prxy5608's picture
Training in progress, step 200, checkpoint
998c936 verified
{
"best_metric": 0.8282029628753662,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 2.150537634408602,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.010752688172043012,
"grad_norm": 6.99516487121582,
"learning_rate": 5e-06,
"loss": 3.1175,
"step": 1
},
{
"epoch": 0.010752688172043012,
"eval_loss": 4.535027027130127,
"eval_runtime": 14.1563,
"eval_samples_per_second": 11.09,
"eval_steps_per_second": 5.581,
"step": 1
},
{
"epoch": 0.021505376344086023,
"grad_norm": 10.445697784423828,
"learning_rate": 1e-05,
"loss": 3.5728,
"step": 2
},
{
"epoch": 0.03225806451612903,
"grad_norm": 12.358306884765625,
"learning_rate": 1.5e-05,
"loss": 3.9348,
"step": 3
},
{
"epoch": 0.043010752688172046,
"grad_norm": 10.28347396850586,
"learning_rate": 2e-05,
"loss": 3.2841,
"step": 4
},
{
"epoch": 0.053763440860215055,
"grad_norm": 10.178171157836914,
"learning_rate": 2.5e-05,
"loss": 3.2135,
"step": 5
},
{
"epoch": 0.06451612903225806,
"grad_norm": 7.281465530395508,
"learning_rate": 3e-05,
"loss": 2.7599,
"step": 6
},
{
"epoch": 0.07526881720430108,
"grad_norm": 9.009861946105957,
"learning_rate": 3.5e-05,
"loss": 2.4959,
"step": 7
},
{
"epoch": 0.08602150537634409,
"grad_norm": 13.042397499084473,
"learning_rate": 4e-05,
"loss": 2.2706,
"step": 8
},
{
"epoch": 0.0967741935483871,
"grad_norm": 7.2071638107299805,
"learning_rate": 4.5e-05,
"loss": 1.94,
"step": 9
},
{
"epoch": 0.10752688172043011,
"grad_norm": 8.721714973449707,
"learning_rate": 5e-05,
"loss": 1.6717,
"step": 10
},
{
"epoch": 0.11827956989247312,
"grad_norm": 9.722580909729004,
"learning_rate": 5.500000000000001e-05,
"loss": 1.821,
"step": 11
},
{
"epoch": 0.12903225806451613,
"grad_norm": 8.217576026916504,
"learning_rate": 6e-05,
"loss": 1.58,
"step": 12
},
{
"epoch": 0.13978494623655913,
"grad_norm": 6.87364387512207,
"learning_rate": 6.500000000000001e-05,
"loss": 1.3209,
"step": 13
},
{
"epoch": 0.15053763440860216,
"grad_norm": 8.912040710449219,
"learning_rate": 7e-05,
"loss": 1.7184,
"step": 14
},
{
"epoch": 0.16129032258064516,
"grad_norm": 7.484908580780029,
"learning_rate": 7.500000000000001e-05,
"loss": 1.0966,
"step": 15
},
{
"epoch": 0.17204301075268819,
"grad_norm": 9.977514266967773,
"learning_rate": 8e-05,
"loss": 1.86,
"step": 16
},
{
"epoch": 0.1827956989247312,
"grad_norm": 6.212277412414551,
"learning_rate": 8.5e-05,
"loss": 1.2687,
"step": 17
},
{
"epoch": 0.1935483870967742,
"grad_norm": 6.920554161071777,
"learning_rate": 9e-05,
"loss": 1.1078,
"step": 18
},
{
"epoch": 0.20430107526881722,
"grad_norm": 6.519954681396484,
"learning_rate": 9.5e-05,
"loss": 1.0365,
"step": 19
},
{
"epoch": 0.21505376344086022,
"grad_norm": 9.026784896850586,
"learning_rate": 0.0001,
"loss": 1.2932,
"step": 20
},
{
"epoch": 0.22580645161290322,
"grad_norm": 8.984053611755371,
"learning_rate": 9.999238475781957e-05,
"loss": 1.0453,
"step": 21
},
{
"epoch": 0.23655913978494625,
"grad_norm": 7.758922100067139,
"learning_rate": 9.99695413509548e-05,
"loss": 1.2444,
"step": 22
},
{
"epoch": 0.24731182795698925,
"grad_norm": 7.274909496307373,
"learning_rate": 9.99314767377287e-05,
"loss": 0.9704,
"step": 23
},
{
"epoch": 0.25806451612903225,
"grad_norm": 8.09979248046875,
"learning_rate": 9.987820251299122e-05,
"loss": 2.5905,
"step": 24
},
{
"epoch": 0.26881720430107525,
"grad_norm": 5.225498676300049,
"learning_rate": 9.980973490458728e-05,
"loss": 2.3784,
"step": 25
},
{
"epoch": 0.27956989247311825,
"grad_norm": 4.132240295410156,
"learning_rate": 9.972609476841367e-05,
"loss": 1.9521,
"step": 26
},
{
"epoch": 0.2903225806451613,
"grad_norm": 4.1804656982421875,
"learning_rate": 9.962730758206611e-05,
"loss": 2.1076,
"step": 27
},
{
"epoch": 0.3010752688172043,
"grad_norm": 3.6776556968688965,
"learning_rate": 9.951340343707852e-05,
"loss": 1.9513,
"step": 28
},
{
"epoch": 0.3118279569892473,
"grad_norm": 7.695979595184326,
"learning_rate": 9.938441702975689e-05,
"loss": 1.6991,
"step": 29
},
{
"epoch": 0.3225806451612903,
"grad_norm": 6.833826065063477,
"learning_rate": 9.924038765061042e-05,
"loss": 1.3824,
"step": 30
},
{
"epoch": 0.3333333333333333,
"grad_norm": 3.845757007598877,
"learning_rate": 9.908135917238321e-05,
"loss": 1.4698,
"step": 31
},
{
"epoch": 0.34408602150537637,
"grad_norm": 3.1820149421691895,
"learning_rate": 9.890738003669029e-05,
"loss": 1.1889,
"step": 32
},
{
"epoch": 0.3548387096774194,
"grad_norm": 4.390655517578125,
"learning_rate": 9.871850323926177e-05,
"loss": 1.1593,
"step": 33
},
{
"epoch": 0.3655913978494624,
"grad_norm": 4.404603958129883,
"learning_rate": 9.851478631379982e-05,
"loss": 0.9826,
"step": 34
},
{
"epoch": 0.3763440860215054,
"grad_norm": 3.859941005706787,
"learning_rate": 9.829629131445342e-05,
"loss": 1.0486,
"step": 35
},
{
"epoch": 0.3870967741935484,
"grad_norm": 7.225592613220215,
"learning_rate": 9.806308479691595e-05,
"loss": 1.0135,
"step": 36
},
{
"epoch": 0.3978494623655914,
"grad_norm": 4.242331504821777,
"learning_rate": 9.781523779815179e-05,
"loss": 0.8687,
"step": 37
},
{
"epoch": 0.40860215053763443,
"grad_norm": 4.794056415557861,
"learning_rate": 9.755282581475769e-05,
"loss": 0.9721,
"step": 38
},
{
"epoch": 0.41935483870967744,
"grad_norm": 3.9935474395751953,
"learning_rate": 9.727592877996585e-05,
"loss": 0.8433,
"step": 39
},
{
"epoch": 0.43010752688172044,
"grad_norm": 3.3857905864715576,
"learning_rate": 9.698463103929542e-05,
"loss": 0.7633,
"step": 40
},
{
"epoch": 0.44086021505376344,
"grad_norm": 3.6143317222595215,
"learning_rate": 9.667902132486009e-05,
"loss": 1.0136,
"step": 41
},
{
"epoch": 0.45161290322580644,
"grad_norm": 3.9374451637268066,
"learning_rate": 9.635919272833938e-05,
"loss": 0.915,
"step": 42
},
{
"epoch": 0.46236559139784944,
"grad_norm": 6.313716411590576,
"learning_rate": 9.602524267262203e-05,
"loss": 1.191,
"step": 43
},
{
"epoch": 0.4731182795698925,
"grad_norm": 3.7929749488830566,
"learning_rate": 9.567727288213005e-05,
"loss": 1.0602,
"step": 44
},
{
"epoch": 0.4838709677419355,
"grad_norm": 5.400207996368408,
"learning_rate": 9.53153893518325e-05,
"loss": 0.8725,
"step": 45
},
{
"epoch": 0.4946236559139785,
"grad_norm": 3.7726528644561768,
"learning_rate": 9.493970231495835e-05,
"loss": 0.5123,
"step": 46
},
{
"epoch": 0.5053763440860215,
"grad_norm": 3.6227760314941406,
"learning_rate": 9.45503262094184e-05,
"loss": 2.2086,
"step": 47
},
{
"epoch": 0.5161290322580645,
"grad_norm": 3.246999979019165,
"learning_rate": 9.414737964294636e-05,
"loss": 2.3014,
"step": 48
},
{
"epoch": 0.5268817204301075,
"grad_norm": 4.355656623840332,
"learning_rate": 9.373098535696979e-05,
"loss": 2.0727,
"step": 49
},
{
"epoch": 0.5376344086021505,
"grad_norm": 2.746245861053467,
"learning_rate": 9.330127018922194e-05,
"loss": 1.5312,
"step": 50
},
{
"epoch": 0.5376344086021505,
"eval_loss": 1.3152074813842773,
"eval_runtime": 14.4699,
"eval_samples_per_second": 10.85,
"eval_steps_per_second": 5.46,
"step": 50
},
{
"epoch": 0.5483870967741935,
"grad_norm": 3.6186563968658447,
"learning_rate": 9.285836503510562e-05,
"loss": 1.7819,
"step": 51
},
{
"epoch": 0.5591397849462365,
"grad_norm": 3.9859840869903564,
"learning_rate": 9.24024048078213e-05,
"loss": 1.2978,
"step": 52
},
{
"epoch": 0.5698924731182796,
"grad_norm": 3.9932148456573486,
"learning_rate": 9.193352839727121e-05,
"loss": 1.447,
"step": 53
},
{
"epoch": 0.5806451612903226,
"grad_norm": 3.5936036109924316,
"learning_rate": 9.145187862775209e-05,
"loss": 1.3967,
"step": 54
},
{
"epoch": 0.5913978494623656,
"grad_norm": 3.375316858291626,
"learning_rate": 9.09576022144496e-05,
"loss": 1.0998,
"step": 55
},
{
"epoch": 0.6021505376344086,
"grad_norm": 3.807166814804077,
"learning_rate": 9.045084971874738e-05,
"loss": 1.1404,
"step": 56
},
{
"epoch": 0.6129032258064516,
"grad_norm": 3.084326982498169,
"learning_rate": 8.993177550236464e-05,
"loss": 0.988,
"step": 57
},
{
"epoch": 0.6236559139784946,
"grad_norm": 3.7563459873199463,
"learning_rate": 8.940053768033609e-05,
"loss": 1.0191,
"step": 58
},
{
"epoch": 0.6344086021505376,
"grad_norm": 4.25908899307251,
"learning_rate": 8.885729807284856e-05,
"loss": 0.8702,
"step": 59
},
{
"epoch": 0.6451612903225806,
"grad_norm": 3.750455617904663,
"learning_rate": 8.83022221559489e-05,
"loss": 0.8746,
"step": 60
},
{
"epoch": 0.6559139784946236,
"grad_norm": 3.6707611083984375,
"learning_rate": 8.773547901113862e-05,
"loss": 0.9337,
"step": 61
},
{
"epoch": 0.6666666666666666,
"grad_norm": 3.612056016921997,
"learning_rate": 8.715724127386972e-05,
"loss": 1.1718,
"step": 62
},
{
"epoch": 0.6774193548387096,
"grad_norm": 4.298011779785156,
"learning_rate": 8.656768508095853e-05,
"loss": 0.628,
"step": 63
},
{
"epoch": 0.6881720430107527,
"grad_norm": 4.248507022857666,
"learning_rate": 8.596699001693255e-05,
"loss": 0.6906,
"step": 64
},
{
"epoch": 0.6989247311827957,
"grad_norm": 3.890913963317871,
"learning_rate": 8.535533905932738e-05,
"loss": 1.0229,
"step": 65
},
{
"epoch": 0.7096774193548387,
"grad_norm": 4.77676248550415,
"learning_rate": 8.473291852294987e-05,
"loss": 0.8174,
"step": 66
},
{
"epoch": 0.7204301075268817,
"grad_norm": 4.285477161407471,
"learning_rate": 8.409991800312493e-05,
"loss": 0.8005,
"step": 67
},
{
"epoch": 0.7311827956989247,
"grad_norm": 3.8125786781311035,
"learning_rate": 8.345653031794292e-05,
"loss": 0.5634,
"step": 68
},
{
"epoch": 0.7419354838709677,
"grad_norm": 4.32075309753418,
"learning_rate": 8.280295144952536e-05,
"loss": 0.6793,
"step": 69
},
{
"epoch": 0.7526881720430108,
"grad_norm": 3.8095755577087402,
"learning_rate": 8.213938048432697e-05,
"loss": 2.0449,
"step": 70
},
{
"epoch": 0.7634408602150538,
"grad_norm": 3.9818124771118164,
"learning_rate": 8.146601955249188e-05,
"loss": 1.9486,
"step": 71
},
{
"epoch": 0.7741935483870968,
"grad_norm": 3.4325692653656006,
"learning_rate": 8.07830737662829e-05,
"loss": 1.8445,
"step": 72
},
{
"epoch": 0.7849462365591398,
"grad_norm": 3.9117205142974854,
"learning_rate": 8.009075115760243e-05,
"loss": 1.9626,
"step": 73
},
{
"epoch": 0.7956989247311828,
"grad_norm": 3.40067195892334,
"learning_rate": 7.938926261462366e-05,
"loss": 1.3113,
"step": 74
},
{
"epoch": 0.8064516129032258,
"grad_norm": 3.8196802139282227,
"learning_rate": 7.86788218175523e-05,
"loss": 1.7274,
"step": 75
},
{
"epoch": 0.8172043010752689,
"grad_norm": 4.589756488800049,
"learning_rate": 7.795964517353735e-05,
"loss": 1.7595,
"step": 76
},
{
"epoch": 0.8279569892473119,
"grad_norm": 4.156907081604004,
"learning_rate": 7.723195175075136e-05,
"loss": 1.5407,
"step": 77
},
{
"epoch": 0.8387096774193549,
"grad_norm": 3.101147174835205,
"learning_rate": 7.649596321166024e-05,
"loss": 0.9117,
"step": 78
},
{
"epoch": 0.8494623655913979,
"grad_norm": 2.96360182762146,
"learning_rate": 7.575190374550272e-05,
"loss": 1.1026,
"step": 79
},
{
"epoch": 0.8602150537634409,
"grad_norm": 2.9797332286834717,
"learning_rate": 7.500000000000001e-05,
"loss": 0.8541,
"step": 80
},
{
"epoch": 0.8709677419354839,
"grad_norm": 3.0931079387664795,
"learning_rate": 7.424048101231686e-05,
"loss": 0.8875,
"step": 81
},
{
"epoch": 0.8817204301075269,
"grad_norm": 2.8962762355804443,
"learning_rate": 7.347357813929454e-05,
"loss": 0.9001,
"step": 82
},
{
"epoch": 0.8924731182795699,
"grad_norm": 3.5031051635742188,
"learning_rate": 7.269952498697734e-05,
"loss": 0.8276,
"step": 83
},
{
"epoch": 0.9032258064516129,
"grad_norm": 3.142409086227417,
"learning_rate": 7.191855733945387e-05,
"loss": 0.823,
"step": 84
},
{
"epoch": 0.9139784946236559,
"grad_norm": 2.962707757949829,
"learning_rate": 7.113091308703498e-05,
"loss": 0.8578,
"step": 85
},
{
"epoch": 0.9247311827956989,
"grad_norm": 3.3335912227630615,
"learning_rate": 7.033683215379002e-05,
"loss": 0.8609,
"step": 86
},
{
"epoch": 0.9354838709677419,
"grad_norm": 4.589385509490967,
"learning_rate": 6.953655642446368e-05,
"loss": 1.0085,
"step": 87
},
{
"epoch": 0.946236559139785,
"grad_norm": 3.5878801345825195,
"learning_rate": 6.873032967079561e-05,
"loss": 0.5649,
"step": 88
},
{
"epoch": 0.956989247311828,
"grad_norm": 3.286903142929077,
"learning_rate": 6.7918397477265e-05,
"loss": 0.4883,
"step": 89
},
{
"epoch": 0.967741935483871,
"grad_norm": 3.4730312824249268,
"learning_rate": 6.710100716628344e-05,
"loss": 0.7835,
"step": 90
},
{
"epoch": 0.978494623655914,
"grad_norm": 3.7954018115997314,
"learning_rate": 6.627840772285784e-05,
"loss": 0.7178,
"step": 91
},
{
"epoch": 0.989247311827957,
"grad_norm": 3.0298449993133545,
"learning_rate": 6.545084971874738e-05,
"loss": 0.4597,
"step": 92
},
{
"epoch": 1.0,
"grad_norm": 11.850226402282715,
"learning_rate": 6.461858523613684e-05,
"loss": 1.5025,
"step": 93
},
{
"epoch": 1.010752688172043,
"grad_norm": 6.41492223739624,
"learning_rate": 6.378186779084995e-05,
"loss": 1.8111,
"step": 94
},
{
"epoch": 1.021505376344086,
"grad_norm": 3.9378912448883057,
"learning_rate": 6.294095225512603e-05,
"loss": 1.5149,
"step": 95
},
{
"epoch": 1.032258064516129,
"grad_norm": 3.533281087875366,
"learning_rate": 6.209609477998338e-05,
"loss": 1.4312,
"step": 96
},
{
"epoch": 1.043010752688172,
"grad_norm": 3.780578136444092,
"learning_rate": 6.124755271719325e-05,
"loss": 1.2979,
"step": 97
},
{
"epoch": 1.053763440860215,
"grad_norm": 2.7388339042663574,
"learning_rate": 6.0395584540887963e-05,
"loss": 0.988,
"step": 98
},
{
"epoch": 1.064516129032258,
"grad_norm": 2.726057767868042,
"learning_rate": 5.9540449768827246e-05,
"loss": 1.0731,
"step": 99
},
{
"epoch": 1.075268817204301,
"grad_norm": 4.266610622406006,
"learning_rate": 5.868240888334653e-05,
"loss": 1.0392,
"step": 100
},
{
"epoch": 1.075268817204301,
"eval_loss": 1.115802526473999,
"eval_runtime": 14.4453,
"eval_samples_per_second": 10.869,
"eval_steps_per_second": 5.469,
"step": 100
},
{
"epoch": 1.086021505376344,
"grad_norm": 3.880851984024048,
"learning_rate": 5.782172325201155e-05,
"loss": 0.7542,
"step": 101
},
{
"epoch": 1.096774193548387,
"grad_norm": 2.7799625396728516,
"learning_rate": 5.695865504800327e-05,
"loss": 0.6733,
"step": 102
},
{
"epoch": 1.10752688172043,
"grad_norm": 2.877377986907959,
"learning_rate": 5.6093467170257374e-05,
"loss": 0.7712,
"step": 103
},
{
"epoch": 1.118279569892473,
"grad_norm": 2.6312854290008545,
"learning_rate": 5.522642316338268e-05,
"loss": 0.5234,
"step": 104
},
{
"epoch": 1.129032258064516,
"grad_norm": 2.5732200145721436,
"learning_rate": 5.435778713738292e-05,
"loss": 0.4727,
"step": 105
},
{
"epoch": 1.139784946236559,
"grad_norm": 2.3443665504455566,
"learning_rate": 5.348782368720626e-05,
"loss": 0.5915,
"step": 106
},
{
"epoch": 1.1505376344086022,
"grad_norm": 2.900613784790039,
"learning_rate": 5.26167978121472e-05,
"loss": 0.4433,
"step": 107
},
{
"epoch": 1.1612903225806452,
"grad_norm": 2.779599666595459,
"learning_rate": 5.174497483512506e-05,
"loss": 0.4696,
"step": 108
},
{
"epoch": 1.1720430107526882,
"grad_norm": 3.2906293869018555,
"learning_rate": 5.0872620321864185e-05,
"loss": 0.6608,
"step": 109
},
{
"epoch": 1.1827956989247312,
"grad_norm": 3.181580066680908,
"learning_rate": 5e-05,
"loss": 0.5271,
"step": 110
},
{
"epoch": 1.1935483870967742,
"grad_norm": 3.3558270931243896,
"learning_rate": 4.912737967813583e-05,
"loss": 0.4357,
"step": 111
},
{
"epoch": 1.2043010752688172,
"grad_norm": 2.8602147102355957,
"learning_rate": 4.825502516487497e-05,
"loss": 0.4805,
"step": 112
},
{
"epoch": 1.2150537634408602,
"grad_norm": 3.4753870964050293,
"learning_rate": 4.738320218785281e-05,
"loss": 0.5753,
"step": 113
},
{
"epoch": 1.2258064516129032,
"grad_norm": 2.4663901329040527,
"learning_rate": 4.6512176312793736e-05,
"loss": 0.319,
"step": 114
},
{
"epoch": 1.2365591397849462,
"grad_norm": 4.225439548492432,
"learning_rate": 4.564221286261709e-05,
"loss": 0.5583,
"step": 115
},
{
"epoch": 1.2473118279569892,
"grad_norm": 2.3677308559417725,
"learning_rate": 4.477357683661734e-05,
"loss": 0.3289,
"step": 116
},
{
"epoch": 1.2580645161290323,
"grad_norm": 3.1596832275390625,
"learning_rate": 4.390653282974264e-05,
"loss": 1.6511,
"step": 117
},
{
"epoch": 1.2688172043010753,
"grad_norm": 3.4493701457977295,
"learning_rate": 4.3041344951996746e-05,
"loss": 1.6611,
"step": 118
},
{
"epoch": 1.2795698924731183,
"grad_norm": 3.0767276287078857,
"learning_rate": 4.2178276747988446e-05,
"loss": 1.4595,
"step": 119
},
{
"epoch": 1.2903225806451613,
"grad_norm": 3.6605751514434814,
"learning_rate": 4.131759111665349e-05,
"loss": 0.9704,
"step": 120
},
{
"epoch": 1.3010752688172043,
"grad_norm": 4.2374420166015625,
"learning_rate": 4.045955023117276e-05,
"loss": 1.2689,
"step": 121
},
{
"epoch": 1.3118279569892473,
"grad_norm": 3.0131664276123047,
"learning_rate": 3.960441545911204e-05,
"loss": 0.9044,
"step": 122
},
{
"epoch": 1.3225806451612903,
"grad_norm": 2.836139678955078,
"learning_rate": 3.875244728280676e-05,
"loss": 0.9114,
"step": 123
},
{
"epoch": 1.3333333333333333,
"grad_norm": 3.379606246948242,
"learning_rate": 3.790390522001662e-05,
"loss": 1.2156,
"step": 124
},
{
"epoch": 1.3440860215053765,
"grad_norm": 3.377495050430298,
"learning_rate": 3.705904774487396e-05,
"loss": 0.6492,
"step": 125
},
{
"epoch": 1.3548387096774195,
"grad_norm": 3.2035529613494873,
"learning_rate": 3.6218132209150045e-05,
"loss": 0.728,
"step": 126
},
{
"epoch": 1.3655913978494625,
"grad_norm": 2.9157145023345947,
"learning_rate": 3.5381414763863166e-05,
"loss": 0.6313,
"step": 127
},
{
"epoch": 1.3763440860215055,
"grad_norm": 2.234398126602173,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.6193,
"step": 128
},
{
"epoch": 1.3870967741935485,
"grad_norm": 2.1972103118896484,
"learning_rate": 3.372159227714218e-05,
"loss": 0.5867,
"step": 129
},
{
"epoch": 1.3978494623655915,
"grad_norm": 3.0726523399353027,
"learning_rate": 3.289899283371657e-05,
"loss": 0.8497,
"step": 130
},
{
"epoch": 1.4086021505376345,
"grad_norm": 2.2349085807800293,
"learning_rate": 3.2081602522734986e-05,
"loss": 0.5208,
"step": 131
},
{
"epoch": 1.4193548387096775,
"grad_norm": 2.5332772731781006,
"learning_rate": 3.12696703292044e-05,
"loss": 0.4452,
"step": 132
},
{
"epoch": 1.4301075268817205,
"grad_norm": 2.002074718475342,
"learning_rate": 3.046344357553632e-05,
"loss": 0.3896,
"step": 133
},
{
"epoch": 1.4408602150537635,
"grad_norm": 2.316570281982422,
"learning_rate": 2.9663167846209998e-05,
"loss": 0.4541,
"step": 134
},
{
"epoch": 1.4516129032258065,
"grad_norm": 2.4266834259033203,
"learning_rate": 2.886908691296504e-05,
"loss": 0.3945,
"step": 135
},
{
"epoch": 1.4623655913978495,
"grad_norm": 2.9699063301086426,
"learning_rate": 2.8081442660546125e-05,
"loss": 0.6487,
"step": 136
},
{
"epoch": 1.4731182795698925,
"grad_norm": 2.3638315200805664,
"learning_rate": 2.7300475013022663e-05,
"loss": 0.2829,
"step": 137
},
{
"epoch": 1.4838709677419355,
"grad_norm": 4.946100234985352,
"learning_rate": 2.6526421860705473e-05,
"loss": 0.4617,
"step": 138
},
{
"epoch": 1.4946236559139785,
"grad_norm": 2.5009000301361084,
"learning_rate": 2.575951898768315e-05,
"loss": 0.2181,
"step": 139
},
{
"epoch": 1.5053763440860215,
"grad_norm": 2.8839311599731445,
"learning_rate": 2.500000000000001e-05,
"loss": 1.7407,
"step": 140
},
{
"epoch": 1.5161290322580645,
"grad_norm": 2.9398391246795654,
"learning_rate": 2.4248096254497288e-05,
"loss": 1.4332,
"step": 141
},
{
"epoch": 1.5268817204301075,
"grad_norm": 2.908121347427368,
"learning_rate": 2.350403678833976e-05,
"loss": 1.2238,
"step": 142
},
{
"epoch": 1.5376344086021505,
"grad_norm": 3.3087892532348633,
"learning_rate": 2.2768048249248648e-05,
"loss": 1.2163,
"step": 143
},
{
"epoch": 1.5483870967741935,
"grad_norm": 2.738379716873169,
"learning_rate": 2.2040354826462668e-05,
"loss": 0.9637,
"step": 144
},
{
"epoch": 1.5591397849462365,
"grad_norm": 3.258429527282715,
"learning_rate": 2.132117818244771e-05,
"loss": 1.1101,
"step": 145
},
{
"epoch": 1.5698924731182795,
"grad_norm": 2.7381410598754883,
"learning_rate": 2.061073738537635e-05,
"loss": 0.9134,
"step": 146
},
{
"epoch": 1.5806451612903225,
"grad_norm": 2.7571890354156494,
"learning_rate": 1.9909248842397584e-05,
"loss": 0.7396,
"step": 147
},
{
"epoch": 1.5913978494623655,
"grad_norm": 2.268134355545044,
"learning_rate": 1.9216926233717085e-05,
"loss": 0.5299,
"step": 148
},
{
"epoch": 1.6021505376344085,
"grad_norm": 2.5643818378448486,
"learning_rate": 1.8533980447508137e-05,
"loss": 0.4746,
"step": 149
},
{
"epoch": 1.6129032258064515,
"grad_norm": 2.6224515438079834,
"learning_rate": 1.7860619515673033e-05,
"loss": 0.5465,
"step": 150
},
{
"epoch": 1.6129032258064515,
"eval_loss": 0.868911862373352,
"eval_runtime": 14.4083,
"eval_samples_per_second": 10.896,
"eval_steps_per_second": 5.483,
"step": 150
},
{
"epoch": 1.6236559139784945,
"grad_norm": 3.0091564655303955,
"learning_rate": 1.7197048550474643e-05,
"loss": 0.7287,
"step": 151
},
{
"epoch": 1.6344086021505375,
"grad_norm": 1.6763337850570679,
"learning_rate": 1.6543469682057106e-05,
"loss": 0.2309,
"step": 152
},
{
"epoch": 1.6451612903225805,
"grad_norm": 2.7991411685943604,
"learning_rate": 1.5900081996875083e-05,
"loss": 0.6103,
"step": 153
},
{
"epoch": 1.6559139784946235,
"grad_norm": 2.1921868324279785,
"learning_rate": 1.526708147705013e-05,
"loss": 0.4928,
"step": 154
},
{
"epoch": 1.6666666666666665,
"grad_norm": 3.5872364044189453,
"learning_rate": 1.4644660940672627e-05,
"loss": 0.6464,
"step": 155
},
{
"epoch": 1.6774193548387095,
"grad_norm": 1.8818094730377197,
"learning_rate": 1.4033009983067452e-05,
"loss": 0.3707,
"step": 156
},
{
"epoch": 1.6881720430107527,
"grad_norm": 2.5614068508148193,
"learning_rate": 1.3432314919041478e-05,
"loss": 0.428,
"step": 157
},
{
"epoch": 1.6989247311827957,
"grad_norm": 2.164555311203003,
"learning_rate": 1.2842758726130283e-05,
"loss": 0.408,
"step": 158
},
{
"epoch": 1.7096774193548387,
"grad_norm": 2.022615432739258,
"learning_rate": 1.22645209888614e-05,
"loss": 0.2232,
"step": 159
},
{
"epoch": 1.7204301075268817,
"grad_norm": 2.447100877761841,
"learning_rate": 1.1697777844051105e-05,
"loss": 0.2983,
"step": 160
},
{
"epoch": 1.7311827956989247,
"grad_norm": 4.865058898925781,
"learning_rate": 1.1142701927151456e-05,
"loss": 0.7972,
"step": 161
},
{
"epoch": 1.7419354838709677,
"grad_norm": 2.682640790939331,
"learning_rate": 1.0599462319663905e-05,
"loss": 0.4589,
"step": 162
},
{
"epoch": 1.7526881720430108,
"grad_norm": 2.2714807987213135,
"learning_rate": 1.006822449763537e-05,
"loss": 1.3321,
"step": 163
},
{
"epoch": 1.7634408602150538,
"grad_norm": 2.663078784942627,
"learning_rate": 9.549150281252633e-06,
"loss": 1.4021,
"step": 164
},
{
"epoch": 1.7741935483870968,
"grad_norm": 2.38264799118042,
"learning_rate": 9.042397785550405e-06,
"loss": 1.2299,
"step": 165
},
{
"epoch": 1.7849462365591398,
"grad_norm": 2.5163092613220215,
"learning_rate": 8.548121372247918e-06,
"loss": 1.0744,
"step": 166
},
{
"epoch": 1.7956989247311828,
"grad_norm": 3.2605020999908447,
"learning_rate": 8.066471602728803e-06,
"loss": 1.1439,
"step": 167
},
{
"epoch": 1.8064516129032258,
"grad_norm": 2.8853485584259033,
"learning_rate": 7.597595192178702e-06,
"loss": 1.0157,
"step": 168
},
{
"epoch": 1.817204301075269,
"grad_norm": 2.8386337757110596,
"learning_rate": 7.1416349648943894e-06,
"loss": 0.9278,
"step": 169
},
{
"epoch": 1.827956989247312,
"grad_norm": 3.006211996078491,
"learning_rate": 6.698729810778065e-06,
"loss": 0.863,
"step": 170
},
{
"epoch": 1.838709677419355,
"grad_norm": 3.339857339859009,
"learning_rate": 6.269014643030213e-06,
"loss": 0.7245,
"step": 171
},
{
"epoch": 1.849462365591398,
"grad_norm": 2.269327402114868,
"learning_rate": 5.852620357053651e-06,
"loss": 0.3663,
"step": 172
},
{
"epoch": 1.860215053763441,
"grad_norm": 2.1220529079437256,
"learning_rate": 5.449673790581611e-06,
"loss": 0.3854,
"step": 173
},
{
"epoch": 1.870967741935484,
"grad_norm": 2.488945722579956,
"learning_rate": 5.060297685041659e-06,
"loss": 0.5545,
"step": 174
},
{
"epoch": 1.881720430107527,
"grad_norm": 2.159728765487671,
"learning_rate": 4.684610648167503e-06,
"loss": 0.4613,
"step": 175
},
{
"epoch": 1.89247311827957,
"grad_norm": 1.9596354961395264,
"learning_rate": 4.322727117869951e-06,
"loss": 0.3092,
"step": 176
},
{
"epoch": 1.903225806451613,
"grad_norm": 2.403904914855957,
"learning_rate": 3.974757327377981e-06,
"loss": 0.4026,
"step": 177
},
{
"epoch": 1.913978494623656,
"grad_norm": 2.790825605392456,
"learning_rate": 3.6408072716606346e-06,
"loss": 0.5259,
"step": 178
},
{
"epoch": 1.924731182795699,
"grad_norm": 1.891068696975708,
"learning_rate": 3.3209786751399187e-06,
"loss": 0.2814,
"step": 179
},
{
"epoch": 1.935483870967742,
"grad_norm": 2.383929967880249,
"learning_rate": 3.0153689607045845e-06,
"loss": 0.4715,
"step": 180
},
{
"epoch": 1.946236559139785,
"grad_norm": 1.4359225034713745,
"learning_rate": 2.724071220034158e-06,
"loss": 0.1781,
"step": 181
},
{
"epoch": 1.956989247311828,
"grad_norm": 2.4820544719696045,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.4342,
"step": 182
},
{
"epoch": 1.967741935483871,
"grad_norm": 1.628670573234558,
"learning_rate": 2.1847622018482283e-06,
"loss": 0.1903,
"step": 183
},
{
"epoch": 1.978494623655914,
"grad_norm": 3.297879934310913,
"learning_rate": 1.9369152030840556e-06,
"loss": 0.3394,
"step": 184
},
{
"epoch": 1.989247311827957,
"grad_norm": 2.3827061653137207,
"learning_rate": 1.70370868554659e-06,
"loss": 0.3608,
"step": 185
},
{
"epoch": 2.0,
"grad_norm": 2.618680238723755,
"learning_rate": 1.4852136862001764e-06,
"loss": 0.8033,
"step": 186
},
{
"epoch": 2.010752688172043,
"grad_norm": 1.7273980379104614,
"learning_rate": 1.2814967607382432e-06,
"loss": 1.237,
"step": 187
},
{
"epoch": 2.021505376344086,
"grad_norm": 2.0505175590515137,
"learning_rate": 1.0926199633097157e-06,
"loss": 1.3056,
"step": 188
},
{
"epoch": 2.032258064516129,
"grad_norm": 2.3544833660125732,
"learning_rate": 9.186408276168013e-07,
"loss": 1.1107,
"step": 189
},
{
"epoch": 2.043010752688172,
"grad_norm": 1.91807222366333,
"learning_rate": 7.596123493895991e-07,
"loss": 0.7513,
"step": 190
},
{
"epoch": 2.053763440860215,
"grad_norm": 2.319427728652954,
"learning_rate": 6.15582970243117e-07,
"loss": 0.8994,
"step": 191
},
{
"epoch": 2.064516129032258,
"grad_norm": 2.4607162475585938,
"learning_rate": 4.865965629214819e-07,
"loss": 0.749,
"step": 192
},
{
"epoch": 2.075268817204301,
"grad_norm": 2.2048518657684326,
"learning_rate": 3.7269241793390085e-07,
"loss": 0.6078,
"step": 193
},
{
"epoch": 2.086021505376344,
"grad_norm": 2.24308443069458,
"learning_rate": 2.7390523158633554e-07,
"loss": 0.6862,
"step": 194
},
{
"epoch": 2.096774193548387,
"grad_norm": 1.80559504032135,
"learning_rate": 1.9026509541272275e-07,
"loss": 0.4276,
"step": 195
},
{
"epoch": 2.10752688172043,
"grad_norm": 1.79929518699646,
"learning_rate": 1.2179748700879012e-07,
"loss": 0.3818,
"step": 196
},
{
"epoch": 2.118279569892473,
"grad_norm": 1.762046456336975,
"learning_rate": 6.852326227130834e-08,
"loss": 0.3659,
"step": 197
},
{
"epoch": 2.129032258064516,
"grad_norm": 1.7550525665283203,
"learning_rate": 3.04586490452119e-08,
"loss": 0.2808,
"step": 198
},
{
"epoch": 2.139784946236559,
"grad_norm": 1.8459057807922363,
"learning_rate": 7.615242180436522e-09,
"loss": 0.325,
"step": 199
},
{
"epoch": 2.150537634408602,
"grad_norm": 1.6853699684143066,
"learning_rate": 0.0,
"loss": 0.2918,
"step": 200
},
{
"epoch": 2.150537634408602,
"eval_loss": 0.8282029628753662,
"eval_runtime": 14.4284,
"eval_samples_per_second": 10.881,
"eval_steps_per_second": 5.475,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.124610424184504e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}