e-06,
"loss": 0.7959,
"step": 43
},
{
"epoch": 0.0022,
"grad_norm": 5.327113151550293,
"learning_rate": 8.6e-06,
"loss": 0.2044,
"step": 44
},
{
"epoch": 0.00225,
"grad_norm": 6.212099075317383,
"learning_rate": 8.8e-06,
"loss": 0.1377,
"step": 45
},
{
"epoch": 0.0023,
"grad_norm": 41.24404525756836,
"learning_rate": 9e-06,
"loss": 0.1969,
"step": 46
},
{
"epoch": 0.00235,
"grad_norm": 55.61800765991211,
"learning_rate": 9.200000000000002e-06,
"loss": 0.8883,
"step": 47
},
{
"epoch": 0.0024,
"grad_norm": 43.88951110839844,
"learning_rate": 9.4e-06,
"loss": 0.3443,
"step": 48
},
{
"epoch": 0.00245,
"grad_norm": 16.66773796081543,
"learning_rate": 9.600000000000001e-06,
"loss": 0.1734,
"step": 49
},
{
"epoch": 0.0025,
"grad_norm": 80.06915283203125,
"learning_rate": 9.800000000000001e-06,
"loss": 0.4474,
"step": 50
},
{
"epoch": 0.00255,
"grad_norm": 66.25089263916016,
"learning_rate": 1e-05,
"loss": 0.6609,
"step": 51
},
{
"epoch": 0.0026,
"grad_norm": 21.153078079223633,
"learning_rate": 1.02e-05,
"loss": 0.1893,
"step": 52
},
{
"epoch": 0.00265,
"grad_norm": 68.32752227783203,
"learning_rate": 1.04e-05,
"loss": 0.4948,
"step": 53
},
{
"epoch": 0.0027,
"grad_norm": 54.980567932128906,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.613,
"step": 54
},
{
"epoch": 0.00275,
"grad_norm": 15.39164924621582,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.1968,
"step": 55
},
{
"epoch": 0.0028,
"grad_norm": 44.391944885253906,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.4588,
"step": 56
},
{
"epoch": 0.00285,
"grad_norm": 10.547879219055176,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.1715,
"step": 57
},
{
"epoch": 0.0029,
"grad_norm": 38.64859390258789,
"learning_rate": 1.14e-05,
"loss": 0.7959,
"step": 58
},
{
"epoch": 0.00295,
"grad_norm": 4.460653781890869,
"learning_rate": 1.16e-05,
"loss": 0.1458,
"step": 59
},
{
"epoch": 0.003,
"grad_norm": 25.946611404418945,
"learning_rate": 1.18e-05,
"loss": 0.566,
"step": 60
},
{
"epoch": 0.00305,
"grad_norm": 18.941125869750977,
"learning_rate": 1.2e-05,
"loss": 0.1629,
"step": 61
},
{
"epoch": 0.0031,
"grad_norm": 35.03038787841797,
"learning_rate": 1.22e-05,
"loss": 0.2886,
"step": 62
},
{
"epoch": 0.00315,
"grad_norm": 47.16691970825195,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.5181,
"step": 63
},
{
"epoch": 0.0032,
"grad_norm": 33.76936340332031,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.3963,
"step": 64
},
{
"epoch": 0.00325,
"grad_norm": 6.850577354431152,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.1536,
"step": 65
},
{
"epoch": 0.0033,
"grad_norm": 20.172531127929688,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.2775,
"step": 66
},
{
"epoch": 0.00335,
"grad_norm": 65.00743103027344,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.7195,
"step": 67
},
{
"epoch": 0.0034,
"grad_norm": 52.568260192871094,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.766,
"step": 68
},
{
"epoch": 0.00345,
"grad_norm": 9.241539001464844,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.1225,
"step": 69
},
{
"epoch": 0.0035,
"grad_norm": 4.773672103881836,
"learning_rate": 1.38e-05,
"loss": 0.1336,
"step": 70
},
{
"epoch": 0.00355,
"grad_norm": 48.60388946533203,
"learning_rate": 1.4e-05,
"loss": 0.5573,
"step": 71
},
{
"epoch": 0.0036,
"grad_norm": 27.8742618560791,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.2066,
"step": 72
},
{
"epoch": 0.00365,
"grad_norm": 26.072481155395508,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.5932,
"step": 73
},
{
"epoch": 0.0037,
"grad_norm": 7.826942443847656,
"learning_rate": 1.46e-05,
"loss": 0.1474,
"step": 74
},
{
"epoch": 0.00375,
"grad_norm": 18.62906837463379,
"learning_rate": 1.48e-05,
"loss": 0.1875,
"step": 75
},
{
"epoch": 0.0038,
"grad_norm": 8.554287910461426,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1298,
"step": 76
},
{
"epoch": 0.00385,
"grad_norm": 5.140439987182617,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.1242,
"step": 77
},
{
"epoch": 0.0039,
"grad_norm": 3.9687063694000244,
"learning_rate": 1.54e-05,
"loss": 0.122,
"step": 78
},
{
"epoch": 0.00395,
"grad_norm": 33.81657028198242,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.239,
"step": 79
},
{
"epoch": 0.004,
"grad_norm": 18.419599533081055,
"learning_rate": 1.58e-05,
"loss": 0.5223,
"step": 80
},
{
"epoch": 0.00405,
"grad_norm": 15.859134674072266,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.197,
"step": 81
},
{
"epoch": 0.0041,
"grad_norm": 52.44767761230469,
"learning_rate": 1.62e-05,
"loss": 0.5309,
"step": 82
},
{
"epoch": 0.00415,
"grad_norm": 8.842545509338379,
"learning_rate": 1.64e-05,
"loss": 0.1521,
"step": 83
},
{
"epoch": 0.0042,
"grad_norm": 2.0674080848693848,
"learning_rate": 1.66e-05,
"loss": 0.113,
"step": 84
},
{
"epoch": 0.00425,
"grad_norm": 1.7234963178634644,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.1075,
"step": 85
},
{
"epoch": 0.0043,
"grad_norm": 7.167241096496582,
"learning_rate": 1.7e-05,
"loss": 0.1667,
"step": 86
},
{
"epoch": 0.00435,
"grad_norm": 39.83475112915039,
"learning_rate": 1.72e-05,
"loss": 0.2723,
"step": 87
},
{
"epoch": 0.0044,
"grad_norm": 21.659135818481445,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.2837,
"step": 88
},
{
"epoch": 0.00445,
"grad_norm": 11.999871253967285,
"learning_rate": 1.76e-05,
"loss": 0.1919,
"step": 89
},
{
"epoch": 0.0045,
"grad_norm": 26.082788467407227,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.1552,
"step": 90
},
{
"epoch": 0.00455,
"grad_norm": 2.2943668365478516,
"learning_rate": 1.8e-05,
"loss": 0.0816,
"step": 91
},
{
"epoch": 0.0046,
"grad_norm": 2.454077959060669,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.105,
"step": 92
},
{
"epoch": 0.00465,
"grad_norm": 4.69025993347168,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.1104,
"step": 93
},
{
"epoch": 0.0047,
"grad_norm": 14.2959566116333,
"learning_rate": 1.86e-05,
"loss": 0.8841,
"step": 94
},
{
"epoch": 0.00475,
"grad_norm": 39.46820068359375,
"learning_rate": 1.88e-05,
"loss": 0.764,
"step": 95
},
{
"epoch": 0.0048,
"grad_norm": 2.3603274822235107,
"learning_rate": 1.9e-05,
"loss": 0.119,
"step": 96
},
{
"epoch": 0.00485,
"grad_norm": 14.50840950012207,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.7894,
"step": 97
},
{
"epoch": 0.0049,
"grad_norm": 55.66713333129883,
"learning_rate": 1.94e-05,
"loss": 0.3007,
"step": 98
},
{
"epoch": 0.00495,
"grad_norm": 73.19747924804688,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.3876,
"step": 99
},
{
"epoch": 0.005,
"grad_norm": 12.380102157592773,
"learning_rate": 1.98e-05,
"loss": 0.4437,
"step": 100
},
{
"epoch": 0.00505,
"grad_norm": 60.36326217651367,
"learning_rate": 2e-05,
"loss": 0.3569,
"step": 101
},
{
"epoch": 0.0051,
"grad_norm": 37.3046989440918,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.4898,
"step": 102
},
{
"epoch": 0.00515,
"grad_norm": 15.11868953704834,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.5021,
"step": 103
},
{
"epoch": 0.0052,
"grad_norm": 15.323163986206055,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.3722,
"step": 104
},
{
"epoch": 0.00525,
"grad_norm": 4.708047389984131,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.1488,
"step": 105
},
{
"epoch": 0.0053,
"grad_norm": 4.781192779541016,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.1934,
"step": 106
},
{
"epoch": 0.00535,
"grad_norm": 56.712425231933594,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.8702,
"step": 107
},
{
"epoch": 0.0054,
"grad_norm": 24.84933853149414,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.7662,
"step": 108
},
{
"epoch": 0.00545,
"grad_norm": 88.01167297363281,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.323,
"step": 109
},
{
"epoch": 0.0055,
"grad_norm": 7.916687965393066,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.1437,
"step": 110
},
{
"epoch": 0.00555,
"grad_norm": 27.291078567504883,
"learning_rate": 1.999390827019096e-05,
"loss": 0.2062,
"step": 111
},
{
"epoch": 0.0056,
"grad_norm": 39.047706604003906,
"learning_rate": 1.999262916410621e-05,
"loss": 0.3426,
"step": 112
},
{
"epoch": 0.00565,
"grad_norm": 85.04743957519531,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.4455,
"step": 113
},
{
"epoch": 0.0057,
"grad_norm": 45.45951843261719,
"learning_rate": 1.998970569790715e-05,
"loss": 1.1456,
"step": 114
},
{
"epoch": 0.00575,
"grad_norm": 26.305465698242188,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.4286,
"step": 115
},
{
"epoch": 0.0058,
"grad_norm": 4.146256446838379,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.1529,
"step": 116
},
{
"epoch": 0.00585,
"grad_norm": 41.741695404052734,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.3335,
"step": 117
},
{
"epoch": 0.0059,
"grad_norm": 4.166099548339844,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.1271,
"step": 118
},
{
"epoch": 0.00595,
"grad_norm": 16.836219787597656,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.3735,
"step": 119
},
{
"epoch": 0.006,
"grad_norm": 12.967101097106934,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.2148,
"step": 120
},
{
"epoch": 0.00605,
"grad_norm": 5.404827117919922,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.1681,
"step": 121
},
{
"epoch": 0.0061,
"grad_norm": 1.8873947858810425,
"learning_rate": 1.997314477224458e-05,
"loss": 0.109,
"step": 122
},
{
"epoch": 0.00615,
"grad_norm": 2.6838910579681396,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1122,
"step": 123
},
{
"epoch": 0.0062,
"grad_norm": 3.9862358570098877,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.1129,
"step": 124
},
{
"epoch": 0.00625,
"grad_norm": 17.899574279785156,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.1527,
"step": 125
},
{
"epoch": 0.0063,
"grad_norm": 27.26593589782715,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.2046,
"step": 126
},
{
"epoch": 0.00635,
"grad_norm": 69.12671661376953,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.3191,
"step": 127
},
{
"epoch": 0.0064,
"grad_norm": 29.109697341918945,
"learning_rate": 1.99556196460308e-05,
"loss": 0.7056,
"step": 128
},
{
"epoch": 0.00645,
"grad_norm": 3.6510722637176514,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.1104,
"step": 129
},
{
"epoch": 0.0065,
"grad_norm": 56.172035217285156,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.5127,
"step": 130
},
{
"epoch": 0.00655,
"grad_norm": 36.64286804199219,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.3449,
"step": 131
},
{
"epoch": 0.0066,
"grad_norm": 6.342191219329834,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.1143,
"step": 132
},
{
"epoch": 0.00665,
"grad_norm": 14.290639877319336,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.1838,
"step": 133
},
{
"epoch": 0.0067,
"grad_norm": 2.2349565029144287,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.0817,
"step": 134
},
{
"epoch": 0.00675,
"grad_norm": 3.0459256172180176,
"learning_rate": 1.992965508106537e-05,
"loss": 0.1034,
"step": 135
},
{
"epoch": 0.0068,
"grad_norm": 54.64921951293945,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.8331,
"step": 136
},
{
"epoch": 0.00685,
"grad_norm": 48.370426177978516,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.2902,
"step": 137
},
{
"epoch": 0.0069,
"grad_norm": 4.709333896636963,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.1155,
"step": 138
},
{
"epoch": 0.00695,
"grad_norm": 16.851655960083008,
"learning_rate": 1.991215540251542e-05,
"loss": 0.6068,
"step": 139
},
{
"epoch": 0.007,
"grad_norm": 4.256289482116699,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.0947,
"step": 140
},
{
"epoch": 0.00705,
"grad_norm": 10.517683029174805,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.0943,
"step": 141
},
{
"epoch": 0.0071,
"grad_norm": 23.348649978637695,
"learning_rate": 1.989776230907789e-05,
"loss": 0.1692,
"step": 142
},
{
"epoch": 0.00715,
"grad_norm": 12.16116714477539,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1112,
"step": 143
},
{
"epoch": 0.0072,
"grad_norm": 1.416853427886963,
"learning_rate": 1.988756381047006e-05,
"loss": 0.0841,
"step": 144
},
{
"epoch": 0.00725,
"grad_norm": 1.5739504098892212,
"learning_rate": 1.988228381446553e-05,
"loss": 0.0962,
"step": 145
},
{
"epoch": 0.0073,
"grad_norm": 69.6449966430664,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.2386,
"step": 146
},
{
"epoch": 0.00735,
"grad_norm": 4.477534294128418,
"learning_rate": 1.987136265072988e-05,
"loss": 0.0814,
"step": 147
},
{
"epoch": 0.0074,
"grad_norm": 2.3994815349578857,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.0919,
"step": 148
},
{
"epoch": 0.00745,
"grad_norm": 2.9941301345825195,
"learning_rate": 1.985996037070505e-05,
"loss": 0.0951,
"step": 149
},
{
"epoch": 0.0075,
"grad_norm": 2.181401252746582,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.1014,
"step": 150
},
{
"epoch": 0.00755,
"grad_norm": 42.53776550292969,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.4801,
"step": 151
},
{
"epoch": 0.0076,
"grad_norm": 37.03893280029297,
"learning_rate": 1.984195607969242e-05,
"loss": 0.4286,
"step": 152
},
{
"epoch": 0.00765,
"grad_norm": 3.7320573329925537,
"learning_rate": 1.983571470813386e-05,
"loss": 0.0751,
"step": 153
},
{
"epoch": 0.0077,
"grad_norm": 36.467708587646484,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.2603,
"step": 154
},
{
"epoch": 0.00775,
"grad_norm": 1.3517638444900513,
"learning_rate": 1.982287250728689e-05,
"loss": 0.0881,
"step": 155
},
{
"epoch": 0.0078,
"grad_norm": 3.801837921142578,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.0931,
"step": 156
},
{
"epoch": 0.00785,
"grad_norm": 3.4581966400146484,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.1433,
"step": 157
},
{
"epoch": 0.0079,
"grad_norm": 10.873891830444336,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.1472,
"step": 158
},
{
"epoch": 0.00795,
"grad_norm": 1.951228380203247,
"learning_rate": 1.979575249599344e-05,
"loss": 0.1118,
"step": 159
},
{
"epoch": 0.008,
"grad_norm": 1.2118356227874756,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.0746,
"step": 160
},
{
"epoch": 0.00805,
"grad_norm": 102.95352935791016,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.6052,
"step": 161
},
{
"epoch": 0.0081,
"grad_norm": 1.6797130107879639,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.0851,
"step": 162
},
{
"epoch": 0.00815,
"grad_norm": 46.02894592285156,
"learning_rate": 1.9766722783341682e-05,
"loss": 1.073,
"step": 163
},
{
"epoch": 0.0082,
"grad_norm": 137.5884246826172,
"learning_rate": 1.9759167619387474e-05,
"loss": 1.3911,
"step": 164
},
{
"epoch": 0.00825,
"grad_norm": 13.85587215423584,
"learning_rate": 1.9751493543055634e-05,
"loss": 1.0056,
"step": 165
},
{
"epoch": 0.0083,
"grad_norm": 2.5946624279022217,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.1312,
"step": 166
},
{
"epoch": 0.00835,
"grad_norm": 38.82013702392578,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.4317,
"step": 167
},
{
"epoch": 0.0084,
"grad_norm": 13.546914100646973,
"learning_rate": 1.972775878209397e-05,
"loss": 0.1253,
"step": 168
},
{
"epoch": 0.00845,
"grad_norm": 3.1792168617248535,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.0878,
"step": 169
},
{
"epoch": 0.0085,
"grad_norm": 1.025294542312622,
"learning_rate": 1.971134279909636e-05,
"loss": 0.0818,
"step": 170
},
{
"epoch": 0.00855,
"grad_norm": 3.16849684715271,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.0923,
"step": 171
},
{
"epoch": 0.0086,
"grad_norm": 22.248876571655273,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.4423,
"step": 172
},
{
"epoch": 0.00865,
"grad_norm": 3.4709620475769043,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.1042,
"step": 173
},
{
"epoch": 0.0087,
"grad_norm": 11.294670104980469,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.1051,
"step": 174
},
{
"epoch": 0.00875,
"grad_norm": 47.32283401489258,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.6385,
"step": 175
},
{
"epoch": 0.0088,
"grad_norm": 2.0372188091278076,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.0944,
"step": 176
},
{
"epoch": 0.00885,
"grad_norm": 18.498504638671875,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.1258,
"step": 177
},
{
"epoch": 0.0089,
"grad_norm": 6.902488708496094,
"learning_rate": 1.96409540423411e-05,
"loss": 0.0864,
"step": 178
},
{
"epoch": 0.00895,
"grad_norm": 3.099083423614502,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.1176,
"step": 179
},
{
"epoch": 0.009,
"grad_norm": 47.037715911865234,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.4287,
"step": 180
},
{
"epoch": 0.00905,
"grad_norm": 28.77621841430664,
"learning_rate": 1.961261695938319e-05,
"loss": 0.6359,
"step": 181
},
{
"epoch": 0.0091,
"grad_norm": 30.206228256225586,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.2638,
"step": 182
},
{
"epoch": 0.00915,
"grad_norm": 2.1811752319335938,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.081,
"step": 183
},
{
"epoch": 0.0092,
"grad_norm": 4.938018798828125,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.0994,
"step": 184
},
{
"epoch": 0.00925,
"grad_norm": 1.8686426877975464,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.0658,
"step": 185
},
{
"epoch": 0.0093,
"grad_norm": 1.452397346496582,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.0799,
"step": 186
},
{
"epoch": 0.00935,
"grad_norm": 9.045763969421387,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.1273,
"step": 187
},
{
"epoch": 0.0094,
"grad_norm": 66.31456756591797,
"learning_rate": 1.954240328516277e-05,
"loss": 0.7519,
"step": 188
},
{
"epoch": 0.00945,
"grad_norm": 2.3969008922576904,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.0895,
"step": 189
},
{
"epoch": 0.0095,
"grad_norm": 4.230534553527832,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.1075,
"step": 190
},
{
"epoch": 0.00955,
"grad_norm": 2.6471550464630127,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.1113,
"step": 191
},
{
"epoch": 0.0096,
"grad_norm": 2.1071929931640625,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.071,
"step": 192
},
{
"epoch": 0.00965,
"grad_norm": 29.577516555786133,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.2003,
"step": 193
},
{
"epoch": 0.0097,
"grad_norm": 2.6140875816345215,
"learning_rate": 1.947768410009586e-05,
"loss": 0.0934,
"step": 194
},
{
"epoch": 0.00975,
"grad_norm": 20.867664337158203,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.6841,
"step": 195
},
{
"epoch": 0.0098,
"grad_norm": 4.463425159454346,
"learning_rate": 1.945518575599317e-05,
"loss": 0.1023,
"step": 196
},
{
"epoch": 0.00985,
"grad_norm": 27.137752532958984,
"learning_rate": 1.944376370237481e-05,
"loss": 0.28,
"step": 197
},
{
"epoch": 0.0099,
"grad_norm": 27.44022560119629,
"learning_rate": 1.943222657947601e-05,
"loss": 0.262,
"step": 198
},
{
"epoch": 0.00995,
"grad_norm": 86.16558074951172,
"learning_rate": 1.942057452787297e-05,
"loss": 0.2155,
"step": 199
},
{
"epoch": 0.01,
"grad_norm": 4.05201530456543,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0818,
"step": 200
},
{
"epoch": 0.01005,
"grad_norm": 2.2698569297790527,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.0961,
"step": 201
},
{
"epoch": 0.0101,
"grad_norm": 3.336587429046631,
"learning_rate": 1.938493022759556e-05,
"loss": 0.0802,
"step": 202
},
{
"epoch": 0.01015,
"grad_norm": 2.88741397857666,
"learning_rate": 1.937281989491892e-05,
"loss": 0.0804,
"step": 203
},
{
"epoch": 0.0102,
"grad_norm": 35.056396484375,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.5725,
"step": 204
},
{
"epoch": 0.01025,
"grad_norm": 4.568849086761475,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.0933,
"step": 205
},
{
"epoch": 0.0103,
"grad_norm": 54.54572296142578,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.3056,
"step": 206
},
{
"epoch": 0.01035,
"grad_norm": 47.241539001464844,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.2674,
"step": 207
},
{
"epoch": 0.0104,
"grad_norm": 17.177762985229492,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.1673,
"step": 208
},
{
"epoch": 0.01045,
"grad_norm": 1.9468616247177124,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.1193,
"step": 209
},
{
"epoch": 0.0105,
"grad_norm": 18.45438575744629,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.2822,
"step": 210
},
{
"epoch": 0.01055,
"grad_norm": 43.80156707763672,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.9663,
"step": 211
},
{
"epoch": 0.0106,
"grad_norm": 2.534038543701172,
"learning_rate": 1.925870584809995e-05,
"loss": 0.0924,
"step": 212
},
{
"epoch": 0.01065,
"grad_norm": 24.710065841674805,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.1385,
"step": 213
},
{
"epoch": 0.0107,
"grad_norm": 33.71287536621094,
"learning_rate": 1.923210217112981e-05,
"loss": 0.1855,
"step": 214
},
{
"epoch": 0.01075,
"grad_norm": 16.790836334228516,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.1124,
"step": 215
},
{
"epoch": 0.0108,
"grad_norm": 4.847932815551758,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.102,
"step": 216
},
{
"epoch": 0.01085,
"grad_norm": 47.94908142089844,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.6322,
"step": 217
},
{
"epoch": 0.0109,
"grad_norm": 1.2221285104751587,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.0878,
"step": 218
},
{
"epoch": 0.01095,
"grad_norm": 21.5382137298584,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.5478,
"step": 219
},
{
"epoch": 0.011,
"grad_norm": 6.854888916015625,
"learning_rate": 1.914959667849825e-05,
"loss": 0.1052,
"step": 220
},
{
"epoch": 0.01105,
"grad_norm": 2.5299158096313477,
"learning_rate": 1.913545457642601e-05,
"loss": 0.0806,
"step": 221
},
{
"epoch": 0.0111,
"grad_norm": 6.152573585510254,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.1596,
"step": 222
},
{
"epoch": 0.01115,
"grad_norm": 4.504411220550537,
"learning_rate": 1.910683660806177e-05,
"loss": 0.1096,
"step": 223
},
{
"epoch": 0.0112,
"grad_norm": 34.63329315185547,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.5244,
"step": 224
},
{
"epoch": 0.01125,
"grad_norm": 2.805756092071533,
"learning_rate": 1.907777478532909e-05,
"loss": 0.0859,
"step": 225
},
{
"epoch": 0.0113,
"grad_norm": 7.697692394256592,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.1499,
"step": 226
},
{
"epoch": 0.01135,
"grad_norm": 3.3117737770080566,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.128,
"step": 227
},
{
"epoch": 0.0114,
"grad_norm": 84.48941802978516,
"learning_rate": 1.903335292863301e-05,
"loss": 0.2558,
"step": 228
},
{
"epoch": 0.01145,
"grad_norm": 5.932098865509033,
"learning_rate": 1.901832526405114e-05,
"loss": 0.152,
"step": 229
},
{
"epoch": 0.0115,
"grad_norm": 2.152769088745117,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.1035,
"step": 230
},
{
"epoch": 0.01155,
"grad_norm": 2.9696671962738037,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.133,
"step": 231
},
{
"epoch": 0.0116,
"grad_norm": 1.1272964477539062,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0709,
"step": 232
},
{
"epoch": 0.01165,
"grad_norm": 54.1812629699707,
"learning_rate": 1.895711760239413e-05,
"loss": 1.4537,
"step": 233
},
{
"epoch": 0.0117,
"grad_norm": 6.666309356689453,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1203,
"step": 234
},
{
"epoch": 0.01175,
"grad_norm": 14.333596229553223,
"learning_rate": 1.892585818452126e-05,
"loss": 0.1418,
"step": 235
},
{
"epoch": 0.0118,
"grad_norm": 9.919900894165039,
"learning_rate": 1.891006524188368e-05,
"loss": 0.1157,
"step": 236
},
{
"epoch": 0.01185,
"grad_norm": 16.620220184326172,
"learning_rate": 1.889416373291298e-05,
"loss": 0.1148,
"step": 237
},
{
"epoch": 0.0119,
"grad_norm": 1.308577537536621,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.0823,
"step": 238
},
{
"epoch": 0.01195,
"grad_norm": 2.889721155166626,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.0808,
"step": 239
},
{
"epoch": 0.012,
"grad_norm": 9.654630661010742,
"learning_rate": 1.884580975215084e-05,
"loss": 0.1474,
"step": 240
},
{
"epoch": 0.01205,
"grad_norm": 1.4449971914291382,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.0831,
"step": 241
},
{
"epoch": 0.0121,
"grad_norm": 2.7385785579681396,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.1073,
"step": 242
},
{
"epoch": 0.01215,
"grad_norm": 2.1100575923919678,
"learning_rate": 1.879648572866617e-05,
"loss": 0.0856,
"step": 243
},
{
"epoch": 0.0122,
"grad_norm": 22.79264259338379,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.1046,
"step": 244
},
{
"epoch": 0.01225,
"grad_norm": 2.072134017944336,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.0798,
"step": 245
},
{
"epoch": 0.0123,
"grad_norm": 28.02008628845215,
"learning_rate": 1.874619707139396e-05,
"loss": 0.3727,
"step": 246
},
{
"epoch": 0.01235,
"grad_norm": 1.2192667722702026,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.0597,
"step": 247
},
{
"epoch": 0.0124,
"grad_norm": 2.538534641265869,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.1174,
"step": 248
},
{
"epoch": 0.01245,
"grad_norm": 3.2410473823547363,
"learning_rate": 1.869494929505219e-05,
"loss": 0.1064,
"step": 249
},
{
"epoch": 0.0125,
"grad_norm": 9.867326736450195,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.7139,
"step": 250
},
{
"epoch": 0.01255,
"grad_norm": 2.04473876953125,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0692,
"step": 251
},
{
"epoch": 0.0126,
"grad_norm": 1.841275930404663,
"learning_rate": 1.864274801953705e-05,
"loss": 0.0794,
"step": 252
},
{
"epoch": 0.01265,
"grad_norm": 1.7441390752792358,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.0713,
"step": 253
},
{
"epoch": 0.0127,
"grad_norm": 89.4349594116211,
"learning_rate": 1.860742027003944e-05,
"loss": 0.317,
"step": 254
},
{
"epoch": 0.01275,
"grad_norm": 59.13755798339844,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.2077,
"step": 255
},
{
"epoch": 0.0128,
"grad_norm": 24.257322311401367,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.3272,
"step": 256
},
{
"epoch": 0.01285,
"grad_norm": 3.105194091796875,
"learning_rate": 1.855364260160507e-05,
"loss": 0.0992,
"step": 257
},
{
"epoch": 0.0129,
"grad_norm": 2.192078113555908,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.069,
"step": 258
},
{
"epoch": 0.01295,
"grad_norm": 21.576236724853516,
"learning_rate": 1.851726934143048e-05,
"loss": 0.54,
"step": 259
},
{
"epoch": 0.013,
"grad_norm": 4.011931419372559,
"learning_rate": 1.849892692986864e-05,
"loss": 0.1241,
"step": 260
},
{
"epoch": 0.01305,
"grad_norm": 1.8716342449188232,
"learning_rate": 1.848048096156426e-05,
"loss": 0.0919,
"step": 261
},
{
"epoch": 0.0131,
"grad_norm": 36.63429641723633,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.8812,
"step": 262
},
{
"epoch": 0.01315,
"grad_norm": 36.83095169067383,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.4525,
"step": 263
},
{
"epoch": 0.0132,
"grad_norm": 2.323366165161133,
"learning_rate": 1.842452397007148e-05,
"loss": 0.0841,
"step": 264
},
{
"epoch": 0.01325,
"grad_norm": 14.090721130371094,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.1602,
"step": 265
},
{
"epoch": 0.0133,
"grad_norm": 12.795496940612793,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.182,
"step": 266
},
{
"epoch": 0.01335,
"grad_norm": 4.722551345825195,
"learning_rate": 1.836764313458962e-05,
"loss": 0.1357,
"step": 267
},
{
"epoch": 0.0134,
"grad_norm": 16.780805587768555,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.3341,
"step": 268
},
{
"epoch": 0.01345,
"grad_norm": 1.7280741930007935,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.0941,
"step": 269
},
{
"epoch": 0.0135,
"grad_norm": 3.2625572681427,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.1768,
"step": 270
},
{
"epoch": 0.01355,
"grad_norm": 2.5432939529418945,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.0925,
"step": 271
},
{
"epoch": 0.0136,
"grad_norm": 57.257667541503906,
"learning_rate": 1.827080574274562e-05,
"loss": 0.3118,
"step": 272
},
{
"epoch": 0.01365,
"grad_norm": 7.988307476043701,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.1129,
"step": 273
},
{
"epoch": 0.0137,
"grad_norm": 1.5597671270370483,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.0753,
"step": 274
},
{
"epoch": 0.01375,
"grad_norm": 18.671035766601562,
"learning_rate": 1.821149209133704e-05,
"loss": 0.1241,
"step": 275
},
{
"epoch": 0.0138,
"grad_norm": 1.7065479755401611,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0809,
"step": 276
},
{
"epoch": 0.01385,
"grad_norm": 31.31546401977539,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.2415,
"step": 277
},
{
"epoch": 0.0139,
"grad_norm": 2.4892303943634033,
"learning_rate": 1.815127795728554e-05,
"loss": 0.096,
"step": 278
},
{
"epoch": 0.01395,
"grad_norm": 2.3767950534820557,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.1194,
"step": 279
},
{
"epoch": 0.014,
"grad_norm": 24.376630783081055,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.2867,
"step": 280
},
{
"epoch": 0.01405,
"grad_norm": 2.7759625911712646,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.1072,
"step": 281
},
{
"epoch": 0.0141,
"grad_norm": 16.7757511138916,
"learning_rate": 1.806960312143802e-05,
"loss": 0.6312,
"step": 282
},
{
"epoch": 0.01415,
"grad_norm": 4.640740871429443,
"learning_rate": 1.804893797355914e-05,
"loss": 0.1236,
"step": 283
},
{
"epoch": 0.0142,
"grad_norm": 1.8363475799560547,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.0859,
"step": 284
},
{
"epoch": 0.01425,
"grad_norm": 1.5419412851333618,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.0685,
"step": 285
},
{
"epoch": 0.0143,
"grad_norm": 45.82707214355469,
"learning_rate": 1.798635510047293e-05,
"loss": 0.552,
"step": 286
},
{
"epoch": 0.01435,
"grad_norm": 1.4965375661849976,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0868,
"step": 287
},
{
"epoch": 0.0144,
"grad_norm": 34.65620422363281,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.851,
"step": 288
},
{
"epoch": 0.01445,
"grad_norm": 2.195761203765869,
"learning_rate": 1.792289643355191e-05,
"loss": 0.1009,
"step": 289
},
{
"epoch": 0.0145,
"grad_norm": 1.3684836626052856,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0731,
"step": 290
},
{
"epoch": 0.01455,
"grad_norm": 1.6004326343536377,
"learning_rate": 1.788010753606722e-05,
"loss": 0.0831,
"step": 291
},
{
"epoch": 0.0146,
"grad_norm": 26.745464324951172,
"learning_rate": 1.785856893175402e-05,
"loss": 0.4283,
"step": 292
},
{
"epoch": 0.01465,
"grad_norm": 78.65120697021484,
"learning_rate": 1.78369345732584e-05,
"loss": 0.3526,
"step": 293
},
{
"epoch": 0.0147,
"grad_norm": 1.7803980112075806,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0837,
"step": 294
},
{
"epoch": 0.01475,
"grad_norm": 8.317599296569824,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.1051,
"step": 295
},
{
"epoch": 0.0148,
"grad_norm": 10.342424392700195,
"learning_rate": 1.777145961456971e-05,
"loss": 0.11,
"step": 296
},
{
"epoch": 0.01485,
"grad_norm": 35.15061950683594,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.3765,
"step": 297
},
{
"epoch": 0.0149,
"grad_norm": 11.10451889038086,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.1197,
"step": 298
},
{
"epoch": 0.01495,
"grad_norm": 1.8988792896270752,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.1009,
"step": 299
},
{
"epoch": 0.015,
"grad_norm": 6.130817413330078,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.1401,
"step": 300
},
{
"epoch": 0.01505,
"grad_norm": 1.9723719358444214,
"learning_rate": 1.766044443118978e-05,
"loss": 0.0877,
"step": 301
},
{
"epoch": 0.0151,
"grad_norm": 2.882294178009033,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.0815,
"step": 302
},
{
"epoch": 0.01515,
"grad_norm": 18.67458152770996,
"learning_rate": 1.761538307536737e-05,
"loss": 0.1527,
"step": 303
},
{
"epoch": 0.0152,
"grad_norm": 34.37173080444336,
"learning_rate": 1.759271307334881e-05,
"loss": 0.6591,
"step": 304
},
{
"epoch": 0.01525,
"grad_norm": 1.8363667726516724,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0853,
"step": 305
},
{
"epoch": 0.0153,
"grad_norm": 1.6190111637115479,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0684,
"step": 306
},
{
"epoch": 0.01535,
"grad_norm": 2.7849435806274414,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.1081,
"step": 307
},
{
"epoch": 0.0154,
"grad_norm": 2.385519504547119,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.0699,
"step": 308
},
{
"epoch": 0.01545,
"grad_norm": 103.18679809570312,
"learning_rate": 1.747798090498532e-05,
"loss": 0.2133,
"step": 309
},
{
"epoch": 0.0155,
"grad_norm": 21.587390899658203,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.1336,
"step": 310
},
{
"epoch": 0.01555,
"grad_norm": 64.16925048828125,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.5498,
"step": 311
},
{
"epoch": 0.0156,
"grad_norm": 1.3994704484939575,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0807,
"step": 312
},
{
"epoch": 0.01565,
"grad_norm": 2.7629127502441406,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.1061,
"step": 313
},
{
"epoch": 0.0157,
"grad_norm": 0.7181133031845093,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.0615,
"step": 314
},
{
"epoch": 0.01575,
"grad_norm": 1.9693636894226074,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.0854,
"step": 315
},
{
"epoch": 0.0158,
"grad_norm": 2.3914601802825928,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.1046,
"step": 316
},
{
"epoch": 0.01585,
"grad_norm": 1.8866361379623413,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.0777,
"step": 317
},
{
"epoch": 0.0159,
"grad_norm": 8.258055686950684,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.5726,
"step": 318
},
{
"epoch": 0.01595,
"grad_norm": 3.7941558361053467,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.1251,
"step": 319
},
{
"epoch": 0.016,
"grad_norm": 2.7949891090393066,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.0826,
"step": 320
},
{
"epoch": 0.01605,
"grad_norm": 27.749094009399414,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.4803,
"step": 321
},
{
"epoch": 0.0161,
"grad_norm": 2.399714708328247,
"learning_rate": 1.716910607650483e-05,
"loss": 0.1006,
"step": 322
},
{
"epoch": 0.01615,
"grad_norm": 5.782076835632324,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.1101,
"step": 323
},
{
"epoch": 0.0162,
"grad_norm": 2.2748818397521973,
"learning_rate": 1.712026045990997e-05,
"loss": 0.1027,
"step": 324
},
{
"epoch": 0.01625,
"grad_norm": 7.266913414001465,
"learning_rate": 1.709570736536521e-05,
"loss": 0.112,
"step": 325
},
{
"epoch": 0.0163,
"grad_norm": 1.6896015405654907,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.0824,
"step": 326
},
{
"epoch": 0.01635,
"grad_norm": 12.456267356872559,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.1114,
"step": 327
},
{
"epoch": 0.0164,
"grad_norm": 7.791019916534424,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.0909,
"step": 328
},
{
"epoch": 0.01645,
"grad_norm": 12.910639762878418,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.1899,
"step": 329
},
{
"epoch": 0.0165,
"grad_norm": 2.853062152862549,
"learning_rate": 1.697165102854565e-05,
"loss": 0.0908,
"step": 330
},
{
"epoch": 0.01655,
"grad_norm": 15.528128623962402,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.4061,
"step": 331
},
{
"epoch": 0.0166,
"grad_norm": 48.358848571777344,
"learning_rate": 1.692143173870407e-05,
"loss": 0.4315,
"step": 332
},
{
"epoch": 0.01665,
"grad_norm": 3.2647809982299805,
"learning_rate": 1.68961954373567e-05,
"loss": 0.1223,
"step": 333
},
{
"epoch": 0.0167,
"grad_norm": 11.482593536376953,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.1696,
"step": 334
},
{
"epoch": 0.01675,
"grad_norm": 1.3692466020584106,
"learning_rate": 1.684547105928689e-05,
"loss": 0.077,
"step": 335
},
{
"epoch": 0.0168,
"grad_norm": 117.79779052734375,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.2993,
"step": 336
},
{
"epoch": 0.01685,
"grad_norm": 2.0206730365753174,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.092,
"step": 337
},
{
"epoch": 0.0169,
"grad_norm": 34.888587951660156,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.1954,
"step": 338
},
{
"epoch": 0.01695,
"grad_norm": 7.209064483642578,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.1342,
"step": 339
},
{
"epoch": 0.017,
"grad_norm": 1.6078732013702393,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.0846,
"step": 340
},
{
"epoch": 0.01705,
"grad_norm": 63.080848693847656,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.411,
"step": 341
},
{
"epoch": 0.0171,
"grad_norm": 29.949630737304688,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.6917,
"step": 342
},
{
"epoch": 0.01715,
"grad_norm": 31.307920455932617,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.2668,
"step": 343
},
{
"epoch": 0.0172,
"grad_norm": 13.792570114135742,
"learning_rate": 1.661311865323652e-05,
"loss": 0.119,
"step": 344
},
{
"epoch": 0.01725,
"grad_norm": 30.71436882019043,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.6006,
"step": 345
},
{
"epoch": 0.0173,
"grad_norm": 2.122511863708496,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1022,
"step": 346
},
{
"epoch": 0.01735,
"grad_norm": 4.920584201812744,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.1127,
"step": 347
},
{
"epoch": 0.0174,
"grad_norm": 76.10581970214844,
"learning_rate": 1.650774217265851e-05,
"loss": 0.2452,
"step": 348
},
{
"epoch": 0.01745,
"grad_norm": 0.8686180114746094,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.0773,
"step": 349
},
{
"epoch": 0.0175,
"grad_norm": 2.1581313610076904,
"learning_rate": 1.645457687723951e-05,
"loss": 0.1051,
"step": 350
},
{
"epoch": 0.01755,
"grad_norm": 1.0424079895019531,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0849,
"step": 351
},
{
"epoch": 0.0176,
"grad_norm": 2.0918571949005127,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.0837,
"step": 352
},
{
"epoch": 0.01765,
"grad_norm": 2.8070149421691895,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0866,
"step": 353
},
{
"epoch": 0.0177,
"grad_norm": 2.4599366188049316,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.0953,
"step": 354
},
{
"epoch": 0.01775,
"grad_norm": 1.8522323369979858,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0858,
"step": 355
},
{
"epoch": 0.0178,
"grad_norm": 29.34244728088379,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.3812,
"step": 356
},
{
"epoch": 0.01785,
"grad_norm": 2.804577112197876,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.106,
"step": 357
},
{
"epoch": 0.0179,
"grad_norm": 1.0157254934310913,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.0745,
"step": 358
},
{
"epoch": 0.01795,
"grad_norm": 1.1413592100143433,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0695,
"step": 359
},
{
"epoch": 0.018,
"grad_norm": 1.7196788787841797,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.0858,
"step": 360
},
{
"epoch": 0.01805,
"grad_norm": 9.58154296875,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.5321,
"step": 361
},
{
"epoch": 0.0181,
"grad_norm": 27.282711029052734,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.2956,
"step": 362
},
{
"epoch": 0.01815,
"grad_norm": 2.1738831996917725,
"learning_rate": 1.610145163901268e-05,
"loss": 0.0872,
"step": 363
},
{
"epoch": 0.0182,
"grad_norm": 54.08341598510742,
"learning_rate": 1.607375839723287e-05,
"loss": 0.2653,
"step": 364
},
{
"epoch": 0.01825,
"grad_norm": 32.51065444946289,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.6676,
"step": 365
},
{
"epoch": 0.0183,
"grad_norm": 2.3353264331817627,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.0746,
"step": 366
},
{
"epoch": 0.01835,
"grad_norm": 2.272801160812378,
"learning_rate": 1.599023598515586e-05,
"loss": 0.088,
"step": 367
},
{
"epoch": 0.0184,
"grad_norm": 3.1989638805389404,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.1,
"step": 368
},
{
"epoch": 0.01845,
"grad_norm": 2.0776147842407227,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.0704,
"step": 369
},
{
"epoch": 0.0185,
"grad_norm": 1.6304274797439575,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.068,
"step": 370
},
{
"epoch": 0.01855,
"grad_norm": 1.6919046640396118,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0871,
"step": 371
},
{
"epoch": 0.0186,
"grad_norm": 15.512767791748047,
"learning_rate": 1.584957674987216e-05,
"loss": 0.2278,
"step": 372
},
{
"epoch": 0.01865,
"grad_norm": 1.8702269792556763,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.1016,
"step": 373
},
{
"epoch": 0.0187,
"grad_norm": 2.2889702320098877,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.0825,
"step": 374
},
{
"epoch": 0.01875,
"grad_norm": 0.818047046661377,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.0647,
"step": 375
},
{
"epoch": 0.0188,
"grad_norm": 6.705317974090576,
"learning_rate": 1.573576436351046e-05,
"loss": 0.1588,
"step": 376
},
{
"epoch": 0.01885,
"grad_norm": 70.91289520263672,
"learning_rate": 1.570713567684432e-05,
"loss": 0.264,
"step": 377
},
{
"epoch": 0.0189,
"grad_norm": 4.24697732925415,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.1044,
"step": 378
},
{
"epoch": 0.01895,
"grad_norm": 1.2441545724868774,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0676,
"step": 379
},
{
"epoch": 0.019,
"grad_norm": 11.383224487304688,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.1717,
"step": 380
},
{
"epoch": 0.01905,
"grad_norm": 1.8020143508911133,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.0915,
"step": 381
},
{
"epoch": 0.0191,
"grad_norm": 1.5403460264205933,
"learning_rate": 1.556295615500305e-05,
"loss": 0.0942,
"step": 382
},
{
"epoch": 0.01915,
"grad_norm": 4.809276103973389,
"learning_rate": 1.553391549243344e-05,
"loss": 0.1307,
"step": 383
},
{
"epoch": 0.0192,
"grad_norm": 1.7769126892089844,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.0981,
"step": 384
},
{
"epoch": 0.01925,
"grad_norm": 21.044578552246094,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.1922,
"step": 385
},
{
"epoch": 0.0193,
"grad_norm": 2.1597342491149902,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0778,
"step": 386
},
{
"epoch": 0.01935,
"grad_norm": 71.87742614746094,
"learning_rate": 1.54170821028274e-05,
"loss": 0.2516,
"step": 387
},
{
"epoch": 0.0194,
"grad_norm": 2.8716869354248047,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0801,
"step": 388
},
{
"epoch": 0.01945,
"grad_norm": 3.6430041790008545,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0982,
"step": 389
},
{
"epoch": 0.0195,
"grad_norm": 3.5003151893615723,
"learning_rate": 1.53287627607073e-05,
"loss": 0.1146,
"step": 390
},
{
"epoch": 0.01955,
"grad_norm": 33.15953826904297,
"learning_rate": 1.529919264233205e-05,
"loss": 0.5638,
"step": 391
},
{
"epoch": 0.0196,
"grad_norm": 68.7998046875,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.2967,
"step": 392
},
{
"epoch": 0.01965,
"grad_norm": 4.234882831573486,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.108,
"step": 393
},
{
"epoch": 0.0197,
"grad_norm": 1.4191491603851318,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.0816,
"step": 394
},
{
"epoch": 0.01975,
"grad_norm": 1.553654670715332,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0849,
"step": 395
},
{
"epoch": 0.0198,
"grad_norm": 28.652915954589844,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.3633,
"step": 396
},
{
"epoch": 0.01985,
"grad_norm": 7.698509216308594,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.1178,
"step": 397
},
{
"epoch": 0.0199,
"grad_norm": 1.612652063369751,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0578,
"step": 398
},
{
"epoch": 0.01995,
"grad_norm": 19.84796714782715,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.321,
"step": 399
},
{
"epoch": 0.02,
"grad_norm": 5.804268836975098,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.1119,
"step": 400
},
{
"epoch": 0.02005,
"grad_norm": 30.93545913696289,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2467,
"step": 401
},
{
"epoch": 0.0201,
"grad_norm": 77.25408172607422,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.3551,
"step": 402
},
{
"epoch": 0.02015,
"grad_norm": 1.7736424207687378,
"learning_rate": 1.493941866584231e-05,
"loss": 0.0929,
"step": 403
},
{
"epoch": 0.0202,
"grad_norm": 5.388272285461426,
"learning_rate": 1.490903753615141e-05,
"loss": 0.1161,
"step": 404
},
{
"epoch": 0.02025,
"grad_norm": 2.175442934036255,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.0694,
"step": 405
},
{
"epoch": 0.0203,
"grad_norm": 1.5696885585784912,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1043,
"step": 406
},
{
"epoch": 0.02035,
"grad_norm": 1.4840455055236816,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.0649,
"step": 407
},
{
"epoch": 0.0204,
"grad_norm": 27.966121673583984,
"learning_rate": 1.478691857940607e-05,
"loss": 0.1131,
"step": 408
},
{
"epoch": 0.02045,
"grad_norm": 1.381766438484192,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0965,
"step": 409
},
{
"epoch": 0.0205,
"grad_norm": 1.185716986656189,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0692,
"step": 410
},
{
"epoch": 0.02055,
"grad_norm": 1.860116720199585,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0943,
"step": 411
},
{
"epoch": 0.0206,
"grad_norm": 1.27677583694458,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.061,
"step": 412
},
{
"epoch": 0.02065,
"grad_norm": 1.2773680686950684,
"learning_rate": 1.463296035119862e-05,
"loss": 0.079,
"step": 413
},
{
"epoch": 0.0207,
"grad_norm": 3.0321786403656006,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.0769,
"step": 414
},
{
"epoch": 0.02075,
"grad_norm": 1.8801988363265991,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0725,
"step": 415
},
{
"epoch": 0.0208,
"grad_norm": 35.990970611572266,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.5037,
"step": 416
},
{
"epoch": 0.02085,
"grad_norm": 21.723236083984375,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.5753,
"step": 417
},
{
"epoch": 0.0209,
"grad_norm": 1.4752088785171509,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.0777,
"step": 418
},
{
"epoch": 0.02095,
"grad_norm": 81.2683334350586,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.2849,
"step": 419
},
{
"epoch": 0.021,
"grad_norm": 1.2820717096328735,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.0606,
"step": 420
},
{
"epoch": 0.02105,
"grad_norm": 43.90168380737305,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.15,
"step": 421
},
{
"epoch": 0.0211,
"grad_norm": 2.168410301208496,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0768,
"step": 422
},
{
"epoch": 0.02115,
"grad_norm": 2.721632242202759,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.0778,
"step": 423
},
{
"epoch": 0.0212,
"grad_norm": 31.003597259521484,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.3962,
"step": 424
},
{
"epoch": 0.02125,
"grad_norm": 2.4142940044403076,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.066,
"step": 425
},
{
"epoch": 0.0213,
"grad_norm": 0.9857182502746582,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0628,
"step": 426
},
{
"epoch": 0.02135,
"grad_norm": 27.239112854003906,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.1553,
"step": 427
},
{
"epoch": 0.0214,
"grad_norm": 1.5095759630203247,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.0623,
"step": 428
},
{
"epoch": 0.02145,
"grad_norm": 1.2446397542953491,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0632,
"step": 429
},
{
"epoch": 0.0215,
"grad_norm": 23.939973831176758,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.0938,
"step": 430
},
{
"epoch": 0.02155,
"grad_norm": 2.7718734741210938,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.1157,
"step": 431
},
{
"epoch": 0.0216,
"grad_norm": 1.762878656387329,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.0837,
"step": 432
},
{
"epoch": 0.02165,
"grad_norm": 1.1785478591918945,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0786,
"step": 433
},
{
"epoch": 0.0217,
"grad_norm": 2.3212571144104004,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0795,
"step": 434
},
{
"epoch": 0.02175,
"grad_norm": 61.90953826904297,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.1883,
"step": 435
},
{
"epoch": 0.0218,
"grad_norm": 2.4108712673187256,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.101,
"step": 436
},
{
"epoch": 0.02185,
"grad_norm": 18.406414031982422,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.124,
"step": 437
},
{
"epoch": 0.0219,
"grad_norm": 2.9103317260742188,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.1076,
"step": 438
},
{
"epoch": 0.02195,
"grad_norm": 3.6682307720184326,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.1115,
"step": 439
},
{
"epoch": 0.022,
"grad_norm": 2.4667904376983643,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0894,
"step": 440
},
{
"epoch": 0.02205,
"grad_norm": 1.5204410552978516,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0696,
"step": 441
},
{
"epoch": 0.0221,
"grad_norm": 2.2093870639801025,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0988,
"step": 442
},
{
"epoch": 0.02215,
"grad_norm": 1.1975420713424683,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.075,
"step": 443
},
{
"epoch": 0.0222,
"grad_norm": 41.98251724243164,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.4722,
"step": 444
},
{
"epoch": 0.02225,
"grad_norm": 2.698796272277832,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.093,
"step": 445
},
{
"epoch": 0.0223,
"grad_norm": 25.365121841430664,
"learning_rate": 1.3583679495453e-05,
"loss": 0.6883,
"step": 446
},
{
"epoch": 0.02235,
"grad_norm": 26.637304306030273,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.2127,
"step": 447
},
{
"epoch": 0.0224,
"grad_norm": 1.6217232942581177,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0728,
"step": 448
},
{
"epoch": 0.02245,
"grad_norm": 2.7400615215301514,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.0927,
"step": 449
},
{
"epoch": 0.0225,
"grad_norm": 2.490354537963867,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.0998,
"step": 450
},
{
"epoch": 0.02255,
"grad_norm": 3.2277610301971436,
"learning_rate": 1.342020143325669e-05,
"loss": 0.1005,
"step": 451
},
{
"epoch": 0.0226,
"grad_norm": 1.15992271900177,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0646,
"step": 452
},
{
"epoch": 0.02265,
"grad_norm": 2.447103261947632,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.0893,
"step": 453
},
{
"epoch": 0.0227,
"grad_norm": 3.7397661209106445,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.0956,
"step": 454
},
{
"epoch": 0.02275,
"grad_norm": 1.829380989074707,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0824,
"step": 455
},
{
"epoch": 0.0228,
"grad_norm": 1.3651745319366455,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.0723,
"step": 456
},
{
"epoch": 0.02285,
"grad_norm": 1.840768575668335,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0835,
"step": 457
},
{
"epoch": 0.0229,
"grad_norm": 34.218894958496094,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.4912,
"step": 458
},
{
"epoch": 0.02295,
"grad_norm": 1.4367423057556152,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0654,
"step": 459
},
{
"epoch": 0.023,
"grad_norm": 1.4354041814804077,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0607,
"step": 460
},
{
"epoch": 0.02305,
"grad_norm": 8.959512710571289,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0942,
"step": 461
},
{
"epoch": 0.0231,
"grad_norm": 12.93722152709961,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.2287,
"step": 462
},
{
"epoch": 0.02315,
"grad_norm": 0.9477676749229431,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0627,
"step": 463
},
{
"epoch": 0.0232,
"grad_norm": 21.566787719726562,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.1503,
"step": 464
},
{
"epoch": 0.02325,
"grad_norm": 1.426946997642517,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.0806,
"step": 465
},
{
"epoch": 0.0233,
"grad_norm": 24.813554763793945,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.4477,
"step": 466
},
{
"epoch": 0.02335,
"grad_norm": 1.1683995723724365,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.0683,
"step": 467
},
{
"epoch": 0.0234,
"grad_norm": 26.862457275390625,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.2431,
"step": 468
},
{
"epoch": 0.02345,
"grad_norm": 1.4641770124435425,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.08,
"step": 469
},
{
"epoch": 0.0235,
"grad_norm": 1.9108537435531616,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0797,
"step": 470
},
{
"epoch": 0.02355,
"grad_norm": 18.305461883544922,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.2025,
"step": 471
},
{
"epoch": 0.0236,
"grad_norm": 9.38343334197998,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.1712,
"step": 472
},
{
"epoch": 0.02365,
"grad_norm": 0.9332784414291382,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.061,
"step": 473
},
{
"epoch": 0.0237,
"grad_norm": 2.740528106689453,
"learning_rate": 1.265556117486809e-05,
"loss": 0.0643,
"step": 474
},
{
"epoch": 0.02375,
"grad_norm": 2.050279378890991,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.0739,
"step": 475
},
{
"epoch": 0.0238,
"grad_norm": 4.4667558670043945,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.0964,
"step": 476
},
{
"epoch": 0.02385,
"grad_norm": 1.703878402709961,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.0775,
"step": 477
},
{
"epoch": 0.0239,
"grad_norm": 1.2993277311325073,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0627,
"step": 478
},
{
"epoch": 0.02395,
"grad_norm": 21.238073348999023,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.188,
"step": 479
},
{
"epoch": 0.024,
"grad_norm": 1.4124360084533691,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.0864,
"step": 480
},
{
"epoch": 0.02405,
"grad_norm": 14.426772117614746,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.4066,
"step": 481
},
{
"epoch": 0.0241,
"grad_norm": 2.3008742332458496,
"learning_rate": 1.238533457578581e-05,
"loss": 0.109,
"step": 482
},
{
"epoch": 0.02415,
"grad_norm": 3.065260648727417,
"learning_rate": 1.23514211310259e-05,
"loss": 0.1144,
"step": 483
},
{
"epoch": 0.0242,
"grad_norm": 3.2149078845977783,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.1031,
"step": 484
},
{
"epoch": 0.02425,
"grad_norm": 1.5617332458496094,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.0558,
"step": 485
},
{
"epoch": 0.0243,
"grad_norm": 21.832523345947266,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.1919,
"step": 486
},
{
"epoch": 0.02435,
"grad_norm": 2.4033219814300537,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.0682,
"step": 487
},
{
"epoch": 0.0244,
"grad_norm": 17.43706512451172,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.4117,
"step": 488
},
{
"epoch": 0.02445,
"grad_norm": 3.9174792766571045,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.0823,
"step": 489
},
{
"epoch": 0.0245,
"grad_norm": 46.03153610229492,
"learning_rate": 1.211324796455389e-05,
"loss": 0.3065,
"step": 490
},
{
"epoch": 0.02455,
"grad_norm": 1.6286877393722534,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0705,
"step": 491
},
{
"epoch": 0.0246,
"grad_norm": 39.38934326171875,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.412,
"step": 492
},
{
"epoch": 0.02465,
"grad_norm": 18.489652633666992,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.1775,
"step": 493
},
{
"epoch": 0.0247,
"grad_norm": 1.5449779033660889,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0777,
"step": 494
},
{
"epoch": 0.02475,
"grad_norm": 1.7571698427200317,
"learning_rate": 1.194234351219972e-05,
"loss": 0.0909,
"step": 495
},
{
"epoch": 0.0248,
"grad_norm": 1.4077978134155273,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0665,
"step": 496
},
{
"epoch": 0.02485,
"grad_norm": 4.160821437835693,
"learning_rate": 1.187381314585725e-05,
"loss": 0.0962,
"step": 497
},
{
"epoch": 0.0249,
"grad_norm": 2.014230728149414,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0773,
"step": 498
},
{
"epoch": 0.02495,
"grad_norm": 2.2846667766571045,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.113,
"step": 499
},
{
"epoch": 0.025,
"grad_norm": 1.8118311166763306,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0598,
"step": 500
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}