1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.03269469692015955,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.691962432861327,
|
|
"epoch": 0.0003269469692015955,
|
|
"grad_norm": 12.5625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8118,
|
|
"mean_token_accuracy": 0.0002268856973387301,
|
|
"num_tokens": 10275.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691912364959716,
|
|
"epoch": 0.000653893938403191,
|
|
"grad_norm": 12.0,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7548,
|
|
"mean_token_accuracy": 0.0008593271079007537,
|
|
"num_tokens": 20293.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.690659618377685,
|
|
"epoch": 0.0009808409076047864,
|
|
"grad_norm": 9.375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5796,
|
|
"mean_token_accuracy": 0.027759611373767257,
|
|
"num_tokens": 30769.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.671999549865722,
|
|
"epoch": 0.001307787876806382,
|
|
"grad_norm": 5.5625,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.333,
|
|
"mean_token_accuracy": 0.04241157788783312,
|
|
"num_tokens": 41351.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.554566478729248,
|
|
"epoch": 0.0016347348460079775,
|
|
"grad_norm": 4.625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.128,
|
|
"mean_token_accuracy": 0.04083434790372849,
|
|
"num_tokens": 51137.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.48172836303711,
|
|
"epoch": 0.001961681815209573,
|
|
"grad_norm": 2.84375,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.0286,
|
|
"mean_token_accuracy": 0.04095044266432524,
|
|
"num_tokens": 61184.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.52005319595337,
|
|
"epoch": 0.0022886287844111684,
|
|
"grad_norm": 2.890625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.9428,
|
|
"mean_token_accuracy": 0.04355649538338184,
|
|
"num_tokens": 71499.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.493383026123047,
|
|
"epoch": 0.002615575753612764,
|
|
"grad_norm": 2.390625,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.9404,
|
|
"mean_token_accuracy": 0.042642039619386195,
|
|
"num_tokens": 81761.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.462896347045898,
|
|
"epoch": 0.0029425227228143595,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.9134,
|
|
"mean_token_accuracy": 0.044699825905263425,
|
|
"num_tokens": 93149.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.480470275878906,
|
|
"epoch": 0.003269469692015955,
|
|
"grad_norm": 2.40625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.8578,
|
|
"mean_token_accuracy": 0.04424864072352648,
|
|
"num_tokens": 103201.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.50320463180542,
|
|
"epoch": 0.0035964166612175506,
|
|
"grad_norm": 2.21875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.8204,
|
|
"mean_token_accuracy": 0.05240613967180252,
|
|
"num_tokens": 114340.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.402948760986328,
|
|
"epoch": 0.003923363630419146,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.7316,
|
|
"mean_token_accuracy": 0.05212643444538116,
|
|
"num_tokens": 125783.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.340407657623292,
|
|
"epoch": 0.004250310599620741,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.6209,
|
|
"mean_token_accuracy": 0.05632065013051033,
|
|
"num_tokens": 135351.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.26957654953003,
|
|
"epoch": 0.004577257568822337,
|
|
"grad_norm": 1.625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.6186,
|
|
"mean_token_accuracy": 0.056864064931869504,
|
|
"num_tokens": 146942.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.386287784576416,
|
|
"epoch": 0.004904204538023932,
|
|
"grad_norm": 1.7109375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.5645,
|
|
"mean_token_accuracy": 0.056648611277341845,
|
|
"num_tokens": 157676.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.315990447998047,
|
|
"epoch": 0.005231151507225528,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.4324,
|
|
"mean_token_accuracy": 0.056685902178287506,
|
|
"num_tokens": 168608.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.201674747467042,
|
|
"epoch": 0.0055580984764271235,
|
|
"grad_norm": 1.7890625,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.3533,
|
|
"mean_token_accuracy": 0.05901941917836666,
|
|
"num_tokens": 179268.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.28765230178833,
|
|
"epoch": 0.005885045445628719,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.3399,
|
|
"mean_token_accuracy": 0.05402175635099411,
|
|
"num_tokens": 189279.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.194643878936768,
|
|
"epoch": 0.006211992414830315,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.196,
|
|
"mean_token_accuracy": 0.06014312915503979,
|
|
"num_tokens": 198791.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.213055515289307,
|
|
"epoch": 0.00653893938403191,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 9.1506,
|
|
"mean_token_accuracy": 0.05811347998678684,
|
|
"num_tokens": 208262.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.124658584594727,
|
|
"epoch": 0.006865886353233506,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 9.0631,
|
|
"mean_token_accuracy": 0.05890461802482605,
|
|
"num_tokens": 219359.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.020441341400147,
|
|
"epoch": 0.007192833322435101,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.9009,
|
|
"mean_token_accuracy": 0.061437368765473364,
|
|
"num_tokens": 229582.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.00902338027954,
|
|
"epoch": 0.007519780291636697,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.9367,
|
|
"mean_token_accuracy": 0.057738688215613364,
|
|
"num_tokens": 239856.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.907579135894775,
|
|
"epoch": 0.007846727260838291,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.7953,
|
|
"mean_token_accuracy": 0.05762846954166889,
|
|
"num_tokens": 249620.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.752235412597656,
|
|
"epoch": 0.008173674230039887,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.7378,
|
|
"mean_token_accuracy": 0.05578424893319607,
|
|
"num_tokens": 260081.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.781640815734864,
|
|
"epoch": 0.008500621199241483,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.5809,
|
|
"mean_token_accuracy": 0.06519303396344185,
|
|
"num_tokens": 269889.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.525392723083495,
|
|
"epoch": 0.008827568168443078,
|
|
"grad_norm": 1.5625,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.6158,
|
|
"mean_token_accuracy": 0.06163802072405815,
|
|
"num_tokens": 279884.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.505074214935302,
|
|
"epoch": 0.009154515137644674,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.5592,
|
|
"mean_token_accuracy": 0.05903047285974026,
|
|
"num_tokens": 290532.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.370058345794678,
|
|
"epoch": 0.00948146210684627,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.4303,
|
|
"mean_token_accuracy": 0.058588751778006556,
|
|
"num_tokens": 300384.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.222196769714355,
|
|
"epoch": 0.009808409076047865,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 8.3183,
|
|
"mean_token_accuracy": 0.06703893281519413,
|
|
"num_tokens": 310486.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.067441463470459,
|
|
"epoch": 0.01013535604524946,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 8.369,
|
|
"mean_token_accuracy": 0.06330439783632755,
|
|
"num_tokens": 320405.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.02968339920044,
|
|
"epoch": 0.010462303014451056,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 8.3285,
|
|
"mean_token_accuracy": 0.06537889093160629,
|
|
"num_tokens": 331041.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.945148849487305,
|
|
"epoch": 0.010789249983652651,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 8.3496,
|
|
"mean_token_accuracy": 0.06229307875037193,
|
|
"num_tokens": 341014.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.904346561431884,
|
|
"epoch": 0.011116196952854247,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 8.299,
|
|
"mean_token_accuracy": 0.06237541511654854,
|
|
"num_tokens": 352840.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.780626392364502,
|
|
"epoch": 0.011443143922055843,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 8.2693,
|
|
"mean_token_accuracy": 0.06663689725100994,
|
|
"num_tokens": 362829.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.713014221191406,
|
|
"epoch": 0.011770090891257438,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 8.1825,
|
|
"mean_token_accuracy": 0.06633488647639751,
|
|
"num_tokens": 373330.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.689865589141846,
|
|
"epoch": 0.012097037860459034,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 8.1874,
|
|
"mean_token_accuracy": 0.06622433178126812,
|
|
"num_tokens": 383488.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.68079071044922,
|
|
"epoch": 0.01242398482966063,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 8.1841,
|
|
"mean_token_accuracy": 0.06749865673482418,
|
|
"num_tokens": 393936.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.624857902526855,
|
|
"epoch": 0.012750931798862225,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 8.2894,
|
|
"mean_token_accuracy": 0.06525838598608971,
|
|
"num_tokens": 404013.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.582938766479492,
|
|
"epoch": 0.01307787876806382,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 8.1825,
|
|
"mean_token_accuracy": 0.06989333964884281,
|
|
"num_tokens": 414009.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.596421909332275,
|
|
"epoch": 0.013404825737265416,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 8.2087,
|
|
"mean_token_accuracy": 0.06921766847372054,
|
|
"num_tokens": 424011.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.601021480560302,
|
|
"epoch": 0.013731772706467011,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 8.1777,
|
|
"mean_token_accuracy": 0.06699340268969536,
|
|
"num_tokens": 434985.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.557740116119385,
|
|
"epoch": 0.014058719675668607,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 8.1858,
|
|
"mean_token_accuracy": 0.06825171709060669,
|
|
"num_tokens": 446276.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.430389785766602,
|
|
"epoch": 0.014385666644870202,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 8.0834,
|
|
"mean_token_accuracy": 0.07472717836499214,
|
|
"num_tokens": 455307.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 8.491849899291992,
|
|
"epoch": 0.014712613614071798,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 8.1734,
|
|
"mean_token_accuracy": 0.07013615183532237,
|
|
"num_tokens": 465253.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 8.517255306243896,
|
|
"epoch": 0.015039560583273394,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 8.1761,
|
|
"mean_token_accuracy": 0.06887588053941726,
|
|
"num_tokens": 475588.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 8.475957870483398,
|
|
"epoch": 0.015366507552474989,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 8.0487,
|
|
"mean_token_accuracy": 0.07346096634864807,
|
|
"num_tokens": 485356.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 8.470558166503906,
|
|
"epoch": 0.015693454521676583,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 8.0824,
|
|
"mean_token_accuracy": 0.0730382464826107,
|
|
"num_tokens": 494798.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 8.372940826416016,
|
|
"epoch": 0.01602040149087818,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000122,
|
|
"loss": 8.1423,
|
|
"mean_token_accuracy": 0.07283038049936294,
|
|
"num_tokens": 504907.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 8.429611682891846,
|
|
"epoch": 0.016347348460079774,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 8.153,
|
|
"mean_token_accuracy": 0.07119653932750225,
|
|
"num_tokens": 515219.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 8.42165298461914,
|
|
"epoch": 0.01667429542928137,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000127,
|
|
"loss": 8.0908,
|
|
"mean_token_accuracy": 0.07306713648140431,
|
|
"num_tokens": 524467.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 8.492340278625488,
|
|
"epoch": 0.017001242398482965,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 8.1286,
|
|
"mean_token_accuracy": 0.07200958207249641,
|
|
"num_tokens": 533573.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 8.316695594787598,
|
|
"epoch": 0.01732818936768456,
|
|
"grad_norm": 1.90625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 8.1454,
|
|
"mean_token_accuracy": 0.07354601062834262,
|
|
"num_tokens": 543373.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 8.469814777374268,
|
|
"epoch": 0.017655136336886156,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 8.0899,
|
|
"mean_token_accuracy": 0.0713756311684847,
|
|
"num_tokens": 554398.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 8.352188110351562,
|
|
"epoch": 0.017982083306087752,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 8.063,
|
|
"mean_token_accuracy": 0.0770417533814907,
|
|
"num_tokens": 564362.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 8.408161735534668,
|
|
"epoch": 0.018309030275289347,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 8.1399,
|
|
"mean_token_accuracy": 0.07096323706209659,
|
|
"num_tokens": 575658.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 8.40747356414795,
|
|
"epoch": 0.018635977244490943,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 8.1079,
|
|
"mean_token_accuracy": 0.07521827667951583,
|
|
"num_tokens": 585744.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 8.309930896759033,
|
|
"epoch": 0.01896292421369254,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.983,
|
|
"mean_token_accuracy": 0.07047860436141491,
|
|
"num_tokens": 595435.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 8.389182662963867,
|
|
"epoch": 0.019289871182894134,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 8.0106,
|
|
"mean_token_accuracy": 0.0744885340332985,
|
|
"num_tokens": 605374.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 8.33745756149292,
|
|
"epoch": 0.01961681815209573,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 8.1112,
|
|
"mean_token_accuracy": 0.07103150337934494,
|
|
"num_tokens": 616418.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 8.340550994873047,
|
|
"epoch": 0.019943765121297325,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.9849,
|
|
"mean_token_accuracy": 0.08127295225858688,
|
|
"num_tokens": 625857.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 8.38352565765381,
|
|
"epoch": 0.02027071209049892,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 8.0612,
|
|
"mean_token_accuracy": 0.07608035206794739,
|
|
"num_tokens": 636469.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 8.316315841674804,
|
|
"epoch": 0.020597659059700516,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000157,
|
|
"loss": 8.0281,
|
|
"mean_token_accuracy": 0.0735818523913622,
|
|
"num_tokens": 648105.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 8.296995544433594,
|
|
"epoch": 0.02092460602890211,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.9422,
|
|
"mean_token_accuracy": 0.0793572224676609,
|
|
"num_tokens": 657214.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 8.264441871643067,
|
|
"epoch": 0.021251552998103707,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000162,
|
|
"loss": 8.0265,
|
|
"mean_token_accuracy": 0.073749577999115,
|
|
"num_tokens": 667228.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 8.35250301361084,
|
|
"epoch": 0.021578499967305303,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.9781,
|
|
"mean_token_accuracy": 0.07626760490238667,
|
|
"num_tokens": 677412.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 8.250011539459228,
|
|
"epoch": 0.0219054469365069,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.9378,
|
|
"mean_token_accuracy": 0.07601722851395606,
|
|
"num_tokens": 687893.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 8.302884387969971,
|
|
"epoch": 0.022232393905708494,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 8.0661,
|
|
"mean_token_accuracy": 0.07206159345805645,
|
|
"num_tokens": 698497.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 8.278049850463868,
|
|
"epoch": 0.02255934087491009,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.9492,
|
|
"mean_token_accuracy": 0.07396932058036328,
|
|
"num_tokens": 708593.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 8.230413436889648,
|
|
"epoch": 0.022886287844111685,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.9611,
|
|
"mean_token_accuracy": 0.07906975001096725,
|
|
"num_tokens": 719781.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 8.322347927093507,
|
|
"epoch": 0.02321323481331328,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.9722,
|
|
"mean_token_accuracy": 0.07742859311401844,
|
|
"num_tokens": 729946.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 8.199689579010009,
|
|
"epoch": 0.023540181782514876,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.9721,
|
|
"mean_token_accuracy": 0.07622929476201534,
|
|
"num_tokens": 740542.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 8.228047943115234,
|
|
"epoch": 0.02386712875171647,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.8555,
|
|
"mean_token_accuracy": 0.08382938280701638,
|
|
"num_tokens": 750387.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 8.169698715209961,
|
|
"epoch": 0.024194075720918067,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.9124,
|
|
"mean_token_accuracy": 0.07616014927625656,
|
|
"num_tokens": 759708.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 8.173001480102538,
|
|
"epoch": 0.024521022690119663,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.861,
|
|
"mean_token_accuracy": 0.08310946524143219,
|
|
"num_tokens": 770655.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 8.247628688812256,
|
|
"epoch": 0.02484796965932126,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.9962,
|
|
"mean_token_accuracy": 0.07753782495856285,
|
|
"num_tokens": 780263.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 8.120156478881835,
|
|
"epoch": 0.025174916628522854,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.8174,
|
|
"mean_token_accuracy": 0.08559585362672806,
|
|
"num_tokens": 789667.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 8.231314945220948,
|
|
"epoch": 0.02550186359772445,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.8745,
|
|
"mean_token_accuracy": 0.07828151807188988,
|
|
"num_tokens": 800099.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 8.123421001434327,
|
|
"epoch": 0.025828810566926045,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.8562,
|
|
"mean_token_accuracy": 0.08755742833018303,
|
|
"num_tokens": 809347.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 8.190375709533692,
|
|
"epoch": 0.02615575753612764,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 7.8895,
|
|
"mean_token_accuracy": 0.0774234376847744,
|
|
"num_tokens": 819721.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 8.068334150314332,
|
|
"epoch": 0.026482704505329236,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000202,
|
|
"loss": 7.8147,
|
|
"mean_token_accuracy": 0.08247106224298477,
|
|
"num_tokens": 830278.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 8.158495712280274,
|
|
"epoch": 0.02680965147453083,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.8692,
|
|
"mean_token_accuracy": 0.07925740331411361,
|
|
"num_tokens": 839760.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 8.091846561431884,
|
|
"epoch": 0.027136598443732427,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.8936,
|
|
"mean_token_accuracy": 0.08189868628978729,
|
|
"num_tokens": 850778.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 8.150493049621582,
|
|
"epoch": 0.027463545412934023,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.8127,
|
|
"mean_token_accuracy": 0.08337080627679824,
|
|
"num_tokens": 860419.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 8.17796688079834,
|
|
"epoch": 0.027790492382135618,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.9115,
|
|
"mean_token_accuracy": 0.08478068634867668,
|
|
"num_tokens": 869939.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 8.120305299758911,
|
|
"epoch": 0.028117439351337214,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.8569,
|
|
"mean_token_accuracy": 0.08082921504974365,
|
|
"num_tokens": 880331.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 8.22661657333374,
|
|
"epoch": 0.02844438632053881,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.9214,
|
|
"mean_token_accuracy": 0.07378799393773079,
|
|
"num_tokens": 891342.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 8.110522651672364,
|
|
"epoch": 0.028771333289740405,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.8951,
|
|
"mean_token_accuracy": 0.07804202027618885,
|
|
"num_tokens": 902253.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 8.053891658782959,
|
|
"epoch": 0.029098280258942,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.8047,
|
|
"mean_token_accuracy": 0.08209186047315598,
|
|
"num_tokens": 913086.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 8.072042274475098,
|
|
"epoch": 0.029425227228143596,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.8269,
|
|
"mean_token_accuracy": 0.0826948344707489,
|
|
"num_tokens": 925082.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 8.100944375991821,
|
|
"epoch": 0.02975217419734519,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 7.822,
|
|
"mean_token_accuracy": 0.07624491900205613,
|
|
"num_tokens": 933955.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 8.06613826751709,
|
|
"epoch": 0.030079121166546787,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 7.7756,
|
|
"mean_token_accuracy": 0.08377034105360508,
|
|
"num_tokens": 943951.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 8.096276569366456,
|
|
"epoch": 0.030406068135748383,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 7.7703,
|
|
"mean_token_accuracy": 0.08263073042035103,
|
|
"num_tokens": 955297.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.890885972976685,
|
|
"epoch": 0.030733015104949978,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 7.8169,
|
|
"mean_token_accuracy": 0.08191272169351578,
|
|
"num_tokens": 967012.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 8.025649976730346,
|
|
"epoch": 0.031059962074151574,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000237,
|
|
"loss": 7.7327,
|
|
"mean_token_accuracy": 0.08218838796019554,
|
|
"num_tokens": 977956.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 8.034643745422363,
|
|
"epoch": 0.031386909043353166,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.7506,
|
|
"mean_token_accuracy": 0.08425655737519264,
|
|
"num_tokens": 987270.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.962606287002563,
|
|
"epoch": 0.03171385601255476,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.7517,
|
|
"mean_token_accuracy": 0.08468098640441894,
|
|
"num_tokens": 998010.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.962759017944336,
|
|
"epoch": 0.03204080298175636,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 7.7869,
|
|
"mean_token_accuracy": 0.09083530306816101,
|
|
"num_tokens": 1008173.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 8.062295961380006,
|
|
"epoch": 0.03236774995095795,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 7.7972,
|
|
"mean_token_accuracy": 0.08286057338118553,
|
|
"num_tokens": 1019037.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 8.059563636779785,
|
|
"epoch": 0.03269469692015955,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 7.7878,
|
|
"mean_token_accuracy": 0.08349623754620553,
|
|
"num_tokens": 1028822.0,
|
|
"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": 1643322746880000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|