2035 lines
54 KiB
JSON
2035 lines
54 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.05904233335301411,
|
|
"eval_steps": 500,
|
|
"global_step": 1000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.691952991485596,
|
|
"epoch": 0.00029521166676507054,
|
|
"grad_norm": 13.5,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.8063,
|
|
"mean_token_accuracy": 0.00022093121660873293,
|
|
"num_tokens": 9044.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691949939727783,
|
|
"epoch": 0.0005904233335301411,
|
|
"grad_norm": 12.3125,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7746,
|
|
"mean_token_accuracy": 0.00021052630618214607,
|
|
"num_tokens": 20049.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.691005229949951,
|
|
"epoch": 0.0008856350002952116,
|
|
"grad_norm": 10.5,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.5243,
|
|
"mean_token_accuracy": 0.02576384658459574,
|
|
"num_tokens": 30441.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.681586647033692,
|
|
"epoch": 0.0011808466670602822,
|
|
"grad_norm": 6.03125,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.1811,
|
|
"mean_token_accuracy": 0.04208611287176609,
|
|
"num_tokens": 41245.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.640601921081544,
|
|
"epoch": 0.0014760583338253527,
|
|
"grad_norm": 4.0625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.9191,
|
|
"mean_token_accuracy": 0.041578793339431284,
|
|
"num_tokens": 51815.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.59462308883667,
|
|
"epoch": 0.0017712700005904232,
|
|
"grad_norm": 3.203125,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.7236,
|
|
"mean_token_accuracy": 0.046854367852211,
|
|
"num_tokens": 62291.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.566697216033935,
|
|
"epoch": 0.002066481667355494,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.6479,
|
|
"mean_token_accuracy": 0.04647400453686714,
|
|
"num_tokens": 72859.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.56708059310913,
|
|
"epoch": 0.0023616933341205643,
|
|
"grad_norm": 2.609375,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.6398,
|
|
"mean_token_accuracy": 0.04794426709413528,
|
|
"num_tokens": 83349.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.572033309936524,
|
|
"epoch": 0.002656905000885635,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.6101,
|
|
"mean_token_accuracy": 0.04311534278094768,
|
|
"num_tokens": 93550.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.580781936645508,
|
|
"epoch": 0.0029521166676507054,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.5518,
|
|
"mean_token_accuracy": 0.0487774446606636,
|
|
"num_tokens": 104354.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.563411712646484,
|
|
"epoch": 0.003247328334415776,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.4688,
|
|
"mean_token_accuracy": 0.05522352792322636,
|
|
"num_tokens": 114901.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.49093132019043,
|
|
"epoch": 0.0035425400011808465,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.3461,
|
|
"mean_token_accuracy": 0.05991581827402115,
|
|
"num_tokens": 125378.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.3884859085083,
|
|
"epoch": 0.0038377516679459172,
|
|
"grad_norm": 2.28125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.2938,
|
|
"mean_token_accuracy": 0.05960136093199253,
|
|
"num_tokens": 136074.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.436979675292969,
|
|
"epoch": 0.004132963334710988,
|
|
"grad_norm": 2.21875,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.1677,
|
|
"mean_token_accuracy": 0.06536128893494605,
|
|
"num_tokens": 147517.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.378404426574708,
|
|
"epoch": 0.004428175001476058,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.0427,
|
|
"mean_token_accuracy": 0.07047605030238628,
|
|
"num_tokens": 158398.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.407978820800782,
|
|
"epoch": 0.004723386668241129,
|
|
"grad_norm": 2.359375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 8.9631,
|
|
"mean_token_accuracy": 0.07082438990473747,
|
|
"num_tokens": 168721.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.277961730957031,
|
|
"epoch": 0.0050185983350062,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 8.9505,
|
|
"mean_token_accuracy": 0.06753718703985215,
|
|
"num_tokens": 179053.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.263121604919434,
|
|
"epoch": 0.00531381000177127,
|
|
"grad_norm": 2.1875,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.81,
|
|
"mean_token_accuracy": 0.06845465488731861,
|
|
"num_tokens": 189377.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.137414455413818,
|
|
"epoch": 0.0056090216685363405,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.6659,
|
|
"mean_token_accuracy": 0.08077330142259598,
|
|
"num_tokens": 199316.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.143539524078369,
|
|
"epoch": 0.005904233335301411,
|
|
"grad_norm": 2.03125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.5703,
|
|
"mean_token_accuracy": 0.07332179993391037,
|
|
"num_tokens": 209735.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.063137626647949,
|
|
"epoch": 0.006199445002066482,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.5061,
|
|
"mean_token_accuracy": 0.07544220425188541,
|
|
"num_tokens": 220153.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.078357124328614,
|
|
"epoch": 0.006494656668831552,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.4012,
|
|
"mean_token_accuracy": 0.07546676695346832,
|
|
"num_tokens": 231226.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 9.911471176147462,
|
|
"epoch": 0.006789868335596623,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.2365,
|
|
"mean_token_accuracy": 0.07300828918814659,
|
|
"num_tokens": 241715.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.870896339416504,
|
|
"epoch": 0.007085080002361693,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.122,
|
|
"mean_token_accuracy": 0.07934067547321319,
|
|
"num_tokens": 253683.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.708064746856689,
|
|
"epoch": 0.007380291669126764,
|
|
"grad_norm": 1.75,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 7.9982,
|
|
"mean_token_accuracy": 0.07759345918893815,
|
|
"num_tokens": 262863.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.583840370178223,
|
|
"epoch": 0.0076755033358918345,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.9066,
|
|
"mean_token_accuracy": 0.07621643394231796,
|
|
"num_tokens": 273237.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.455964183807373,
|
|
"epoch": 0.007970715002656906,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.8192,
|
|
"mean_token_accuracy": 0.0756739940494299,
|
|
"num_tokens": 283772.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.128956699371338,
|
|
"epoch": 0.008265926669421976,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.686,
|
|
"mean_token_accuracy": 0.08342597484588624,
|
|
"num_tokens": 294776.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 8.970587921142577,
|
|
"epoch": 0.008561138336187046,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.6361,
|
|
"mean_token_accuracy": 0.07761476486921311,
|
|
"num_tokens": 305244.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.766981601715088,
|
|
"epoch": 0.008856350002952117,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.5218,
|
|
"mean_token_accuracy": 0.08236418589949608,
|
|
"num_tokens": 315175.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.568089389801026,
|
|
"epoch": 0.009151561669717187,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.444,
|
|
"mean_token_accuracy": 0.07824666872620582,
|
|
"num_tokens": 326648.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.377489757537841,
|
|
"epoch": 0.009446773336482257,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.4133,
|
|
"mean_token_accuracy": 0.07897001653909683,
|
|
"num_tokens": 336905.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.255996799468994,
|
|
"epoch": 0.009741985003247328,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.3357,
|
|
"mean_token_accuracy": 0.08415433838963508,
|
|
"num_tokens": 348163.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.156648921966553,
|
|
"epoch": 0.0100371966700124,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.3681,
|
|
"mean_token_accuracy": 0.08406007140874863,
|
|
"num_tokens": 359420.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.059180402755738,
|
|
"epoch": 0.01033240833677747,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.2938,
|
|
"mean_token_accuracy": 0.08481760621070862,
|
|
"num_tokens": 369712.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 7.922717761993408,
|
|
"epoch": 0.01062762000354254,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.2317,
|
|
"mean_token_accuracy": 0.08849364891648293,
|
|
"num_tokens": 380413.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 7.878174924850464,
|
|
"epoch": 0.01092283167030761,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.2576,
|
|
"mean_token_accuracy": 0.08616138696670532,
|
|
"num_tokens": 391428.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 7.907813787460327,
|
|
"epoch": 0.011218043337072681,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.1902,
|
|
"mean_token_accuracy": 0.08708868958055974,
|
|
"num_tokens": 402383.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 7.741433143615723,
|
|
"epoch": 0.011513255003837751,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.2219,
|
|
"mean_token_accuracy": 0.08469741642475129,
|
|
"num_tokens": 412692.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.7598718166351315,
|
|
"epoch": 0.011808466670602822,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.1954,
|
|
"mean_token_accuracy": 0.08851397931575775,
|
|
"num_tokens": 423175.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.6599428176879885,
|
|
"epoch": 0.012103678337367892,
|
|
"grad_norm": 1.4609375,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.164,
|
|
"mean_token_accuracy": 0.08638672009110451,
|
|
"num_tokens": 433076.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.721544361114502,
|
|
"epoch": 0.012398890004132964,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.1491,
|
|
"mean_token_accuracy": 0.08772207424044609,
|
|
"num_tokens": 443338.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.719466781616211,
|
|
"epoch": 0.012694101670898034,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.0918,
|
|
"mean_token_accuracy": 0.08484251573681831,
|
|
"num_tokens": 453674.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.642396736145019,
|
|
"epoch": 0.012989313337663105,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.112,
|
|
"mean_token_accuracy": 0.0881990872323513,
|
|
"num_tokens": 464660.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.623027515411377,
|
|
"epoch": 0.013284525004428175,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.135,
|
|
"mean_token_accuracy": 0.08962138071656227,
|
|
"num_tokens": 475478.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.573812961578369,
|
|
"epoch": 0.013579736671193245,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.0507,
|
|
"mean_token_accuracy": 0.0912325844168663,
|
|
"num_tokens": 486279.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.559884262084961,
|
|
"epoch": 0.013874948337958316,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.1265,
|
|
"mean_token_accuracy": 0.08900480046868324,
|
|
"num_tokens": 497947.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.493877267837524,
|
|
"epoch": 0.014170160004723386,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 6.948,
|
|
"mean_token_accuracy": 0.09850702658295632,
|
|
"num_tokens": 508492.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.5359265327453615,
|
|
"epoch": 0.014465371671488458,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.034,
|
|
"mean_token_accuracy": 0.09341262727975845,
|
|
"num_tokens": 519231.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.547327709197998,
|
|
"epoch": 0.014760583338253528,
|
|
"grad_norm": 1.5,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.0388,
|
|
"mean_token_accuracy": 0.08856186792254447,
|
|
"num_tokens": 529936.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.387069749832153,
|
|
"epoch": 0.015055795005018599,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.0339,
|
|
"mean_token_accuracy": 0.09349972456693649,
|
|
"num_tokens": 540315.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.444060134887695,
|
|
"epoch": 0.015351006671783669,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 6.9648,
|
|
"mean_token_accuracy": 0.097850751131773,
|
|
"num_tokens": 551587.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.412183332443237,
|
|
"epoch": 0.01564621833854874,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.000132,
|
|
"loss": 6.9316,
|
|
"mean_token_accuracy": 0.09379192665219308,
|
|
"num_tokens": 561857.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.429265975952148,
|
|
"epoch": 0.01594143000531381,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 6.9717,
|
|
"mean_token_accuracy": 0.09063394144177436,
|
|
"num_tokens": 571784.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.434524583816528,
|
|
"epoch": 0.01623664167207888,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.0056,
|
|
"mean_token_accuracy": 0.09408731907606124,
|
|
"num_tokens": 583475.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.450810766220092,
|
|
"epoch": 0.016531853338843952,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 6.9856,
|
|
"mean_token_accuracy": 0.08944389000535011,
|
|
"num_tokens": 594752.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.2732994556427,
|
|
"epoch": 0.016827065005609022,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 6.9268,
|
|
"mean_token_accuracy": 0.09354007542133332,
|
|
"num_tokens": 604459.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.404242277145386,
|
|
"epoch": 0.017122276672374093,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 6.9605,
|
|
"mean_token_accuracy": 0.09190556928515434,
|
|
"num_tokens": 615477.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.318359518051148,
|
|
"epoch": 0.017417488339139163,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 6.9085,
|
|
"mean_token_accuracy": 0.09038040712475777,
|
|
"num_tokens": 626527.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.3704808235168455,
|
|
"epoch": 0.017712700005904233,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 6.9326,
|
|
"mean_token_accuracy": 0.08688042685389519,
|
|
"num_tokens": 638236.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.292983913421631,
|
|
"epoch": 0.018007911672669304,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000152,
|
|
"loss": 6.9029,
|
|
"mean_token_accuracy": 0.0930778980255127,
|
|
"num_tokens": 648836.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.261309051513672,
|
|
"epoch": 0.018303123339434374,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 6.8546,
|
|
"mean_token_accuracy": 0.09702132865786553,
|
|
"num_tokens": 659610.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.29406886100769,
|
|
"epoch": 0.018598335006199444,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 6.9224,
|
|
"mean_token_accuracy": 0.0894104540348053,
|
|
"num_tokens": 671423.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.343613338470459,
|
|
"epoch": 0.018893546672964515,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 6.8996,
|
|
"mean_token_accuracy": 0.08960375934839249,
|
|
"num_tokens": 682645.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.207725191116333,
|
|
"epoch": 0.019188758339729585,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000162,
|
|
"loss": 6.8131,
|
|
"mean_token_accuracy": 0.09437809884548187,
|
|
"num_tokens": 692315.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.3380321025848385,
|
|
"epoch": 0.019483970006494655,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.8965,
|
|
"mean_token_accuracy": 0.09695785492658615,
|
|
"num_tokens": 702973.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.216346549987793,
|
|
"epoch": 0.019779181673259726,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 6.8629,
|
|
"mean_token_accuracy": 0.0920400932431221,
|
|
"num_tokens": 713787.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.180256462097168,
|
|
"epoch": 0.0200743933400248,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.7804,
|
|
"mean_token_accuracy": 0.09937777146697044,
|
|
"num_tokens": 725357.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.094199705123901,
|
|
"epoch": 0.02036960500678987,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.7488,
|
|
"mean_token_accuracy": 0.10237349718809127,
|
|
"num_tokens": 735003.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.314488554000855,
|
|
"epoch": 0.02066481667355494,
|
|
"grad_norm": 1.2890625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.8222,
|
|
"mean_token_accuracy": 0.09937268868088722,
|
|
"num_tokens": 747278.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.053475475311279,
|
|
"epoch": 0.02096002834032001,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 6.8199,
|
|
"mean_token_accuracy": 0.0992422841489315,
|
|
"num_tokens": 757974.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.2850106716156,
|
|
"epoch": 0.02125524000708508,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 6.8103,
|
|
"mean_token_accuracy": 0.09952985718846322,
|
|
"num_tokens": 767584.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.131100606918335,
|
|
"epoch": 0.02155045167385015,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.7109,
|
|
"mean_token_accuracy": 0.09645930156111718,
|
|
"num_tokens": 777397.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.148231458663941,
|
|
"epoch": 0.02184566334061522,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.7905,
|
|
"mean_token_accuracy": 0.09331258609890938,
|
|
"num_tokens": 788227.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.114551830291748,
|
|
"epoch": 0.02214087500738029,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.7458,
|
|
"mean_token_accuracy": 0.09811887294054031,
|
|
"num_tokens": 799168.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.097161912918091,
|
|
"epoch": 0.022436086674145362,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.7355,
|
|
"mean_token_accuracy": 0.10040534660220146,
|
|
"num_tokens": 810199.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.139192867279053,
|
|
"epoch": 0.022731298340910432,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.7409,
|
|
"mean_token_accuracy": 0.09671718701720237,
|
|
"num_tokens": 822391.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.088096618652344,
|
|
"epoch": 0.023026510007675503,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8565,
|
|
"mean_token_accuracy": 0.09625174552202224,
|
|
"num_tokens": 834350.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.0904669761657715,
|
|
"epoch": 0.023321721674440573,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.7523,
|
|
"mean_token_accuracy": 0.09790571630001069,
|
|
"num_tokens": 845709.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 6.981272554397583,
|
|
"epoch": 0.023616933341205643,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.7085,
|
|
"mean_token_accuracy": 0.10158270671963691,
|
|
"num_tokens": 855092.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.02293906211853,
|
|
"epoch": 0.023912145007970714,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.7193,
|
|
"mean_token_accuracy": 0.08962539583444595,
|
|
"num_tokens": 865943.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.03565149307251,
|
|
"epoch": 0.024207356674735784,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.647,
|
|
"mean_token_accuracy": 0.09619486778974533,
|
|
"num_tokens": 876649.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 6.996511030197143,
|
|
"epoch": 0.024502568341500858,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.6861,
|
|
"mean_token_accuracy": 0.10710784569382667,
|
|
"num_tokens": 886221.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.017127513885498,
|
|
"epoch": 0.024797780008265928,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.7047,
|
|
"mean_token_accuracy": 0.10007574260234833,
|
|
"num_tokens": 896234.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 6.983069896697998,
|
|
"epoch": 0.025092991675031,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.6479,
|
|
"mean_token_accuracy": 0.09976074621081352,
|
|
"num_tokens": 907198.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 6.8837158203125,
|
|
"epoch": 0.02538820334179607,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.702,
|
|
"mean_token_accuracy": 0.10629369989037514,
|
|
"num_tokens": 917579.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 6.978274822235107,
|
|
"epoch": 0.02568341500856114,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.6347,
|
|
"mean_token_accuracy": 0.10065946727991104,
|
|
"num_tokens": 928763.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 6.867664194107055,
|
|
"epoch": 0.02597862667532621,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.6455,
|
|
"mean_token_accuracy": 0.1076437845826149,
|
|
"num_tokens": 939383.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.063400411605835,
|
|
"epoch": 0.02627383834209128,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.7118,
|
|
"mean_token_accuracy": 0.09779355004429817,
|
|
"num_tokens": 949967.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 6.977987289428711,
|
|
"epoch": 0.02656905000885635,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.7177,
|
|
"mean_token_accuracy": 0.09856124073266984,
|
|
"num_tokens": 960438.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 6.9721705436706545,
|
|
"epoch": 0.02686426167562142,
|
|
"grad_norm": 1.2265625,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.678,
|
|
"mean_token_accuracy": 0.1020229034125805,
|
|
"num_tokens": 971771.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.0487203121185305,
|
|
"epoch": 0.02715947334238649,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.7437,
|
|
"mean_token_accuracy": 0.09785925075411797,
|
|
"num_tokens": 982431.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 6.8306797504425045,
|
|
"epoch": 0.02745468500915156,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.69,
|
|
"mean_token_accuracy": 0.09700809046626091,
|
|
"num_tokens": 992763.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 6.909020566940308,
|
|
"epoch": 0.02774989667591663,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.6022,
|
|
"mean_token_accuracy": 0.107904601842165,
|
|
"num_tokens": 1002977.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 6.922635841369629,
|
|
"epoch": 0.0280451083426817,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.6653,
|
|
"mean_token_accuracy": 0.10020340532064438,
|
|
"num_tokens": 1014211.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 6.923598337173462,
|
|
"epoch": 0.028340320009446772,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.6715,
|
|
"mean_token_accuracy": 0.10355789959430695,
|
|
"num_tokens": 1025625.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 6.909690856933594,
|
|
"epoch": 0.028635531676211842,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.686,
|
|
"mean_token_accuracy": 0.10261352136731147,
|
|
"num_tokens": 1036430.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 6.986854982376099,
|
|
"epoch": 0.028930743342976916,
|
|
"grad_norm": 0.9921875,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.7083,
|
|
"mean_token_accuracy": 0.10415773317217827,
|
|
"num_tokens": 1047246.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 6.8165546417236325,
|
|
"epoch": 0.029225955009741986,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.5998,
|
|
"mean_token_accuracy": 0.10099067613482475,
|
|
"num_tokens": 1057004.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 6.876231527328491,
|
|
"epoch": 0.029521166676507057,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.6397,
|
|
"mean_token_accuracy": 0.1102587789297104,
|
|
"num_tokens": 1067224.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 6.866609764099121,
|
|
"epoch": 0.029816378343272127,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000252,
|
|
"loss": 6.6158,
|
|
"mean_token_accuracy": 0.1031361848115921,
|
|
"num_tokens": 1077523.0,
|
|
"step": 505
|
|
},
|
|
{
|
|
"entropy": 6.851018238067627,
|
|
"epoch": 0.030111590010037197,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0002545,
|
|
"loss": 6.5845,
|
|
"mean_token_accuracy": 0.1057098388671875,
|
|
"num_tokens": 1088591.0,
|
|
"step": 510
|
|
},
|
|
{
|
|
"entropy": 6.780634880065918,
|
|
"epoch": 0.030406801676802268,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000257,
|
|
"loss": 6.6068,
|
|
"mean_token_accuracy": 0.10300912559032441,
|
|
"num_tokens": 1100634.0,
|
|
"step": 515
|
|
},
|
|
{
|
|
"entropy": 6.772457027435303,
|
|
"epoch": 0.030702013343567338,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.0002595,
|
|
"loss": 6.5604,
|
|
"mean_token_accuracy": 0.10456541404128075,
|
|
"num_tokens": 1111231.0,
|
|
"step": 520
|
|
},
|
|
{
|
|
"entropy": 6.864635372161866,
|
|
"epoch": 0.03099722501033241,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.000262,
|
|
"loss": 6.5368,
|
|
"mean_token_accuracy": 0.11013480946421624,
|
|
"num_tokens": 1121112.0,
|
|
"step": 525
|
|
},
|
|
{
|
|
"entropy": 6.817177820205688,
|
|
"epoch": 0.03129243667709748,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00026450000000000003,
|
|
"loss": 6.6807,
|
|
"mean_token_accuracy": 0.10330734625458718,
|
|
"num_tokens": 1130803.0,
|
|
"step": 530
|
|
},
|
|
{
|
|
"entropy": 6.852666234970092,
|
|
"epoch": 0.03158764834386255,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.00026700000000000004,
|
|
"loss": 6.6568,
|
|
"mean_token_accuracy": 0.09623496234416962,
|
|
"num_tokens": 1141621.0,
|
|
"step": 535
|
|
},
|
|
{
|
|
"entropy": 6.851540040969849,
|
|
"epoch": 0.03188286001062762,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00026950000000000005,
|
|
"loss": 6.6752,
|
|
"mean_token_accuracy": 0.10156160518527031,
|
|
"num_tokens": 1152860.0,
|
|
"step": 540
|
|
},
|
|
{
|
|
"entropy": 6.910813045501709,
|
|
"epoch": 0.03217807167739269,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.00027200000000000005,
|
|
"loss": 6.5423,
|
|
"mean_token_accuracy": 0.10722377449274063,
|
|
"num_tokens": 1162958.0,
|
|
"step": 545
|
|
},
|
|
{
|
|
"entropy": 6.7858928680419925,
|
|
"epoch": 0.03247328334415776,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.0002745,
|
|
"loss": 6.6395,
|
|
"mean_token_accuracy": 0.10390800461173058,
|
|
"num_tokens": 1174205.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 6.813560676574707,
|
|
"epoch": 0.032768495010922834,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000277,
|
|
"loss": 6.5218,
|
|
"mean_token_accuracy": 0.10657795593142509,
|
|
"num_tokens": 1184336.0,
|
|
"step": 555
|
|
},
|
|
{
|
|
"entropy": 6.759924650192261,
|
|
"epoch": 0.033063706677687904,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.0002795,
|
|
"loss": 6.5285,
|
|
"mean_token_accuracy": 0.11047871261835099,
|
|
"num_tokens": 1194696.0,
|
|
"step": 560
|
|
},
|
|
{
|
|
"entropy": 6.675027894973755,
|
|
"epoch": 0.033358918344452974,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00028199999999999997,
|
|
"loss": 6.4689,
|
|
"mean_token_accuracy": 0.10344288945198059,
|
|
"num_tokens": 1206440.0,
|
|
"step": 565
|
|
},
|
|
{
|
|
"entropy": 6.843683671951294,
|
|
"epoch": 0.033654130011218045,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0002845,
|
|
"loss": 6.545,
|
|
"mean_token_accuracy": 0.10641451999545097,
|
|
"num_tokens": 1216229.0,
|
|
"step": 570
|
|
},
|
|
{
|
|
"entropy": 6.739560270309449,
|
|
"epoch": 0.033949341677983115,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 0.000287,
|
|
"loss": 6.6295,
|
|
"mean_token_accuracy": 0.1049153484404087,
|
|
"num_tokens": 1228077.0,
|
|
"step": 575
|
|
},
|
|
{
|
|
"entropy": 6.755360412597656,
|
|
"epoch": 0.034244553344748185,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0002895,
|
|
"loss": 6.6024,
|
|
"mean_token_accuracy": 0.10700471252202988,
|
|
"num_tokens": 1239047.0,
|
|
"step": 580
|
|
},
|
|
{
|
|
"entropy": 6.835167455673218,
|
|
"epoch": 0.034539765011513256,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000292,
|
|
"loss": 6.5143,
|
|
"mean_token_accuracy": 0.11052370145916939,
|
|
"num_tokens": 1249633.0,
|
|
"step": 585
|
|
},
|
|
{
|
|
"entropy": 6.736393117904663,
|
|
"epoch": 0.034834976678278326,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002945,
|
|
"loss": 6.5858,
|
|
"mean_token_accuracy": 0.10385486260056495,
|
|
"num_tokens": 1261085.0,
|
|
"step": 590
|
|
},
|
|
{
|
|
"entropy": 6.705540657043457,
|
|
"epoch": 0.035130188345043396,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000297,
|
|
"loss": 6.4767,
|
|
"mean_token_accuracy": 0.11535756587982178,
|
|
"num_tokens": 1271161.0,
|
|
"step": 595
|
|
},
|
|
{
|
|
"entropy": 6.711465167999267,
|
|
"epoch": 0.03542540001180847,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0002995,
|
|
"loss": 6.5664,
|
|
"mean_token_accuracy": 0.1055572934448719,
|
|
"num_tokens": 1281776.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 6.741065502166748,
|
|
"epoch": 0.03572061167857354,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000302,
|
|
"loss": 6.5441,
|
|
"mean_token_accuracy": 0.10771351680159569,
|
|
"num_tokens": 1291759.0,
|
|
"step": 605
|
|
},
|
|
{
|
|
"entropy": 6.709814834594726,
|
|
"epoch": 0.03601582334533861,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.0003045,
|
|
"loss": 6.4237,
|
|
"mean_token_accuracy": 0.1066345103085041,
|
|
"num_tokens": 1301783.0,
|
|
"step": 610
|
|
},
|
|
{
|
|
"entropy": 6.701594305038452,
|
|
"epoch": 0.03631103501210368,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000307,
|
|
"loss": 6.5523,
|
|
"mean_token_accuracy": 0.10823088437318802,
|
|
"num_tokens": 1312442.0,
|
|
"step": 615
|
|
},
|
|
{
|
|
"entropy": 6.710081911087036,
|
|
"epoch": 0.03660624667886875,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0003095,
|
|
"loss": 6.5222,
|
|
"mean_token_accuracy": 0.11155530214309692,
|
|
"num_tokens": 1321787.0,
|
|
"step": 620
|
|
},
|
|
{
|
|
"entropy": 6.720113277435303,
|
|
"epoch": 0.03690145834563382,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000312,
|
|
"loss": 6.522,
|
|
"mean_token_accuracy": 0.10557924285531044,
|
|
"num_tokens": 1331724.0,
|
|
"step": 625
|
|
},
|
|
{
|
|
"entropy": 6.668059158325195,
|
|
"epoch": 0.03719667001239889,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0003145,
|
|
"loss": 6.5261,
|
|
"mean_token_accuracy": 0.10685517415404319,
|
|
"num_tokens": 1342272.0,
|
|
"step": 630
|
|
},
|
|
{
|
|
"entropy": 6.737295341491699,
|
|
"epoch": 0.03749188167916396,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000317,
|
|
"loss": 6.5756,
|
|
"mean_token_accuracy": 0.10473631396889686,
|
|
"num_tokens": 1352806.0,
|
|
"step": 635
|
|
},
|
|
{
|
|
"entropy": 6.6583808898925785,
|
|
"epoch": 0.03778709334592903,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0003195,
|
|
"loss": 6.4683,
|
|
"mean_token_accuracy": 0.10982004255056381,
|
|
"num_tokens": 1363668.0,
|
|
"step": 640
|
|
},
|
|
{
|
|
"entropy": 6.669088649749756,
|
|
"epoch": 0.0380823050126941,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.000322,
|
|
"loss": 6.5768,
|
|
"mean_token_accuracy": 0.10401294007897377,
|
|
"num_tokens": 1376374.0,
|
|
"step": 645
|
|
},
|
|
{
|
|
"entropy": 6.727249479293823,
|
|
"epoch": 0.03837751667945917,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00032450000000000003,
|
|
"loss": 6.5341,
|
|
"mean_token_accuracy": 0.10710064172744752,
|
|
"num_tokens": 1387332.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 6.656798219680786,
|
|
"epoch": 0.03867272834622424,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00032700000000000003,
|
|
"loss": 6.4362,
|
|
"mean_token_accuracy": 0.103745087236166,
|
|
"num_tokens": 1398761.0,
|
|
"step": 655
|
|
},
|
|
{
|
|
"entropy": 6.517245817184448,
|
|
"epoch": 0.03896794001298931,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.00032950000000000004,
|
|
"loss": 6.4851,
|
|
"mean_token_accuracy": 0.11338173747062683,
|
|
"num_tokens": 1409818.0,
|
|
"step": 660
|
|
},
|
|
{
|
|
"entropy": 6.730355882644654,
|
|
"epoch": 0.03926315167975438,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00033200000000000005,
|
|
"loss": 6.504,
|
|
"mean_token_accuracy": 0.10808097869157791,
|
|
"num_tokens": 1420839.0,
|
|
"step": 665
|
|
},
|
|
{
|
|
"entropy": 6.628807258605957,
|
|
"epoch": 0.03955836334651945,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00033450000000000005,
|
|
"loss": 6.5066,
|
|
"mean_token_accuracy": 0.10702955350279808,
|
|
"num_tokens": 1434314.0,
|
|
"step": 670
|
|
},
|
|
{
|
|
"entropy": 6.578580474853515,
|
|
"epoch": 0.03985357501328453,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.000337,
|
|
"loss": 6.4204,
|
|
"mean_token_accuracy": 0.11155755072832108,
|
|
"num_tokens": 1445404.0,
|
|
"step": 675
|
|
},
|
|
{
|
|
"entropy": 6.606809616088867,
|
|
"epoch": 0.0401487866800496,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0003395,
|
|
"loss": 6.555,
|
|
"mean_token_accuracy": 0.11048336997628212,
|
|
"num_tokens": 1456860.0,
|
|
"step": 680
|
|
},
|
|
{
|
|
"entropy": 6.589565467834473,
|
|
"epoch": 0.04044399834681467,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000342,
|
|
"loss": 6.4932,
|
|
"mean_token_accuracy": 0.11120514348149299,
|
|
"num_tokens": 1467597.0,
|
|
"step": 685
|
|
},
|
|
{
|
|
"entropy": 6.696408987045288,
|
|
"epoch": 0.04073921001357974,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.00034449999999999997,
|
|
"loss": 6.4581,
|
|
"mean_token_accuracy": 0.11563440561294555,
|
|
"num_tokens": 1478093.0,
|
|
"step": 690
|
|
},
|
|
{
|
|
"entropy": 6.656173992156982,
|
|
"epoch": 0.04103442168034481,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000347,
|
|
"loss": 6.5137,
|
|
"mean_token_accuracy": 0.11482275947928429,
|
|
"num_tokens": 1488527.0,
|
|
"step": 695
|
|
},
|
|
{
|
|
"entropy": 6.550519704818726,
|
|
"epoch": 0.04132963334710988,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.0003495,
|
|
"loss": 6.3829,
|
|
"mean_token_accuracy": 0.11443194523453712,
|
|
"num_tokens": 1499370.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 6.566233539581299,
|
|
"epoch": 0.04162484501387495,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000352,
|
|
"loss": 6.4064,
|
|
"mean_token_accuracy": 0.11202551424503326,
|
|
"num_tokens": 1510518.0,
|
|
"step": 705
|
|
},
|
|
{
|
|
"entropy": 6.535747432708741,
|
|
"epoch": 0.04192005668064002,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0003545,
|
|
"loss": 6.4342,
|
|
"mean_token_accuracy": 0.11363685354590417,
|
|
"num_tokens": 1520791.0,
|
|
"step": 710
|
|
},
|
|
{
|
|
"entropy": 6.632448720932007,
|
|
"epoch": 0.04221526834740509,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.000357,
|
|
"loss": 6.5126,
|
|
"mean_token_accuracy": 0.10954305157065392,
|
|
"num_tokens": 1532122.0,
|
|
"step": 715
|
|
},
|
|
{
|
|
"entropy": 6.5457751750946045,
|
|
"epoch": 0.04251048001417016,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0003595,
|
|
"loss": 6.3958,
|
|
"mean_token_accuracy": 0.11548751443624497,
|
|
"num_tokens": 1542719.0,
|
|
"step": 720
|
|
},
|
|
{
|
|
"entropy": 6.563694810867309,
|
|
"epoch": 0.04280569168093523,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.000362,
|
|
"loss": 6.4559,
|
|
"mean_token_accuracy": 0.11078683212399483,
|
|
"num_tokens": 1553397.0,
|
|
"step": 725
|
|
},
|
|
{
|
|
"entropy": 6.572269773483276,
|
|
"epoch": 0.0431009033477003,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0003645,
|
|
"loss": 6.4017,
|
|
"mean_token_accuracy": 0.11554159820079804,
|
|
"num_tokens": 1563952.0,
|
|
"step": 730
|
|
},
|
|
{
|
|
"entropy": 6.650298690795898,
|
|
"epoch": 0.04339611501446537,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.000367,
|
|
"loss": 6.4341,
|
|
"mean_token_accuracy": 0.11468867510557175,
|
|
"num_tokens": 1574248.0,
|
|
"step": 735
|
|
},
|
|
{
|
|
"entropy": 6.517534494400024,
|
|
"epoch": 0.04369132668123044,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0003695,
|
|
"loss": 6.4001,
|
|
"mean_token_accuracy": 0.11306158527731895,
|
|
"num_tokens": 1584325.0,
|
|
"step": 740
|
|
},
|
|
{
|
|
"entropy": 6.510380697250366,
|
|
"epoch": 0.04398653834799551,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000372,
|
|
"loss": 6.3622,
|
|
"mean_token_accuracy": 0.1204331710934639,
|
|
"num_tokens": 1593737.0,
|
|
"step": 745
|
|
},
|
|
{
|
|
"entropy": 6.5366417407989506,
|
|
"epoch": 0.04428175001476058,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.0003745,
|
|
"loss": 6.42,
|
|
"mean_token_accuracy": 0.11730596050620079,
|
|
"num_tokens": 1604161.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 6.487906742095947,
|
|
"epoch": 0.044576961681525654,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000377,
|
|
"loss": 6.3659,
|
|
"mean_token_accuracy": 0.11893260702490807,
|
|
"num_tokens": 1614719.0,
|
|
"step": 755
|
|
},
|
|
{
|
|
"entropy": 6.4263794898986815,
|
|
"epoch": 0.044872173348290724,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0003795,
|
|
"loss": 6.3783,
|
|
"mean_token_accuracy": 0.11692238673567772,
|
|
"num_tokens": 1626450.0,
|
|
"step": 760
|
|
},
|
|
{
|
|
"entropy": 6.564861059188843,
|
|
"epoch": 0.045167385015055794,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000382,
|
|
"loss": 6.444,
|
|
"mean_token_accuracy": 0.10918903425335884,
|
|
"num_tokens": 1638297.0,
|
|
"step": 765
|
|
},
|
|
{
|
|
"entropy": 6.559746789932251,
|
|
"epoch": 0.045462596681820865,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0003845,
|
|
"loss": 6.4014,
|
|
"mean_token_accuracy": 0.11586136147379875,
|
|
"num_tokens": 1648744.0,
|
|
"step": 770
|
|
},
|
|
{
|
|
"entropy": 6.450838375091553,
|
|
"epoch": 0.045757808348585935,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.00038700000000000003,
|
|
"loss": 6.3877,
|
|
"mean_token_accuracy": 0.11436291560530662,
|
|
"num_tokens": 1660291.0,
|
|
"step": 775
|
|
},
|
|
{
|
|
"entropy": 6.512089347839355,
|
|
"epoch": 0.046053020015351005,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00038950000000000003,
|
|
"loss": 6.3968,
|
|
"mean_token_accuracy": 0.12081583142280579,
|
|
"num_tokens": 1670425.0,
|
|
"step": 780
|
|
},
|
|
{
|
|
"entropy": 6.560306644439697,
|
|
"epoch": 0.046348231682116076,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.00039200000000000004,
|
|
"loss": 6.3979,
|
|
"mean_token_accuracy": 0.11076491773128509,
|
|
"num_tokens": 1681390.0,
|
|
"step": 785
|
|
},
|
|
{
|
|
"entropy": 6.450212621688843,
|
|
"epoch": 0.046643443348881146,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00039450000000000005,
|
|
"loss": 6.4569,
|
|
"mean_token_accuracy": 0.11279196292161942,
|
|
"num_tokens": 1690856.0,
|
|
"step": 790
|
|
},
|
|
{
|
|
"entropy": 6.431796646118164,
|
|
"epoch": 0.046938655015646216,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.00039700000000000005,
|
|
"loss": 6.3975,
|
|
"mean_token_accuracy": 0.11749225705862046,
|
|
"num_tokens": 1701985.0,
|
|
"step": 795
|
|
},
|
|
{
|
|
"entropy": 6.531089401245117,
|
|
"epoch": 0.04723386668241129,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.0003995,
|
|
"loss": 6.33,
|
|
"mean_token_accuracy": 0.1177647665143013,
|
|
"num_tokens": 1713125.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 6.485761880874634,
|
|
"epoch": 0.04752907834917636,
|
|
"grad_norm": 1.0078125,
|
|
"learning_rate": 0.000402,
|
|
"loss": 6.4297,
|
|
"mean_token_accuracy": 0.11685430556535721,
|
|
"num_tokens": 1725409.0,
|
|
"step": 805
|
|
},
|
|
{
|
|
"entropy": 6.4987870216369625,
|
|
"epoch": 0.04782429001594143,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004045,
|
|
"loss": 6.4797,
|
|
"mean_token_accuracy": 0.11249323934316635,
|
|
"num_tokens": 1736579.0,
|
|
"step": 810
|
|
},
|
|
{
|
|
"entropy": 6.412155485153198,
|
|
"epoch": 0.0481195016827065,
|
|
"grad_norm": 1.2109375,
|
|
"learning_rate": 0.00040699999999999997,
|
|
"loss": 6.3084,
|
|
"mean_token_accuracy": 0.11493954733014107,
|
|
"num_tokens": 1747029.0,
|
|
"step": 815
|
|
},
|
|
{
|
|
"entropy": 6.421283435821533,
|
|
"epoch": 0.04841471334947157,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004095,
|
|
"loss": 6.2559,
|
|
"mean_token_accuracy": 0.11367330402135849,
|
|
"num_tokens": 1757444.0,
|
|
"step": 820
|
|
},
|
|
{
|
|
"entropy": 6.466654539108276,
|
|
"epoch": 0.048709925016236645,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.000412,
|
|
"loss": 6.4014,
|
|
"mean_token_accuracy": 0.11653372645378113,
|
|
"num_tokens": 1767576.0,
|
|
"step": 825
|
|
},
|
|
{
|
|
"entropy": 6.4442650318145756,
|
|
"epoch": 0.049005136683001715,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0004145,
|
|
"loss": 6.4257,
|
|
"mean_token_accuracy": 0.11447887197136879,
|
|
"num_tokens": 1777714.0,
|
|
"step": 830
|
|
},
|
|
{
|
|
"entropy": 6.418978452682495,
|
|
"epoch": 0.049300348349766786,
|
|
"grad_norm": 1.09375,
|
|
"learning_rate": 0.000417,
|
|
"loss": 6.4399,
|
|
"mean_token_accuracy": 0.11199806556105614,
|
|
"num_tokens": 1788857.0,
|
|
"step": 835
|
|
},
|
|
{
|
|
"entropy": 6.483746528625488,
|
|
"epoch": 0.049595560016531856,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004195,
|
|
"loss": 6.3862,
|
|
"mean_token_accuracy": 0.11579035371541976,
|
|
"num_tokens": 1800743.0,
|
|
"step": 840
|
|
},
|
|
{
|
|
"entropy": 6.482276964187622,
|
|
"epoch": 0.049890771683296926,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.000422,
|
|
"loss": 6.3287,
|
|
"mean_token_accuracy": 0.11087024435400963,
|
|
"num_tokens": 1810477.0,
|
|
"step": 845
|
|
},
|
|
{
|
|
"entropy": 6.3517107486724855,
|
|
"epoch": 0.050185983350062,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0004245,
|
|
"loss": 6.3575,
|
|
"mean_token_accuracy": 0.11524266377091408,
|
|
"num_tokens": 1821516.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 6.587709379196167,
|
|
"epoch": 0.05048119501682707,
|
|
"grad_norm": 1.0,
|
|
"learning_rate": 0.000427,
|
|
"loss": 6.4653,
|
|
"mean_token_accuracy": 0.11049215123057365,
|
|
"num_tokens": 1832784.0,
|
|
"step": 855
|
|
},
|
|
{
|
|
"entropy": 6.34732494354248,
|
|
"epoch": 0.05077640668359214,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.0004295,
|
|
"loss": 6.2728,
|
|
"mean_token_accuracy": 0.1179530993103981,
|
|
"num_tokens": 1843737.0,
|
|
"step": 860
|
|
},
|
|
{
|
|
"entropy": 6.438384008407593,
|
|
"epoch": 0.05107161835035721,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000432,
|
|
"loss": 6.3462,
|
|
"mean_token_accuracy": 0.12044489830732345,
|
|
"num_tokens": 1854215.0,
|
|
"step": 865
|
|
},
|
|
{
|
|
"entropy": 6.3278967380523685,
|
|
"epoch": 0.05136683001712228,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0004345,
|
|
"loss": 6.3059,
|
|
"mean_token_accuracy": 0.1170185461640358,
|
|
"num_tokens": 1864813.0,
|
|
"step": 870
|
|
},
|
|
{
|
|
"entropy": 6.467825937271118,
|
|
"epoch": 0.05166204168388735,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000437,
|
|
"loss": 6.364,
|
|
"mean_token_accuracy": 0.11428075954318047,
|
|
"num_tokens": 1874962.0,
|
|
"step": 875
|
|
},
|
|
{
|
|
"entropy": 6.507443571090699,
|
|
"epoch": 0.05195725335065242,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.0004395,
|
|
"loss": 6.4321,
|
|
"mean_token_accuracy": 0.1131198450922966,
|
|
"num_tokens": 1886476.0,
|
|
"step": 880
|
|
},
|
|
{
|
|
"entropy": 6.3558813571929935,
|
|
"epoch": 0.05225246501741749,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.000442,
|
|
"loss": 6.3187,
|
|
"mean_token_accuracy": 0.12074669077992439,
|
|
"num_tokens": 1896860.0,
|
|
"step": 885
|
|
},
|
|
{
|
|
"entropy": 6.31374831199646,
|
|
"epoch": 0.05254767668418256,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.0004445,
|
|
"loss": 6.3,
|
|
"mean_token_accuracy": 0.12395407184958458,
|
|
"num_tokens": 1908727.0,
|
|
"step": 890
|
|
},
|
|
{
|
|
"entropy": 6.462383651733399,
|
|
"epoch": 0.05284288835094763,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000447,
|
|
"loss": 6.3968,
|
|
"mean_token_accuracy": 0.11613713800907136,
|
|
"num_tokens": 1920175.0,
|
|
"step": 895
|
|
},
|
|
{
|
|
"entropy": 6.438568925857544,
|
|
"epoch": 0.0531381000177127,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.00044950000000000003,
|
|
"loss": 6.3363,
|
|
"mean_token_accuracy": 0.12365170270204544,
|
|
"num_tokens": 1930267.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 6.340644979476929,
|
|
"epoch": 0.05343331168447777,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 0.00045200000000000004,
|
|
"loss": 6.3596,
|
|
"mean_token_accuracy": 0.12366607934236526,
|
|
"num_tokens": 1940527.0,
|
|
"step": 905
|
|
},
|
|
{
|
|
"entropy": 6.441576862335205,
|
|
"epoch": 0.05372852335124284,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.00045450000000000004,
|
|
"loss": 6.4129,
|
|
"mean_token_accuracy": 0.11123256385326385,
|
|
"num_tokens": 1952557.0,
|
|
"step": 910
|
|
},
|
|
{
|
|
"entropy": 6.354656839370728,
|
|
"epoch": 0.05402373501800791,
|
|
"grad_norm": 1.140625,
|
|
"learning_rate": 0.00045700000000000005,
|
|
"loss": 6.272,
|
|
"mean_token_accuracy": 0.11737576723098755,
|
|
"num_tokens": 1962858.0,
|
|
"step": 915
|
|
},
|
|
{
|
|
"entropy": 6.502988862991333,
|
|
"epoch": 0.05431894668477298,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00045950000000000006,
|
|
"loss": 6.4562,
|
|
"mean_token_accuracy": 0.10999115817248821,
|
|
"num_tokens": 1974579.0,
|
|
"step": 920
|
|
},
|
|
{
|
|
"entropy": 6.265300655364991,
|
|
"epoch": 0.05461415835153805,
|
|
"grad_norm": 1.03125,
|
|
"learning_rate": 0.000462,
|
|
"loss": 6.2954,
|
|
"mean_token_accuracy": 0.129633379727602,
|
|
"num_tokens": 1984937.0,
|
|
"step": 925
|
|
},
|
|
{
|
|
"entropy": 6.484027338027954,
|
|
"epoch": 0.05490937001830312,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0004645,
|
|
"loss": 6.3673,
|
|
"mean_token_accuracy": 0.11584793254733086,
|
|
"num_tokens": 1995046.0,
|
|
"step": 930
|
|
},
|
|
{
|
|
"entropy": 6.3297584533691404,
|
|
"epoch": 0.05520458168506819,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000467,
|
|
"loss": 6.3379,
|
|
"mean_token_accuracy": 0.12105886042118072,
|
|
"num_tokens": 2006261.0,
|
|
"step": 935
|
|
},
|
|
{
|
|
"entropy": 6.4803060531616214,
|
|
"epoch": 0.05549979335183326,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0004695,
|
|
"loss": 6.3509,
|
|
"mean_token_accuracy": 0.11742576733231544,
|
|
"num_tokens": 2017470.0,
|
|
"step": 940
|
|
},
|
|
{
|
|
"entropy": 6.31043176651001,
|
|
"epoch": 0.05579500501859833,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.000472,
|
|
"loss": 6.3327,
|
|
"mean_token_accuracy": 0.12294093891978264,
|
|
"num_tokens": 2028104.0,
|
|
"step": 945
|
|
},
|
|
{
|
|
"entropy": 6.356010007858276,
|
|
"epoch": 0.0560902166853634,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.0004745,
|
|
"loss": 6.3971,
|
|
"mean_token_accuracy": 0.11537264212965966,
|
|
"num_tokens": 2039347.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 6.297523021697998,
|
|
"epoch": 0.056385428352128474,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.000477,
|
|
"loss": 6.2869,
|
|
"mean_token_accuracy": 0.11124283596873283,
|
|
"num_tokens": 2050464.0,
|
|
"step": 955
|
|
},
|
|
{
|
|
"entropy": 6.3561155796051025,
|
|
"epoch": 0.056680640018893544,
|
|
"grad_norm": 1.0703125,
|
|
"learning_rate": 0.0004795,
|
|
"loss": 6.2763,
|
|
"mean_token_accuracy": 0.12250063344836234,
|
|
"num_tokens": 2061554.0,
|
|
"step": 960
|
|
},
|
|
{
|
|
"entropy": 6.351288223266602,
|
|
"epoch": 0.056975851685658614,
|
|
"grad_norm": 1.0859375,
|
|
"learning_rate": 0.000482,
|
|
"loss": 6.3241,
|
|
"mean_token_accuracy": 0.11924508661031723,
|
|
"num_tokens": 2072217.0,
|
|
"step": 965
|
|
},
|
|
{
|
|
"entropy": 6.317080402374268,
|
|
"epoch": 0.057271063352423685,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.0004845,
|
|
"loss": 6.2585,
|
|
"mean_token_accuracy": 0.12346376478672028,
|
|
"num_tokens": 2082088.0,
|
|
"step": 970
|
|
},
|
|
{
|
|
"entropy": 6.334501934051514,
|
|
"epoch": 0.05756627501918876,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.000487,
|
|
"loss": 6.333,
|
|
"mean_token_accuracy": 0.11622191295027733,
|
|
"num_tokens": 2091959.0,
|
|
"step": 975
|
|
},
|
|
{
|
|
"entropy": 6.329196119308472,
|
|
"epoch": 0.05786148668595383,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 0.0004895,
|
|
"loss": 6.2826,
|
|
"mean_token_accuracy": 0.12484293133020401,
|
|
"num_tokens": 2102907.0,
|
|
"step": 980
|
|
},
|
|
{
|
|
"entropy": 6.299886989593506,
|
|
"epoch": 0.0581566983527189,
|
|
"grad_norm": 1.046875,
|
|
"learning_rate": 0.000492,
|
|
"loss": 6.3273,
|
|
"mean_token_accuracy": 0.1229906715452671,
|
|
"num_tokens": 2112818.0,
|
|
"step": 985
|
|
},
|
|
{
|
|
"entropy": 6.340411424636841,
|
|
"epoch": 0.05845191001948397,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0004945,
|
|
"loss": 6.2827,
|
|
"mean_token_accuracy": 0.12152156233787537,
|
|
"num_tokens": 2122847.0,
|
|
"step": 990
|
|
},
|
|
{
|
|
"entropy": 6.255873918533325,
|
|
"epoch": 0.05874712168624904,
|
|
"grad_norm": 1.015625,
|
|
"learning_rate": 0.000497,
|
|
"loss": 6.2182,
|
|
"mean_token_accuracy": 0.12108674049377441,
|
|
"num_tokens": 2133317.0,
|
|
"step": 995
|
|
},
|
|
{
|
|
"entropy": 6.304410934448242,
|
|
"epoch": 0.05904233335301411,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.0004995,
|
|
"loss": 6.3561,
|
|
"mean_token_accuracy": 0.11512539386749268,
|
|
"num_tokens": 2144568.0,
|
|
"step": 1000
|
|
}
|
|
],
|
|
"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": 3114674509824000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|