1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.033488496701383076,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.692002010345458,
|
|
"epoch": 0.00033488496701383075,
|
|
"grad_norm": 10.5625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.798,
|
|
"mean_token_accuracy": 0.00014925373252481223,
|
|
"num_tokens": 10113.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.69187297821045,
|
|
"epoch": 0.0006697699340276615,
|
|
"grad_norm": 11.375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.749,
|
|
"mean_token_accuracy": 0.000525744108017534,
|
|
"num_tokens": 19598.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.690606594085693,
|
|
"epoch": 0.0010046549010414923,
|
|
"grad_norm": 8.5,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5809,
|
|
"mean_token_accuracy": 0.0301290619187057,
|
|
"num_tokens": 29705.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.670234203338623,
|
|
"epoch": 0.001339539868055323,
|
|
"grad_norm": 8.75,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.3539,
|
|
"mean_token_accuracy": 0.04047959893941879,
|
|
"num_tokens": 39871.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.557568454742432,
|
|
"epoch": 0.0016744248350691537,
|
|
"grad_norm": 4.3125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.1453,
|
|
"mean_token_accuracy": 0.04189771674573421,
|
|
"num_tokens": 49805.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.465806770324708,
|
|
"epoch": 0.0020093098020829846,
|
|
"grad_norm": 2.90625,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.9507,
|
|
"mean_token_accuracy": 0.04296911805868149,
|
|
"num_tokens": 59845.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.461338233947753,
|
|
"epoch": 0.0023441947690968153,
|
|
"grad_norm": 2.859375,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.9614,
|
|
"mean_token_accuracy": 0.0368592556566,
|
|
"num_tokens": 69836.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.483565521240234,
|
|
"epoch": 0.002679079736110646,
|
|
"grad_norm": 2.40625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.8929,
|
|
"mean_token_accuracy": 0.046063635125756265,
|
|
"num_tokens": 80088.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.410326480865479,
|
|
"epoch": 0.0030139647031244767,
|
|
"grad_norm": 2.359375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.8768,
|
|
"mean_token_accuracy": 0.04229466691613197,
|
|
"num_tokens": 89458.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.434312725067139,
|
|
"epoch": 0.0033488496701383074,
|
|
"grad_norm": 2.25,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.8494,
|
|
"mean_token_accuracy": 0.04813771955668926,
|
|
"num_tokens": 99523.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.504166984558106,
|
|
"epoch": 0.003683734637152138,
|
|
"grad_norm": 2.0625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.8242,
|
|
"mean_token_accuracy": 0.05405057594180107,
|
|
"num_tokens": 110080.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.344751262664795,
|
|
"epoch": 0.004018619604165969,
|
|
"grad_norm": 2.046875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.7129,
|
|
"mean_token_accuracy": 0.05208342745900154,
|
|
"num_tokens": 119412.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.211891746520996,
|
|
"epoch": 0.0043535045711798,
|
|
"grad_norm": 1.9453125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.6776,
|
|
"mean_token_accuracy": 0.051375290378928185,
|
|
"num_tokens": 129432.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.302197170257568,
|
|
"epoch": 0.0046883895381936306,
|
|
"grad_norm": 1.8828125,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.6112,
|
|
"mean_token_accuracy": 0.059913215786218645,
|
|
"num_tokens": 139756.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.31943874359131,
|
|
"epoch": 0.005023274505207461,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.5806,
|
|
"mean_token_accuracy": 0.05665694735944271,
|
|
"num_tokens": 148846.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.25365629196167,
|
|
"epoch": 0.005358159472221292,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.5087,
|
|
"mean_token_accuracy": 0.06129935123026371,
|
|
"num_tokens": 158647.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.23855152130127,
|
|
"epoch": 0.005693044439235123,
|
|
"grad_norm": 1.9921875,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.3883,
|
|
"mean_token_accuracy": 0.060005851462483405,
|
|
"num_tokens": 167701.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.186964225769042,
|
|
"epoch": 0.006027929406248953,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.3761,
|
|
"mean_token_accuracy": 0.06175476983189583,
|
|
"num_tokens": 178428.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.192708683013915,
|
|
"epoch": 0.006362814373262784,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.3126,
|
|
"mean_token_accuracy": 0.057852984964847566,
|
|
"num_tokens": 187737.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.148875999450684,
|
|
"epoch": 0.006697699340276615,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.3047,
|
|
"mean_token_accuracy": 0.05756702236831188,
|
|
"num_tokens": 198374.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.143085479736328,
|
|
"epoch": 0.007032584307290445,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.2348,
|
|
"mean_token_accuracy": 0.05544475130736828,
|
|
"num_tokens": 208483.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.065252590179444,
|
|
"epoch": 0.007367469274304276,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 9.1179,
|
|
"mean_token_accuracy": 0.0532416757196188,
|
|
"num_tokens": 218715.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 9.994503688812255,
|
|
"epoch": 0.007702354241318108,
|
|
"grad_norm": 1.7265625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9596,
|
|
"mean_token_accuracy": 0.06599014401435851,
|
|
"num_tokens": 227777.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.866777515411377,
|
|
"epoch": 0.008037239208331938,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.982,
|
|
"mean_token_accuracy": 0.05981399100273847,
|
|
"num_tokens": 238192.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.84700288772583,
|
|
"epoch": 0.008372124175345769,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.827,
|
|
"mean_token_accuracy": 0.06517648994922638,
|
|
"num_tokens": 247259.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.726642894744874,
|
|
"epoch": 0.0087070091423596,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.8551,
|
|
"mean_token_accuracy": 0.054852675274014474,
|
|
"num_tokens": 258100.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.696339225769043,
|
|
"epoch": 0.00904189410937343,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.7497,
|
|
"mean_token_accuracy": 0.061625052243471146,
|
|
"num_tokens": 267620.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.46251974105835,
|
|
"epoch": 0.009376779076387261,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.6451,
|
|
"mean_token_accuracy": 0.07104777507483959,
|
|
"num_tokens": 277654.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.472876358032227,
|
|
"epoch": 0.009711664043401092,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.6622,
|
|
"mean_token_accuracy": 0.061386901512742044,
|
|
"num_tokens": 288195.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.343764114379884,
|
|
"epoch": 0.010046549010414923,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.6348,
|
|
"mean_token_accuracy": 0.0647747952491045,
|
|
"num_tokens": 298860.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.256326389312743,
|
|
"epoch": 0.010381433977428753,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.5569,
|
|
"mean_token_accuracy": 0.06752786003053188,
|
|
"num_tokens": 307799.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.192852115631103,
|
|
"epoch": 0.010716318944442584,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.5493,
|
|
"mean_token_accuracy": 0.06289746835827828,
|
|
"num_tokens": 318257.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.057830619812012,
|
|
"epoch": 0.011051203911456415,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.5351,
|
|
"mean_token_accuracy": 0.06436791345477104,
|
|
"num_tokens": 329065.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.072613048553468,
|
|
"epoch": 0.011386088878470245,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.4825,
|
|
"mean_token_accuracy": 0.06475077606737614,
|
|
"num_tokens": 338809.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.959860801696777,
|
|
"epoch": 0.011720973845484076,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.4588,
|
|
"mean_token_accuracy": 0.06541862487792968,
|
|
"num_tokens": 349318.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.849260616302491,
|
|
"epoch": 0.012055858812497907,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.4348,
|
|
"mean_token_accuracy": 0.06766009479761123,
|
|
"num_tokens": 358760.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.860241985321045,
|
|
"epoch": 0.012390743779511737,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.3168,
|
|
"mean_token_accuracy": 0.06955580413341522,
|
|
"num_tokens": 367589.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.835146236419678,
|
|
"epoch": 0.012725628746525568,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.3624,
|
|
"mean_token_accuracy": 0.06997261717915534,
|
|
"num_tokens": 377012.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.665368556976318,
|
|
"epoch": 0.013060513713539399,
|
|
"grad_norm": 0.96875,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.3276,
|
|
"mean_token_accuracy": 0.07220411114394665,
|
|
"num_tokens": 386791.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.820245933532714,
|
|
"epoch": 0.01339539868055323,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.4404,
|
|
"mean_token_accuracy": 0.06381001807749272,
|
|
"num_tokens": 396620.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.68319730758667,
|
|
"epoch": 0.01373028364756706,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.3971,
|
|
"mean_token_accuracy": 0.06604428850114345,
|
|
"num_tokens": 406384.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.685719966888428,
|
|
"epoch": 0.01406516861458089,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.3622,
|
|
"mean_token_accuracy": 0.07016287520527839,
|
|
"num_tokens": 416481.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.711257934570312,
|
|
"epoch": 0.014400053581594721,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.3841,
|
|
"mean_token_accuracy": 0.07033630795776843,
|
|
"num_tokens": 425697.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.652897453308105,
|
|
"epoch": 0.014734938548608552,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.2995,
|
|
"mean_token_accuracy": 0.0704294566065073,
|
|
"num_tokens": 435836.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.68178768157959,
|
|
"epoch": 0.015069823515622383,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.3569,
|
|
"mean_token_accuracy": 0.06939754635095596,
|
|
"num_tokens": 445926.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.638516330718994,
|
|
"epoch": 0.015404708482636215,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.2983,
|
|
"mean_token_accuracy": 0.07235002666711807,
|
|
"num_tokens": 455594.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.570248413085938,
|
|
"epoch": 0.015739593449650046,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.3461,
|
|
"mean_token_accuracy": 0.0698947947472334,
|
|
"num_tokens": 465928.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.62375259399414,
|
|
"epoch": 0.016074478416663877,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.2851,
|
|
"mean_token_accuracy": 0.07294886186718941,
|
|
"num_tokens": 475946.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.62733097076416,
|
|
"epoch": 0.016409363383677707,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.2878,
|
|
"mean_token_accuracy": 0.07264770828187465,
|
|
"num_tokens": 485285.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.576191997528076,
|
|
"epoch": 0.016744248350691538,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.2942,
|
|
"mean_token_accuracy": 0.07046032212674617,
|
|
"num_tokens": 496516.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.584659576416016,
|
|
"epoch": 0.01707913331770537,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.2829,
|
|
"mean_token_accuracy": 0.07218111529946328,
|
|
"num_tokens": 506330.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.551284503936767,
|
|
"epoch": 0.0174140182847192,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.3443,
|
|
"mean_token_accuracy": 0.07309619709849358,
|
|
"num_tokens": 515825.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.580291271209717,
|
|
"epoch": 0.01774890325173303,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.2687,
|
|
"mean_token_accuracy": 0.06971881128847598,
|
|
"num_tokens": 526303.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.48112440109253,
|
|
"epoch": 0.01808378821874686,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.3759,
|
|
"mean_token_accuracy": 0.07506775110960007,
|
|
"num_tokens": 536805.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.600781631469726,
|
|
"epoch": 0.01841867318576069,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.2216,
|
|
"mean_token_accuracy": 0.07659760937094688,
|
|
"num_tokens": 546381.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.565075874328613,
|
|
"epoch": 0.018753558152774522,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.2444,
|
|
"mean_token_accuracy": 0.0764175396412611,
|
|
"num_tokens": 556278.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.52633352279663,
|
|
"epoch": 0.019088443119788353,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.2569,
|
|
"mean_token_accuracy": 0.07796998918056489,
|
|
"num_tokens": 565312.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.471183109283448,
|
|
"epoch": 0.019423328086802184,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 8.2313,
|
|
"mean_token_accuracy": 0.07882754877209663,
|
|
"num_tokens": 575304.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.441308689117431,
|
|
"epoch": 0.019758213053816014,
|
|
"grad_norm": 0.99609375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.1941,
|
|
"mean_token_accuracy": 0.07772204428911209,
|
|
"num_tokens": 585262.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.575240802764892,
|
|
"epoch": 0.020093098020829845,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.2345,
|
|
"mean_token_accuracy": 0.07687565684318542,
|
|
"num_tokens": 595297.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.531679725646972,
|
|
"epoch": 0.020427982987843676,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000152,
|
|
"loss": 8.2113,
|
|
"mean_token_accuracy": 0.07619660049676895,
|
|
"num_tokens": 604354.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.408655452728272,
|
|
"epoch": 0.020762867954857506,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.3286,
|
|
"mean_token_accuracy": 0.07043087370693683,
|
|
"num_tokens": 616283.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.580455112457276,
|
|
"epoch": 0.021097752921871337,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.2259,
|
|
"mean_token_accuracy": 0.07452421076595783,
|
|
"num_tokens": 626059.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.43509645462036,
|
|
"epoch": 0.021432637888885168,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 8.1612,
|
|
"mean_token_accuracy": 0.07599973231554032,
|
|
"num_tokens": 634483.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.409455680847168,
|
|
"epoch": 0.021767522855899,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.1607,
|
|
"mean_token_accuracy": 0.08002224043011666,
|
|
"num_tokens": 643111.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.471851444244384,
|
|
"epoch": 0.02210240782291283,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 8.1549,
|
|
"mean_token_accuracy": 0.08036141917109489,
|
|
"num_tokens": 653650.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.432329463958741,
|
|
"epoch": 0.02243729278992666,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 8.1509,
|
|
"mean_token_accuracy": 0.08708992898464203,
|
|
"num_tokens": 663143.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.438332271575927,
|
|
"epoch": 0.02277217775694049,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.1221,
|
|
"mean_token_accuracy": 0.08310596123337746,
|
|
"num_tokens": 672434.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.37921953201294,
|
|
"epoch": 0.02310706272395432,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 8.1224,
|
|
"mean_token_accuracy": 0.07925173193216324,
|
|
"num_tokens": 682973.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.429475784301758,
|
|
"epoch": 0.023441947690968152,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 8.1876,
|
|
"mean_token_accuracy": 0.07883969098329544,
|
|
"num_tokens": 692136.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.41495771408081,
|
|
"epoch": 0.023776832657981983,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 8.0919,
|
|
"mean_token_accuracy": 0.08360464051365853,
|
|
"num_tokens": 702257.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.438455581665039,
|
|
"epoch": 0.024111717624995813,
|
|
"grad_norm": 0.96484375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 8.1034,
|
|
"mean_token_accuracy": 0.08223021924495696,
|
|
"num_tokens": 712444.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.402928638458253,
|
|
"epoch": 0.024446602592009644,
|
|
"grad_norm": 2.59375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 8.2461,
|
|
"mean_token_accuracy": 0.07820370681583881,
|
|
"num_tokens": 721614.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.494019985198975,
|
|
"epoch": 0.024781487559023475,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 8.2242,
|
|
"mean_token_accuracy": 0.07433447390794753,
|
|
"num_tokens": 731189.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.423804473876952,
|
|
"epoch": 0.025116372526037305,
|
|
"grad_norm": 0.9609375,
|
|
"learning_rate": 0.000187,
|
|
"loss": 8.1387,
|
|
"mean_token_accuracy": 0.08087190091609955,
|
|
"num_tokens": 741555.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.242571067810058,
|
|
"epoch": 0.025451257493051136,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 8.0677,
|
|
"mean_token_accuracy": 0.084798514097929,
|
|
"num_tokens": 751560.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.45106019973755,
|
|
"epoch": 0.025786142460064967,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 8.1193,
|
|
"mean_token_accuracy": 0.07877458333969116,
|
|
"num_tokens": 761471.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.435086345672607,
|
|
"epoch": 0.026121027427078797,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 8.1184,
|
|
"mean_token_accuracy": 0.08121965527534485,
|
|
"num_tokens": 771565.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.258657455444336,
|
|
"epoch": 0.026455912394092628,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 8.0685,
|
|
"mean_token_accuracy": 0.08140704035758972,
|
|
"num_tokens": 781248.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.397636890411377,
|
|
"epoch": 0.02679079736110646,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 8.1233,
|
|
"mean_token_accuracy": 0.0783721823245287,
|
|
"num_tokens": 790907.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.232399845123291,
|
|
"epoch": 0.02712568232812029,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 8.1384,
|
|
"mean_token_accuracy": 0.07922295071184635,
|
|
"num_tokens": 800073.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.517311382293702,
|
|
"epoch": 0.02746056729513412,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 8.099,
|
|
"mean_token_accuracy": 0.0789324052631855,
|
|
"num_tokens": 809782.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.192907619476319,
|
|
"epoch": 0.02779545226214795,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 8.0579,
|
|
"mean_token_accuracy": 0.08063366934657097,
|
|
"num_tokens": 820329.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.31554536819458,
|
|
"epoch": 0.02813033722916178,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.9757,
|
|
"mean_token_accuracy": 0.09008945301175117,
|
|
"num_tokens": 829405.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.286150169372558,
|
|
"epoch": 0.028465222196175612,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.000212,
|
|
"loss": 8.0418,
|
|
"mean_token_accuracy": 0.08079335764050484,
|
|
"num_tokens": 838700.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.263776969909667,
|
|
"epoch": 0.028800107163189443,
|
|
"grad_norm": 1.0390625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 8.1043,
|
|
"mean_token_accuracy": 0.083824248239398,
|
|
"num_tokens": 848892.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.295380115509033,
|
|
"epoch": 0.029134992130203274,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 8.0354,
|
|
"mean_token_accuracy": 0.08374309539794922,
|
|
"num_tokens": 857737.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.229405975341797,
|
|
"epoch": 0.029469877097217104,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.968,
|
|
"mean_token_accuracy": 0.08430874347686768,
|
|
"num_tokens": 867758.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.303557300567627,
|
|
"epoch": 0.029804762064230935,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.9853,
|
|
"mean_token_accuracy": 0.08112291395664215,
|
|
"num_tokens": 877273.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.286294078826904,
|
|
"epoch": 0.030139647031244766,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 8.087,
|
|
"mean_token_accuracy": 0.08337052017450333,
|
|
"num_tokens": 887109.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.264797973632813,
|
|
"epoch": 0.030474531998258596,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 8.0057,
|
|
"mean_token_accuracy": 0.08394731357693672,
|
|
"num_tokens": 896157.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.19506025314331,
|
|
"epoch": 0.03080941696527243,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 8.011,
|
|
"mean_token_accuracy": 0.07882361188530922,
|
|
"num_tokens": 907231.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.185263442993165,
|
|
"epoch": 0.03114430193228626,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.9223,
|
|
"mean_token_accuracy": 0.0873487912118435,
|
|
"num_tokens": 916756.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 8.243685722351074,
|
|
"epoch": 0.03147918689930009,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 8.0922,
|
|
"mean_token_accuracy": 0.07646922953426838,
|
|
"num_tokens": 927714.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.29259443283081,
|
|
"epoch": 0.03181407186631392,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.9389,
|
|
"mean_token_accuracy": 0.08775902166962624,
|
|
"num_tokens": 936620.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.152761650085449,
|
|
"epoch": 0.03214895683332775,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.9753,
|
|
"mean_token_accuracy": 0.08129271119832993,
|
|
"num_tokens": 945942.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 8.177737617492676,
|
|
"epoch": 0.032483841800341584,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 8.0066,
|
|
"mean_token_accuracy": 0.08826668187975883,
|
|
"num_tokens": 956685.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 8.196662044525146,
|
|
"epoch": 0.032818726767355415,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.9442,
|
|
"mean_token_accuracy": 0.08767017722129822,
|
|
"num_tokens": 967453.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.19489345550537,
|
|
"epoch": 0.033153611734369245,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.9764,
|
|
"mean_token_accuracy": 0.08438041880726814,
|
|
"num_tokens": 978220.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.135752773284912,
|
|
"epoch": 0.033488496701383076,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.9185,
|
|
"mean_token_accuracy": 0.09252278879284859,
|
|
"num_tokens": 988120.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"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": 1575019376640000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|