CaffeineThief's picture
Upload saved model files
db02f52 verified
{
"best_global_step": 109,
"best_metric": 0.0,
"best_model_checkpoint": "./cysecbert-ttp-bert-base_data/checkpoint-109",
"epoch": 2.0,
"eval_steps": 500,
"global_step": 218,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009174311926605505,
"grad_norm": 46530.15234375,
"learning_rate": 0.0,
"loss": 0.7023,
"step": 1
},
{
"epoch": 0.01834862385321101,
"grad_norm": 49571.35546875,
"learning_rate": 5.000000000000001e-07,
"loss": 0.7006,
"step": 2
},
{
"epoch": 0.027522935779816515,
"grad_norm": 45885.55078125,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.7033,
"step": 3
},
{
"epoch": 0.03669724770642202,
"grad_norm": 43333.73046875,
"learning_rate": 1.5e-06,
"loss": 0.7014,
"step": 4
},
{
"epoch": 0.045871559633027525,
"grad_norm": 44340.41796875,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.6992,
"step": 5
},
{
"epoch": 0.05504587155963303,
"grad_norm": 48282.54296875,
"learning_rate": 2.5e-06,
"loss": 0.6924,
"step": 6
},
{
"epoch": 0.06422018348623854,
"grad_norm": 48561.56640625,
"learning_rate": 3e-06,
"loss": 0.6948,
"step": 7
},
{
"epoch": 0.07339449541284404,
"grad_norm": 43580.078125,
"learning_rate": 3.5000000000000004e-06,
"loss": 0.6916,
"step": 8
},
{
"epoch": 0.08256880733944955,
"grad_norm": 46552.28515625,
"learning_rate": 4.000000000000001e-06,
"loss": 0.6854,
"step": 9
},
{
"epoch": 0.09174311926605505,
"grad_norm": 49042.15234375,
"learning_rate": 4.5e-06,
"loss": 0.6822,
"step": 10
},
{
"epoch": 0.10091743119266056,
"grad_norm": 97274.8515625,
"learning_rate": 5e-06,
"loss": 0.6788,
"step": 11
},
{
"epoch": 0.11009174311926606,
"grad_norm": 55310.375,
"learning_rate": 5.500000000000001e-06,
"loss": 0.6781,
"step": 12
},
{
"epoch": 0.11926605504587157,
"grad_norm": 70317.6328125,
"learning_rate": 6e-06,
"loss": 0.6754,
"step": 13
},
{
"epoch": 0.12844036697247707,
"grad_norm": 46825.9765625,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.6626,
"step": 14
},
{
"epoch": 0.13761467889908258,
"grad_norm": 50745.15625,
"learning_rate": 7.000000000000001e-06,
"loss": 0.655,
"step": 15
},
{
"epoch": 0.14678899082568808,
"grad_norm": 51342.453125,
"learning_rate": 7.5e-06,
"loss": 0.6491,
"step": 16
},
{
"epoch": 0.1559633027522936,
"grad_norm": 51815.421875,
"learning_rate": 8.000000000000001e-06,
"loss": 0.6464,
"step": 17
},
{
"epoch": 0.1651376146788991,
"grad_norm": 54610.65234375,
"learning_rate": 8.500000000000002e-06,
"loss": 0.6298,
"step": 18
},
{
"epoch": 0.1743119266055046,
"grad_norm": 61936.31640625,
"learning_rate": 9e-06,
"loss": 0.623,
"step": 19
},
{
"epoch": 0.1834862385321101,
"grad_norm": 56021.69921875,
"learning_rate": 9.5e-06,
"loss": 0.6156,
"step": 20
},
{
"epoch": 0.1926605504587156,
"grad_norm": 60806.10546875,
"learning_rate": 1e-05,
"loss": 0.5981,
"step": 21
},
{
"epoch": 0.2018348623853211,
"grad_norm": 58582.4140625,
"learning_rate": 1.05e-05,
"loss": 0.5869,
"step": 22
},
{
"epoch": 0.21100917431192662,
"grad_norm": 59099.16015625,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.5732,
"step": 23
},
{
"epoch": 0.22018348623853212,
"grad_norm": 52971.81640625,
"learning_rate": 1.1500000000000002e-05,
"loss": 0.5587,
"step": 24
},
{
"epoch": 0.22935779816513763,
"grad_norm": 51724.43359375,
"learning_rate": 1.2e-05,
"loss": 0.5477,
"step": 25
},
{
"epoch": 0.23853211009174313,
"grad_norm": 58133.453125,
"learning_rate": 1.25e-05,
"loss": 0.5422,
"step": 26
},
{
"epoch": 0.24770642201834864,
"grad_norm": 54134.3359375,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.5319,
"step": 27
},
{
"epoch": 0.25688073394495414,
"grad_norm": 78195.7890625,
"learning_rate": 1.3500000000000001e-05,
"loss": 0.5259,
"step": 28
},
{
"epoch": 0.26605504587155965,
"grad_norm": 47713.20703125,
"learning_rate": 1.4000000000000001e-05,
"loss": 0.5105,
"step": 29
},
{
"epoch": 0.27522935779816515,
"grad_norm": 45838.0703125,
"learning_rate": 1.45e-05,
"loss": 0.499,
"step": 30
},
{
"epoch": 0.28440366972477066,
"grad_norm": 46735.58203125,
"learning_rate": 1.5e-05,
"loss": 0.4914,
"step": 31
},
{
"epoch": 0.29357798165137616,
"grad_norm": 45432.8828125,
"learning_rate": 1.55e-05,
"loss": 0.4783,
"step": 32
},
{
"epoch": 0.30275229357798167,
"grad_norm": 46758.58984375,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.4758,
"step": 33
},
{
"epoch": 0.3119266055045872,
"grad_norm": 45245.82421875,
"learning_rate": 1.65e-05,
"loss": 0.4608,
"step": 34
},
{
"epoch": 0.3211009174311927,
"grad_norm": 43336.75390625,
"learning_rate": 1.7000000000000003e-05,
"loss": 0.4567,
"step": 35
},
{
"epoch": 0.3302752293577982,
"grad_norm": 41445.90625,
"learning_rate": 1.75e-05,
"loss": 0.4465,
"step": 36
},
{
"epoch": 0.3394495412844037,
"grad_norm": 42278.80078125,
"learning_rate": 1.8e-05,
"loss": 0.4395,
"step": 37
},
{
"epoch": 0.3486238532110092,
"grad_norm": 42039.28515625,
"learning_rate": 1.85e-05,
"loss": 0.4289,
"step": 38
},
{
"epoch": 0.3577981651376147,
"grad_norm": 41411.61328125,
"learning_rate": 1.9e-05,
"loss": 0.423,
"step": 39
},
{
"epoch": 0.3669724770642202,
"grad_norm": 40133.84375,
"learning_rate": 1.9500000000000003e-05,
"loss": 0.4191,
"step": 40
},
{
"epoch": 0.3761467889908257,
"grad_norm": 39758.8828125,
"learning_rate": 2e-05,
"loss": 0.4071,
"step": 41
},
{
"epoch": 0.3853211009174312,
"grad_norm": 38760.36328125,
"learning_rate": 2.05e-05,
"loss": 0.3996,
"step": 42
},
{
"epoch": 0.3944954128440367,
"grad_norm": 38552.80078125,
"learning_rate": 2.1e-05,
"loss": 0.3918,
"step": 43
},
{
"epoch": 0.4036697247706422,
"grad_norm": 38561.7578125,
"learning_rate": 2.15e-05,
"loss": 0.3865,
"step": 44
},
{
"epoch": 0.41284403669724773,
"grad_norm": 37616.85546875,
"learning_rate": 2.2000000000000003e-05,
"loss": 0.3794,
"step": 45
},
{
"epoch": 0.42201834862385323,
"grad_norm": 37784.78515625,
"learning_rate": 2.25e-05,
"loss": 0.3745,
"step": 46
},
{
"epoch": 0.43119266055045874,
"grad_norm": 36332.9140625,
"learning_rate": 2.3000000000000003e-05,
"loss": 0.3711,
"step": 47
},
{
"epoch": 0.44036697247706424,
"grad_norm": 38899.73046875,
"learning_rate": 2.35e-05,
"loss": 0.3643,
"step": 48
},
{
"epoch": 0.44954128440366975,
"grad_norm": 36174.9765625,
"learning_rate": 2.4e-05,
"loss": 0.3575,
"step": 49
},
{
"epoch": 0.45871559633027525,
"grad_norm": 35699.83203125,
"learning_rate": 2.45e-05,
"loss": 0.347,
"step": 50
},
{
"epoch": 0.46788990825688076,
"grad_norm": 35562.3671875,
"learning_rate": 2.5e-05,
"loss": 0.3438,
"step": 51
},
{
"epoch": 0.47706422018348627,
"grad_norm": 35428.0625,
"learning_rate": 2.5500000000000003e-05,
"loss": 0.3325,
"step": 52
},
{
"epoch": 0.48623853211009177,
"grad_norm": 34396.80078125,
"learning_rate": 2.6000000000000002e-05,
"loss": 0.3302,
"step": 53
},
{
"epoch": 0.4954128440366973,
"grad_norm": 38376.41796875,
"learning_rate": 2.6500000000000004e-05,
"loss": 0.3218,
"step": 54
},
{
"epoch": 0.5045871559633027,
"grad_norm": 33996.23046875,
"learning_rate": 2.7000000000000002e-05,
"loss": 0.3131,
"step": 55
},
{
"epoch": 0.5137614678899083,
"grad_norm": 33070.55859375,
"learning_rate": 2.7500000000000004e-05,
"loss": 0.3129,
"step": 56
},
{
"epoch": 0.5229357798165137,
"grad_norm": 33200.44140625,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.3003,
"step": 57
},
{
"epoch": 0.5321100917431193,
"grad_norm": 33099.1015625,
"learning_rate": 2.8499999999999998e-05,
"loss": 0.2904,
"step": 58
},
{
"epoch": 0.5412844036697247,
"grad_norm": 31352.298828125,
"learning_rate": 2.9e-05,
"loss": 0.2911,
"step": 59
},
{
"epoch": 0.5504587155963303,
"grad_norm": 31965.087890625,
"learning_rate": 2.95e-05,
"loss": 0.2815,
"step": 60
},
{
"epoch": 0.5596330275229358,
"grad_norm": 30810.849609375,
"learning_rate": 3e-05,
"loss": 0.2818,
"step": 61
},
{
"epoch": 0.5688073394495413,
"grad_norm": 30034.779296875,
"learning_rate": 3.05e-05,
"loss": 0.2731,
"step": 62
},
{
"epoch": 0.5779816513761468,
"grad_norm": 30009.12109375,
"learning_rate": 3.1e-05,
"loss": 0.266,
"step": 63
},
{
"epoch": 0.5871559633027523,
"grad_norm": 29305.173828125,
"learning_rate": 3.15e-05,
"loss": 0.2609,
"step": 64
},
{
"epoch": 0.5963302752293578,
"grad_norm": 29081.853515625,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.2543,
"step": 65
},
{
"epoch": 0.6055045871559633,
"grad_norm": 28217.021484375,
"learning_rate": 3.2500000000000004e-05,
"loss": 0.252,
"step": 66
},
{
"epoch": 0.6146788990825688,
"grad_norm": 26448.12890625,
"learning_rate": 3.3e-05,
"loss": 0.2565,
"step": 67
},
{
"epoch": 0.6238532110091743,
"grad_norm": 27198.80859375,
"learning_rate": 3.35e-05,
"loss": 0.2342,
"step": 68
},
{
"epoch": 0.6330275229357798,
"grad_norm": 26946.30859375,
"learning_rate": 3.4000000000000007e-05,
"loss": 0.2273,
"step": 69
},
{
"epoch": 0.6422018348623854,
"grad_norm": 26236.7265625,
"learning_rate": 3.45e-05,
"loss": 0.2219,
"step": 70
},
{
"epoch": 0.6513761467889908,
"grad_norm": 25535.818359375,
"learning_rate": 3.5e-05,
"loss": 0.2197,
"step": 71
},
{
"epoch": 0.6605504587155964,
"grad_norm": 24871.8515625,
"learning_rate": 3.55e-05,
"loss": 0.2144,
"step": 72
},
{
"epoch": 0.6697247706422018,
"grad_norm": 24484.55078125,
"learning_rate": 3.6e-05,
"loss": 0.2066,
"step": 73
},
{
"epoch": 0.6788990825688074,
"grad_norm": 23655.677734375,
"learning_rate": 3.65e-05,
"loss": 0.206,
"step": 74
},
{
"epoch": 0.6880733944954128,
"grad_norm": 23129.076171875,
"learning_rate": 3.7e-05,
"loss": 0.1979,
"step": 75
},
{
"epoch": 0.6972477064220184,
"grad_norm": 21871.90625,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.1973,
"step": 76
},
{
"epoch": 0.7064220183486238,
"grad_norm": 21561.58984375,
"learning_rate": 3.8e-05,
"loss": 0.1935,
"step": 77
},
{
"epoch": 0.7155963302752294,
"grad_norm": 22210.29296875,
"learning_rate": 3.85e-05,
"loss": 0.1848,
"step": 78
},
{
"epoch": 0.7247706422018348,
"grad_norm": 20555.775390625,
"learning_rate": 3.9000000000000006e-05,
"loss": 0.1801,
"step": 79
},
{
"epoch": 0.7339449541284404,
"grad_norm": 20406.75,
"learning_rate": 3.9500000000000005e-05,
"loss": 0.1746,
"step": 80
},
{
"epoch": 0.7431192660550459,
"grad_norm": 19150.931640625,
"learning_rate": 4e-05,
"loss": 0.1688,
"step": 81
},
{
"epoch": 0.7522935779816514,
"grad_norm": 18422.65625,
"learning_rate": 4.05e-05,
"loss": 0.1661,
"step": 82
},
{
"epoch": 0.7614678899082569,
"grad_norm": 16746.20703125,
"learning_rate": 4.1e-05,
"loss": 0.1711,
"step": 83
},
{
"epoch": 0.7706422018348624,
"grad_norm": 18229.41015625,
"learning_rate": 4.15e-05,
"loss": 0.1646,
"step": 84
},
{
"epoch": 0.7798165137614679,
"grad_norm": 17156.267578125,
"learning_rate": 4.2e-05,
"loss": 0.1588,
"step": 85
},
{
"epoch": 0.7889908256880734,
"grad_norm": 15970.611328125,
"learning_rate": 4.25e-05,
"loss": 0.1547,
"step": 86
},
{
"epoch": 0.7981651376146789,
"grad_norm": 16997.103515625,
"learning_rate": 4.3e-05,
"loss": 0.1493,
"step": 87
},
{
"epoch": 0.8073394495412844,
"grad_norm": 15529.58984375,
"learning_rate": 4.35e-05,
"loss": 0.1439,
"step": 88
},
{
"epoch": 0.8165137614678899,
"grad_norm": 15099.9052734375,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.1432,
"step": 89
},
{
"epoch": 0.8256880733944955,
"grad_norm": 14261.3232421875,
"learning_rate": 4.4500000000000004e-05,
"loss": 0.1396,
"step": 90
},
{
"epoch": 0.8348623853211009,
"grad_norm": 13915.87890625,
"learning_rate": 4.5e-05,
"loss": 0.1317,
"step": 91
},
{
"epoch": 0.8440366972477065,
"grad_norm": 13518.7568359375,
"learning_rate": 4.55e-05,
"loss": 0.1289,
"step": 92
},
{
"epoch": 0.8532110091743119,
"grad_norm": 12170.1865234375,
"learning_rate": 4.600000000000001e-05,
"loss": 0.1371,
"step": 93
},
{
"epoch": 0.8623853211009175,
"grad_norm": 11745.9775390625,
"learning_rate": 4.6500000000000005e-05,
"loss": 0.1434,
"step": 94
},
{
"epoch": 0.8715596330275229,
"grad_norm": 12265.5556640625,
"learning_rate": 4.7e-05,
"loss": 0.1343,
"step": 95
},
{
"epoch": 0.8807339449541285,
"grad_norm": 11386.3291015625,
"learning_rate": 4.75e-05,
"loss": 0.1317,
"step": 96
},
{
"epoch": 0.8899082568807339,
"grad_norm": 10869.4599609375,
"learning_rate": 4.8e-05,
"loss": 0.1295,
"step": 97
},
{
"epoch": 0.8990825688073395,
"grad_norm": 10752.25,
"learning_rate": 4.85e-05,
"loss": 0.124,
"step": 98
},
{
"epoch": 0.908256880733945,
"grad_norm": 9687.6376953125,
"learning_rate": 4.9e-05,
"loss": 0.1271,
"step": 99
},
{
"epoch": 0.9174311926605505,
"grad_norm": 9697.15234375,
"learning_rate": 4.9500000000000004e-05,
"loss": 0.1194,
"step": 100
},
{
"epoch": 0.926605504587156,
"grad_norm": 9423.21875,
"learning_rate": 5e-05,
"loss": 0.1176,
"step": 101
},
{
"epoch": 0.9357798165137615,
"grad_norm": 10253.8564453125,
"learning_rate": 4.994949494949495e-05,
"loss": 0.121,
"step": 102
},
{
"epoch": 0.944954128440367,
"grad_norm": 9756.7646484375,
"learning_rate": 4.98989898989899e-05,
"loss": 0.1173,
"step": 103
},
{
"epoch": 0.9541284403669725,
"grad_norm": 8482.177734375,
"learning_rate": 4.984848484848485e-05,
"loss": 0.1233,
"step": 104
},
{
"epoch": 0.963302752293578,
"grad_norm": 7755.4892578125,
"learning_rate": 4.97979797979798e-05,
"loss": 0.1251,
"step": 105
},
{
"epoch": 0.9724770642201835,
"grad_norm": 8553.6689453125,
"learning_rate": 4.974747474747475e-05,
"loss": 0.1287,
"step": 106
},
{
"epoch": 0.981651376146789,
"grad_norm": 8794.58984375,
"learning_rate": 4.9696969696969694e-05,
"loss": 0.1178,
"step": 107
},
{
"epoch": 0.9908256880733946,
"grad_norm": 8048.97509765625,
"learning_rate": 4.964646464646465e-05,
"loss": 0.1112,
"step": 108
},
{
"epoch": 1.0,
"grad_norm": 24349.96484375,
"learning_rate": 4.9595959595959594e-05,
"loss": 0.112,
"step": 109
},
{
"epoch": 1.0,
"eval_f1_macro": 0.0,
"eval_f1_micro": 0.0,
"eval_loss": 0.1133684441447258,
"eval_precision": 0.0,
"eval_recall": 0.0,
"eval_runtime": 0.8578,
"eval_samples_per_second": 336.892,
"eval_steps_per_second": 15.154,
"step": 109
},
{
"epoch": 1.0091743119266054,
"grad_norm": 7603.02587890625,
"learning_rate": 4.9545454545454553e-05,
"loss": 0.1133,
"step": 110
},
{
"epoch": 1.018348623853211,
"grad_norm": 8602.8974609375,
"learning_rate": 4.94949494949495e-05,
"loss": 0.1192,
"step": 111
},
{
"epoch": 1.0275229357798166,
"grad_norm": 8388.9609375,
"learning_rate": 4.9444444444444446e-05,
"loss": 0.1246,
"step": 112
},
{
"epoch": 1.036697247706422,
"grad_norm": 9368.150390625,
"learning_rate": 4.93939393939394e-05,
"loss": 0.1089,
"step": 113
},
{
"epoch": 1.0458715596330275,
"grad_norm": 7053.083984375,
"learning_rate": 4.9343434343434346e-05,
"loss": 0.1042,
"step": 114
},
{
"epoch": 1.0550458715596331,
"grad_norm": 9539.36328125,
"learning_rate": 4.92929292929293e-05,
"loss": 0.1049,
"step": 115
},
{
"epoch": 1.0642201834862386,
"grad_norm": 6956.2763671875,
"learning_rate": 4.9242424242424245e-05,
"loss": 0.109,
"step": 116
},
{
"epoch": 1.073394495412844,
"grad_norm": 6597.9580078125,
"learning_rate": 4.919191919191919e-05,
"loss": 0.1124,
"step": 117
},
{
"epoch": 1.0825688073394495,
"grad_norm": 8124.76123046875,
"learning_rate": 4.9141414141414145e-05,
"loss": 0.1226,
"step": 118
},
{
"epoch": 1.091743119266055,
"grad_norm": 7030.619140625,
"learning_rate": 4.909090909090909e-05,
"loss": 0.0989,
"step": 119
},
{
"epoch": 1.1009174311926606,
"grad_norm": 9377.66796875,
"learning_rate": 4.9040404040404044e-05,
"loss": 0.1038,
"step": 120
},
{
"epoch": 1.110091743119266,
"grad_norm": 9298.802734375,
"learning_rate": 4.898989898989899e-05,
"loss": 0.1216,
"step": 121
},
{
"epoch": 1.1192660550458715,
"grad_norm": 7350.28369140625,
"learning_rate": 4.8939393939393944e-05,
"loss": 0.0915,
"step": 122
},
{
"epoch": 1.1284403669724772,
"grad_norm": 8066.943359375,
"learning_rate": 4.888888888888889e-05,
"loss": 0.1291,
"step": 123
},
{
"epoch": 1.1376146788990826,
"grad_norm": 8210.3095703125,
"learning_rate": 4.8838383838383836e-05,
"loss": 0.1196,
"step": 124
},
{
"epoch": 1.146788990825688,
"grad_norm": 7568.0234375,
"learning_rate": 4.878787878787879e-05,
"loss": 0.1037,
"step": 125
},
{
"epoch": 1.1559633027522935,
"grad_norm": 6394.8896484375,
"learning_rate": 4.8737373737373736e-05,
"loss": 0.0961,
"step": 126
},
{
"epoch": 1.165137614678899,
"grad_norm": 7549.9951171875,
"learning_rate": 4.868686868686869e-05,
"loss": 0.1075,
"step": 127
},
{
"epoch": 1.1743119266055047,
"grad_norm": 7983.36865234375,
"learning_rate": 4.863636363636364e-05,
"loss": 0.1055,
"step": 128
},
{
"epoch": 1.18348623853211,
"grad_norm": 7613.1455078125,
"learning_rate": 4.858585858585859e-05,
"loss": 0.1097,
"step": 129
},
{
"epoch": 1.1926605504587156,
"grad_norm": 7754.9228515625,
"learning_rate": 4.853535353535354e-05,
"loss": 0.1157,
"step": 130
},
{
"epoch": 1.2018348623853212,
"grad_norm": 8360.388671875,
"learning_rate": 4.848484848484849e-05,
"loss": 0.1019,
"step": 131
},
{
"epoch": 1.2110091743119267,
"grad_norm": 8300.9169921875,
"learning_rate": 4.843434343434344e-05,
"loss": 0.1098,
"step": 132
},
{
"epoch": 1.2201834862385321,
"grad_norm": 7554.8017578125,
"learning_rate": 4.838383838383839e-05,
"loss": 0.1075,
"step": 133
},
{
"epoch": 1.2293577981651376,
"grad_norm": 6510.427734375,
"learning_rate": 4.8333333333333334e-05,
"loss": 0.1041,
"step": 134
},
{
"epoch": 1.238532110091743,
"grad_norm": 5649.77978515625,
"learning_rate": 4.828282828282829e-05,
"loss": 0.1012,
"step": 135
},
{
"epoch": 1.2477064220183487,
"grad_norm": 7662.76513671875,
"learning_rate": 4.823232323232323e-05,
"loss": 0.1357,
"step": 136
},
{
"epoch": 1.2568807339449541,
"grad_norm": 7261.087890625,
"learning_rate": 4.8181818181818186e-05,
"loss": 0.1149,
"step": 137
},
{
"epoch": 1.2660550458715596,
"grad_norm": 6818.208984375,
"learning_rate": 4.813131313131313e-05,
"loss": 0.1067,
"step": 138
},
{
"epoch": 1.2752293577981653,
"grad_norm": 8839.0810546875,
"learning_rate": 4.808080808080808e-05,
"loss": 0.089,
"step": 139
},
{
"epoch": 1.2844036697247707,
"grad_norm": 8774.0615234375,
"learning_rate": 4.803030303030303e-05,
"loss": 0.1013,
"step": 140
},
{
"epoch": 1.2935779816513762,
"grad_norm": 7869.935546875,
"learning_rate": 4.797979797979798e-05,
"loss": 0.1187,
"step": 141
},
{
"epoch": 1.3027522935779816,
"grad_norm": 9808.5986328125,
"learning_rate": 4.792929292929293e-05,
"loss": 0.13,
"step": 142
},
{
"epoch": 1.311926605504587,
"grad_norm": 7122.400390625,
"learning_rate": 4.787878787878788e-05,
"loss": 0.1089,
"step": 143
},
{
"epoch": 1.3211009174311927,
"grad_norm": 6575.39892578125,
"learning_rate": 4.782828282828283e-05,
"loss": 0.1011,
"step": 144
},
{
"epoch": 1.3302752293577982,
"grad_norm": 8656.4345703125,
"learning_rate": 4.7777777777777784e-05,
"loss": 0.0998,
"step": 145
},
{
"epoch": 1.3394495412844036,
"grad_norm": 7247.53466796875,
"learning_rate": 4.772727272727273e-05,
"loss": 0.1255,
"step": 146
},
{
"epoch": 1.3486238532110093,
"grad_norm": 7125.67822265625,
"learning_rate": 4.7676767676767684e-05,
"loss": 0.1123,
"step": 147
},
{
"epoch": 1.3577981651376148,
"grad_norm": 7758.84375,
"learning_rate": 4.762626262626263e-05,
"loss": 0.1138,
"step": 148
},
{
"epoch": 1.3669724770642202,
"grad_norm": 6958.53173828125,
"learning_rate": 4.7575757575757576e-05,
"loss": 0.0968,
"step": 149
},
{
"epoch": 1.3761467889908257,
"grad_norm": 7246.337890625,
"learning_rate": 4.752525252525253e-05,
"loss": 0.0924,
"step": 150
},
{
"epoch": 1.385321100917431,
"grad_norm": 7635.5615234375,
"learning_rate": 4.7474747474747476e-05,
"loss": 0.1152,
"step": 151
},
{
"epoch": 1.3944954128440368,
"grad_norm": 7913.57080078125,
"learning_rate": 4.742424242424243e-05,
"loss": 0.0951,
"step": 152
},
{
"epoch": 1.4036697247706422,
"grad_norm": 8531.388671875,
"learning_rate": 4.7373737373737375e-05,
"loss": 0.1161,
"step": 153
},
{
"epoch": 1.4128440366972477,
"grad_norm": 9724.3291015625,
"learning_rate": 4.732323232323232e-05,
"loss": 0.1253,
"step": 154
},
{
"epoch": 1.4220183486238533,
"grad_norm": 6726.74365234375,
"learning_rate": 4.7272727272727275e-05,
"loss": 0.1037,
"step": 155
},
{
"epoch": 1.4311926605504588,
"grad_norm": 6435.69970703125,
"learning_rate": 4.722222222222222e-05,
"loss": 0.1036,
"step": 156
},
{
"epoch": 1.4403669724770642,
"grad_norm": 6851.59814453125,
"learning_rate": 4.7171717171717174e-05,
"loss": 0.0954,
"step": 157
},
{
"epoch": 1.4495412844036697,
"grad_norm": 7785.31640625,
"learning_rate": 4.712121212121212e-05,
"loss": 0.0996,
"step": 158
},
{
"epoch": 1.4587155963302751,
"grad_norm": 6317.5576171875,
"learning_rate": 4.7070707070707074e-05,
"loss": 0.0977,
"step": 159
},
{
"epoch": 1.4678899082568808,
"grad_norm": 8045.7197265625,
"learning_rate": 4.702020202020202e-05,
"loss": 0.1101,
"step": 160
},
{
"epoch": 1.4770642201834863,
"grad_norm": 7386.67529296875,
"learning_rate": 4.696969696969697e-05,
"loss": 0.1137,
"step": 161
},
{
"epoch": 1.4862385321100917,
"grad_norm": 8017.5791015625,
"learning_rate": 4.6919191919191926e-05,
"loss": 0.0959,
"step": 162
},
{
"epoch": 1.4954128440366974,
"grad_norm": 7198.42138671875,
"learning_rate": 4.686868686868687e-05,
"loss": 0.1005,
"step": 163
},
{
"epoch": 1.5045871559633026,
"grad_norm": 9139.7900390625,
"learning_rate": 4.681818181818182e-05,
"loss": 0.0918,
"step": 164
},
{
"epoch": 1.5137614678899083,
"grad_norm": 6384.1640625,
"learning_rate": 4.676767676767677e-05,
"loss": 0.0948,
"step": 165
},
{
"epoch": 1.5229357798165137,
"grad_norm": 5053.224609375,
"learning_rate": 4.671717171717172e-05,
"loss": 0.1056,
"step": 166
},
{
"epoch": 1.5321100917431192,
"grad_norm": 9213.654296875,
"learning_rate": 4.666666666666667e-05,
"loss": 0.0931,
"step": 167
},
{
"epoch": 1.5412844036697249,
"grad_norm": 7414.76171875,
"learning_rate": 4.661616161616162e-05,
"loss": 0.1123,
"step": 168
},
{
"epoch": 1.5504587155963303,
"grad_norm": 6406.48583984375,
"learning_rate": 4.656565656565657e-05,
"loss": 0.0992,
"step": 169
},
{
"epoch": 1.5596330275229358,
"grad_norm": 7213.58837890625,
"learning_rate": 4.651515151515152e-05,
"loss": 0.0952,
"step": 170
},
{
"epoch": 1.5688073394495414,
"grad_norm": 7537.673828125,
"learning_rate": 4.6464646464646464e-05,
"loss": 0.0961,
"step": 171
},
{
"epoch": 1.5779816513761467,
"grad_norm": 7876.294921875,
"learning_rate": 4.641414141414142e-05,
"loss": 0.1308,
"step": 172
},
{
"epoch": 1.5871559633027523,
"grad_norm": 9550.125,
"learning_rate": 4.636363636363636e-05,
"loss": 0.0941,
"step": 173
},
{
"epoch": 1.5963302752293578,
"grad_norm": 6364.3330078125,
"learning_rate": 4.6313131313131316e-05,
"loss": 0.0976,
"step": 174
},
{
"epoch": 1.6055045871559632,
"grad_norm": 6976.2138671875,
"learning_rate": 4.626262626262626e-05,
"loss": 0.091,
"step": 175
},
{
"epoch": 1.614678899082569,
"grad_norm": 7259.40234375,
"learning_rate": 4.621212121212121e-05,
"loss": 0.1096,
"step": 176
},
{
"epoch": 1.6238532110091743,
"grad_norm": 8176.20849609375,
"learning_rate": 4.616161616161616e-05,
"loss": 0.1029,
"step": 177
},
{
"epoch": 1.6330275229357798,
"grad_norm": 5628.04345703125,
"learning_rate": 4.6111111111111115e-05,
"loss": 0.1008,
"step": 178
},
{
"epoch": 1.6422018348623855,
"grad_norm": 6802.91064453125,
"learning_rate": 4.606060606060607e-05,
"loss": 0.1014,
"step": 179
},
{
"epoch": 1.6513761467889907,
"grad_norm": 14422.4482421875,
"learning_rate": 4.6010101010101015e-05,
"loss": 0.0912,
"step": 180
},
{
"epoch": 1.6605504587155964,
"grad_norm": 8855.7744140625,
"learning_rate": 4.595959595959596e-05,
"loss": 0.0983,
"step": 181
},
{
"epoch": 1.6697247706422018,
"grad_norm": 7411.93603515625,
"learning_rate": 4.5909090909090914e-05,
"loss": 0.1099,
"step": 182
},
{
"epoch": 1.6788990825688073,
"grad_norm": 8138.41748046875,
"learning_rate": 4.585858585858586e-05,
"loss": 0.1054,
"step": 183
},
{
"epoch": 1.688073394495413,
"grad_norm": 6157.80908203125,
"learning_rate": 4.5808080808080814e-05,
"loss": 0.1054,
"step": 184
},
{
"epoch": 1.6972477064220184,
"grad_norm": 6791.07421875,
"learning_rate": 4.575757575757576e-05,
"loss": 0.1022,
"step": 185
},
{
"epoch": 1.7064220183486238,
"grad_norm": 9991.0537109375,
"learning_rate": 4.5707070707070706e-05,
"loss": 0.1131,
"step": 186
},
{
"epoch": 1.7155963302752295,
"grad_norm": 7240.28955078125,
"learning_rate": 4.565656565656566e-05,
"loss": 0.1005,
"step": 187
},
{
"epoch": 1.7247706422018347,
"grad_norm": 6442.6259765625,
"learning_rate": 4.5606060606060606e-05,
"loss": 0.1015,
"step": 188
},
{
"epoch": 1.7339449541284404,
"grad_norm": 8573.3955078125,
"learning_rate": 4.555555555555556e-05,
"loss": 0.1176,
"step": 189
},
{
"epoch": 1.7431192660550459,
"grad_norm": 7330.88525390625,
"learning_rate": 4.5505050505050505e-05,
"loss": 0.0923,
"step": 190
},
{
"epoch": 1.7522935779816513,
"grad_norm": 7870.916015625,
"learning_rate": 4.545454545454546e-05,
"loss": 0.1113,
"step": 191
},
{
"epoch": 1.761467889908257,
"grad_norm": 7789.8564453125,
"learning_rate": 4.5404040404040405e-05,
"loss": 0.1062,
"step": 192
},
{
"epoch": 1.7706422018348624,
"grad_norm": 7983.59326171875,
"learning_rate": 4.535353535353535e-05,
"loss": 0.1078,
"step": 193
},
{
"epoch": 1.7798165137614679,
"grad_norm": 6533.43994140625,
"learning_rate": 4.5303030303030304e-05,
"loss": 0.1033,
"step": 194
},
{
"epoch": 1.7889908256880735,
"grad_norm": 7317.1318359375,
"learning_rate": 4.525252525252526e-05,
"loss": 0.1106,
"step": 195
},
{
"epoch": 1.7981651376146788,
"grad_norm": 7829.7880859375,
"learning_rate": 4.5202020202020204e-05,
"loss": 0.1163,
"step": 196
},
{
"epoch": 1.8073394495412844,
"grad_norm": 7524.41357421875,
"learning_rate": 4.515151515151516e-05,
"loss": 0.1043,
"step": 197
},
{
"epoch": 1.81651376146789,
"grad_norm": 7487.89892578125,
"learning_rate": 4.51010101010101e-05,
"loss": 0.1091,
"step": 198
},
{
"epoch": 1.8256880733944953,
"grad_norm": 6964.3154296875,
"learning_rate": 4.5050505050505056e-05,
"loss": 0.0986,
"step": 199
},
{
"epoch": 1.834862385321101,
"grad_norm": 7087.05029296875,
"learning_rate": 4.5e-05,
"loss": 0.0997,
"step": 200
},
{
"epoch": 1.8440366972477065,
"grad_norm": 7148.7578125,
"learning_rate": 4.494949494949495e-05,
"loss": 0.0989,
"step": 201
},
{
"epoch": 1.853211009174312,
"grad_norm": 7850.2470703125,
"learning_rate": 4.48989898989899e-05,
"loss": 0.0957,
"step": 202
},
{
"epoch": 1.8623853211009176,
"grad_norm": 7752.841796875,
"learning_rate": 4.484848484848485e-05,
"loss": 0.1292,
"step": 203
},
{
"epoch": 1.8715596330275228,
"grad_norm": 6098.35595703125,
"learning_rate": 4.47979797979798e-05,
"loss": 0.1001,
"step": 204
},
{
"epoch": 1.8807339449541285,
"grad_norm": 7094.8173828125,
"learning_rate": 4.474747474747475e-05,
"loss": 0.1286,
"step": 205
},
{
"epoch": 1.889908256880734,
"grad_norm": 7456.2265625,
"learning_rate": 4.46969696969697e-05,
"loss": 0.1256,
"step": 206
},
{
"epoch": 1.8990825688073394,
"grad_norm": 8136.77197265625,
"learning_rate": 4.464646464646465e-05,
"loss": 0.1014,
"step": 207
},
{
"epoch": 1.908256880733945,
"grad_norm": 7520.1650390625,
"learning_rate": 4.4595959595959594e-05,
"loss": 0.0919,
"step": 208
},
{
"epoch": 1.9174311926605505,
"grad_norm": 8064.79296875,
"learning_rate": 4.454545454545455e-05,
"loss": 0.0985,
"step": 209
},
{
"epoch": 1.926605504587156,
"grad_norm": 6153.19677734375,
"learning_rate": 4.4494949494949493e-05,
"loss": 0.1101,
"step": 210
},
{
"epoch": 1.9357798165137616,
"grad_norm": 8089.42431640625,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.0988,
"step": 211
},
{
"epoch": 1.9449541284403669,
"grad_norm": 8096.140625,
"learning_rate": 4.43939393939394e-05,
"loss": 0.0992,
"step": 212
},
{
"epoch": 1.9541284403669725,
"grad_norm": 7865.43798828125,
"learning_rate": 4.4343434343434346e-05,
"loss": 0.0945,
"step": 213
},
{
"epoch": 1.963302752293578,
"grad_norm": 6380.67724609375,
"learning_rate": 4.42929292929293e-05,
"loss": 0.1073,
"step": 214
},
{
"epoch": 1.9724770642201834,
"grad_norm": 8548.4345703125,
"learning_rate": 4.4242424242424246e-05,
"loss": 0.1121,
"step": 215
},
{
"epoch": 1.981651376146789,
"grad_norm": 8465.8466796875,
"learning_rate": 4.41919191919192e-05,
"loss": 0.1078,
"step": 216
},
{
"epoch": 1.9908256880733946,
"grad_norm": 8595.419921875,
"learning_rate": 4.4141414141414145e-05,
"loss": 0.0924,
"step": 217
},
{
"epoch": 2.0,
"grad_norm": 22229.833984375,
"learning_rate": 4.409090909090909e-05,
"loss": 0.0785,
"step": 218
},
{
"epoch": 2.0,
"eval_f1_macro": 0.0,
"eval_f1_micro": 0.0,
"eval_loss": 0.10520372539758682,
"eval_precision": 0.0,
"eval_recall": 0.0,
"eval_runtime": 0.8127,
"eval_samples_per_second": 355.611,
"eval_steps_per_second": 15.996,
"step": 218
}
],
"logging_steps": 1,
"max_steps": 1090,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1366134890434560.0,
"train_batch_size": 24,
"trial_name": null,
"trial_params": null
}