1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.011172310545543924,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.74203462600708,
|
|
"epoch": 0.00011172310545543924,
|
|
"grad_norm": 7.40625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.545,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 4264.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.74201259613037,
|
|
"epoch": 0.00022344621091087847,
|
|
"grad_norm": 6.84375,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.5105,
|
|
"mean_token_accuracy": 0.0002816901309415698,
|
|
"num_tokens": 8413.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.742000961303711,
|
|
"epoch": 0.0003351693163663177,
|
|
"grad_norm": 6.78125,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.4826,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 12394.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.742033767700196,
|
|
"epoch": 0.00044689242182175694,
|
|
"grad_norm": 5.90625,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.4451,
|
|
"mean_token_accuracy": 0.0004269361263141036,
|
|
"num_tokens": 16610.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.742008590698243,
|
|
"epoch": 0.0005586155272771962,
|
|
"grad_norm": 5.9375,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 10.3813,
|
|
"mean_token_accuracy": 0.0017635513795539737,
|
|
"num_tokens": 21243.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.74174461364746,
|
|
"epoch": 0.0006703386327326354,
|
|
"grad_norm": 5.1875,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 10.2219,
|
|
"mean_token_accuracy": 0.019105370296165346,
|
|
"num_tokens": 25571.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.741024494171143,
|
|
"epoch": 0.0007820617381880747,
|
|
"grad_norm": 4.75,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 10.0233,
|
|
"mean_token_accuracy": 0.0476666621863842,
|
|
"num_tokens": 29412.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.739530563354492,
|
|
"epoch": 0.0008937848436435139,
|
|
"grad_norm": 3.75,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.943,
|
|
"mean_token_accuracy": 0.04776074178516865,
|
|
"num_tokens": 33893.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.736891174316407,
|
|
"epoch": 0.0010055079490989532,
|
|
"grad_norm": 3.09375,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.7721,
|
|
"mean_token_accuracy": 0.05200008656829595,
|
|
"num_tokens": 38043.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.733825874328613,
|
|
"epoch": 0.0011172310545543925,
|
|
"grad_norm": 2.75,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.7095,
|
|
"mean_token_accuracy": 0.05134744010865688,
|
|
"num_tokens": 42222.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.731979370117188,
|
|
"epoch": 0.0012289541600098315,
|
|
"grad_norm": 2.765625,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.6035,
|
|
"mean_token_accuracy": 0.05346306636929512,
|
|
"num_tokens": 46321.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.729666996002198,
|
|
"epoch": 0.0013406772654652708,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.5985,
|
|
"mean_token_accuracy": 0.053918526880443095,
|
|
"num_tokens": 50816.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.726247787475586,
|
|
"epoch": 0.0014524003709207101,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.5526,
|
|
"mean_token_accuracy": 0.05777559094130993,
|
|
"num_tokens": 55331.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.722254943847656,
|
|
"epoch": 0.0015641234763761494,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.4754,
|
|
"mean_token_accuracy": 0.060246657207608224,
|
|
"num_tokens": 59379.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.717647266387939,
|
|
"epoch": 0.0016758465818315885,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.3716,
|
|
"mean_token_accuracy": 0.05563458241522312,
|
|
"num_tokens": 63463.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.709985065460206,
|
|
"epoch": 0.0017875696872870278,
|
|
"grad_norm": 2.234375,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.4054,
|
|
"mean_token_accuracy": 0.04782464448362589,
|
|
"num_tokens": 67897.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.698591709136963,
|
|
"epoch": 0.001899292792742467,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 9.2081,
|
|
"mean_token_accuracy": 0.05416969805955887,
|
|
"num_tokens": 71995.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.679736518859864,
|
|
"epoch": 0.0020110158981979064,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 9.0655,
|
|
"mean_token_accuracy": 0.06419909372925758,
|
|
"num_tokens": 75745.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.639767360687255,
|
|
"epoch": 0.0021227390036533456,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 9.0156,
|
|
"mean_token_accuracy": 0.06227423809468746,
|
|
"num_tokens": 79641.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.604651832580567,
|
|
"epoch": 0.002234462109108785,
|
|
"grad_norm": 2.328125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.9827,
|
|
"mean_token_accuracy": 0.07007253468036652,
|
|
"num_tokens": 84368.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.576740074157716,
|
|
"epoch": 0.0023461852145642242,
|
|
"grad_norm": 3.0,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.9633,
|
|
"mean_token_accuracy": 0.06716309636831283,
|
|
"num_tokens": 89177.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.532594108581543,
|
|
"epoch": 0.002457908320019663,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.7625,
|
|
"mean_token_accuracy": 0.06927448399364948,
|
|
"num_tokens": 93266.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.469588470458984,
|
|
"epoch": 0.0025696314254751024,
|
|
"grad_norm": 1.9921875,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.6738,
|
|
"mean_token_accuracy": 0.0684173908084631,
|
|
"num_tokens": 97385.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 10.414745044708251,
|
|
"epoch": 0.0026813545309305417,
|
|
"grad_norm": 2.203125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.4918,
|
|
"mean_token_accuracy": 0.06782151162624359,
|
|
"num_tokens": 100911.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 10.343620300292969,
|
|
"epoch": 0.002793077636385981,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.4104,
|
|
"mean_token_accuracy": 0.07619427219033241,
|
|
"num_tokens": 104785.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 10.282330799102784,
|
|
"epoch": 0.0029048007418414202,
|
|
"grad_norm": 2.09375,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 8.3421,
|
|
"mean_token_accuracy": 0.0715565849095583,
|
|
"num_tokens": 109434.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 10.177853965759278,
|
|
"epoch": 0.0030165238472968595,
|
|
"grad_norm": 1.7578125,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 8.3045,
|
|
"mean_token_accuracy": 0.07821577228605747,
|
|
"num_tokens": 113869.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 10.01875,
|
|
"epoch": 0.003128246952752299,
|
|
"grad_norm": 1.640625,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 8.2069,
|
|
"mean_token_accuracy": 0.07112646363675594,
|
|
"num_tokens": 118332.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.93814001083374,
|
|
"epoch": 0.003239970058207738,
|
|
"grad_norm": 1.65625,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 8.0893,
|
|
"mean_token_accuracy": 0.06839369647204877,
|
|
"num_tokens": 122140.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 9.799712753295898,
|
|
"epoch": 0.003351693163663177,
|
|
"grad_norm": 1.6875,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.9249,
|
|
"mean_token_accuracy": 0.067073168233037,
|
|
"num_tokens": 126515.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 9.603694915771484,
|
|
"epoch": 0.0034634162691186163,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.897,
|
|
"mean_token_accuracy": 0.07342451699078083,
|
|
"num_tokens": 130397.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 9.402643203735352,
|
|
"epoch": 0.0035751393745740555,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.815,
|
|
"mean_token_accuracy": 0.07550931498408317,
|
|
"num_tokens": 134428.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 9.27815113067627,
|
|
"epoch": 0.003686862480029495,
|
|
"grad_norm": 1.5546875,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.6921,
|
|
"mean_token_accuracy": 0.08235657699406147,
|
|
"num_tokens": 138294.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 9.126440715789794,
|
|
"epoch": 0.003798585585484934,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.6466,
|
|
"mean_token_accuracy": 0.0780983179807663,
|
|
"num_tokens": 142750.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.904253101348877,
|
|
"epoch": 0.003910308690940373,
|
|
"grad_norm": 1.1328125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.5833,
|
|
"mean_token_accuracy": 0.06832590401172638,
|
|
"num_tokens": 147220.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.704329586029052,
|
|
"epoch": 0.004022031796395813,
|
|
"grad_norm": 1.0546875,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.464,
|
|
"mean_token_accuracy": 0.08249393478035927,
|
|
"num_tokens": 151396.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.544238567352295,
|
|
"epoch": 0.0041337549018512516,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.4503,
|
|
"mean_token_accuracy": 0.07549951747059822,
|
|
"num_tokens": 155336.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.501252269744873,
|
|
"epoch": 0.004245478007306691,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.4842,
|
|
"mean_token_accuracy": 0.07688803523778916,
|
|
"num_tokens": 159752.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.432751655578613,
|
|
"epoch": 0.00435720111276213,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.5426,
|
|
"mean_token_accuracy": 0.07533748783171176,
|
|
"num_tokens": 163895.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 8.257897663116456,
|
|
"epoch": 0.00446892421821757,
|
|
"grad_norm": 1.0234375,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.3421,
|
|
"mean_token_accuracy": 0.08197390548884868,
|
|
"num_tokens": 168312.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 8.259009075164794,
|
|
"epoch": 0.004580647323673009,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.3701,
|
|
"mean_token_accuracy": 0.08555959761142731,
|
|
"num_tokens": 172358.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 8.100054454803466,
|
|
"epoch": 0.0046923704291284484,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.357,
|
|
"mean_token_accuracy": 0.08119111023843288,
|
|
"num_tokens": 177085.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 8.163977813720702,
|
|
"epoch": 0.004804093534583887,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.4691,
|
|
"mean_token_accuracy": 0.07897469773888588,
|
|
"num_tokens": 181410.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 8.064214897155761,
|
|
"epoch": 0.004915816640039326,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.2085,
|
|
"mean_token_accuracy": 0.08347514495253563,
|
|
"num_tokens": 185649.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.9723944664001465,
|
|
"epoch": 0.005027539745494766,
|
|
"grad_norm": 1.0625,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.2571,
|
|
"mean_token_accuracy": 0.08833960294723511,
|
|
"num_tokens": 190055.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.958705711364746,
|
|
"epoch": 0.005139262850950205,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2784,
|
|
"mean_token_accuracy": 0.08142814487218857,
|
|
"num_tokens": 194267.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.980498743057251,
|
|
"epoch": 0.0052509859564056445,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.2042,
|
|
"mean_token_accuracy": 0.08119908273220063,
|
|
"num_tokens": 198483.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.903439283370972,
|
|
"epoch": 0.005362709061861083,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.1137,
|
|
"mean_token_accuracy": 0.09260561317205429,
|
|
"num_tokens": 202514.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.820244312286377,
|
|
"epoch": 0.005474432167316523,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2591,
|
|
"mean_token_accuracy": 0.08432262316346169,
|
|
"num_tokens": 206748.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.817278146743774,
|
|
"epoch": 0.005586155272771962,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.1337,
|
|
"mean_token_accuracy": 0.08795059770345688,
|
|
"num_tokens": 210981.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.902036476135254,
|
|
"epoch": 0.005697878378227402,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.2718,
|
|
"mean_token_accuracy": 0.08792312145233154,
|
|
"num_tokens": 215551.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.720875978469849,
|
|
"epoch": 0.0058096014836828405,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.0933,
|
|
"mean_token_accuracy": 0.09327159300446511,
|
|
"num_tokens": 219677.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.8553773880004885,
|
|
"epoch": 0.005921324589138279,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.1915,
|
|
"mean_token_accuracy": 0.08988388553261757,
|
|
"num_tokens": 223457.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.8154174327850345,
|
|
"epoch": 0.006033047694593719,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.026,
|
|
"mean_token_accuracy": 0.10610299035906792,
|
|
"num_tokens": 227135.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.805374622344971,
|
|
"epoch": 0.006144770800049158,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.2242,
|
|
"mean_token_accuracy": 0.0845766805112362,
|
|
"num_tokens": 231393.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.793313646316529,
|
|
"epoch": 0.006256493905504598,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.2039,
|
|
"mean_token_accuracy": 0.09036076590418815,
|
|
"num_tokens": 235959.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.797818899154663,
|
|
"epoch": 0.0063682170109600365,
|
|
"grad_norm": 1.265625,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.0391,
|
|
"mean_token_accuracy": 0.10008666068315505,
|
|
"num_tokens": 239875.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.643310832977295,
|
|
"epoch": 0.006479940116415476,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.1816,
|
|
"mean_token_accuracy": 0.10088859647512435,
|
|
"num_tokens": 244397.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.749200963973999,
|
|
"epoch": 0.006591663221870915,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.1022,
|
|
"mean_token_accuracy": 0.09432346150279045,
|
|
"num_tokens": 248441.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.724449968338012,
|
|
"epoch": 0.006703386327326354,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.1223,
|
|
"mean_token_accuracy": 0.09665866494178772,
|
|
"num_tokens": 252693.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.719059801101684,
|
|
"epoch": 0.006815109432781794,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.0852,
|
|
"mean_token_accuracy": 0.0973274439573288,
|
|
"num_tokens": 257135.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.689913558959961,
|
|
"epoch": 0.0069268325382372325,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.1788,
|
|
"mean_token_accuracy": 0.09386017993092537,
|
|
"num_tokens": 261919.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.723158597946167,
|
|
"epoch": 0.007038555643692672,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.0899,
|
|
"mean_token_accuracy": 0.10055064782500267,
|
|
"num_tokens": 266656.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.58135347366333,
|
|
"epoch": 0.007150278749148111,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.0484,
|
|
"mean_token_accuracy": 0.0989685259759426,
|
|
"num_tokens": 270879.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.633032703399659,
|
|
"epoch": 0.007262001854603551,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.0017,
|
|
"mean_token_accuracy": 0.11045164242386818,
|
|
"num_tokens": 274884.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.55345048904419,
|
|
"epoch": 0.00737372496005899,
|
|
"grad_norm": 1.328125,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 6.9434,
|
|
"mean_token_accuracy": 0.10104935392737388,
|
|
"num_tokens": 279096.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.626837635040284,
|
|
"epoch": 0.007485448065514429,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.0338,
|
|
"mean_token_accuracy": 0.10009614899754524,
|
|
"num_tokens": 282923.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.528380966186523,
|
|
"epoch": 0.007597171170969868,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 6.99,
|
|
"mean_token_accuracy": 0.10532780215144158,
|
|
"num_tokens": 287283.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.572710466384888,
|
|
"epoch": 0.007708894276425307,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 6.9504,
|
|
"mean_token_accuracy": 0.10530123338103295,
|
|
"num_tokens": 291409.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.547561979293823,
|
|
"epoch": 0.007820617381880746,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 6.9732,
|
|
"mean_token_accuracy": 0.10650015398859977,
|
|
"num_tokens": 295157.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.669625616073608,
|
|
"epoch": 0.007932340487336187,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0005,
|
|
"mean_token_accuracy": 0.10243007764220238,
|
|
"num_tokens": 299370.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.543743228912353,
|
|
"epoch": 0.008044063592791625,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.0093,
|
|
"mean_token_accuracy": 0.11244696080684662,
|
|
"num_tokens": 303955.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.570756626129151,
|
|
"epoch": 0.008155786698247064,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.000182,
|
|
"loss": 6.9899,
|
|
"mean_token_accuracy": 0.10562622845172882,
|
|
"num_tokens": 308047.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.594445466995239,
|
|
"epoch": 0.008267509803702503,
|
|
"grad_norm": 1.609375,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 6.9313,
|
|
"mean_token_accuracy": 0.11213268861174583,
|
|
"num_tokens": 312050.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.5787379264831545,
|
|
"epoch": 0.008379232909157944,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 0.000187,
|
|
"loss": 6.9007,
|
|
"mean_token_accuracy": 0.11451447457075119,
|
|
"num_tokens": 315901.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.617064571380615,
|
|
"epoch": 0.008490956014613383,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 6.9233,
|
|
"mean_token_accuracy": 0.11718286573886871,
|
|
"num_tokens": 320077.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.479556846618652,
|
|
"epoch": 0.008602679120068821,
|
|
"grad_norm": 1.546875,
|
|
"learning_rate": 0.000192,
|
|
"loss": 6.8783,
|
|
"mean_token_accuracy": 0.11937321722507477,
|
|
"num_tokens": 324365.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.485933542251587,
|
|
"epoch": 0.00871440222552426,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 6.8292,
|
|
"mean_token_accuracy": 0.11517802253365517,
|
|
"num_tokens": 327971.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.489345073699951,
|
|
"epoch": 0.0088261253309797,
|
|
"grad_norm": 1.5390625,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 6.8634,
|
|
"mean_token_accuracy": 0.12188869342207909,
|
|
"num_tokens": 331791.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.405192232131958,
|
|
"epoch": 0.00893784843643514,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.9151,
|
|
"mean_token_accuracy": 0.11866991743445396,
|
|
"num_tokens": 335616.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.442100572586059,
|
|
"epoch": 0.009049571541890579,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8337,
|
|
"mean_token_accuracy": 0.1216049998998642,
|
|
"num_tokens": 339815.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.466716241836548,
|
|
"epoch": 0.009161294647346017,
|
|
"grad_norm": 1.4140625,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 6.8153,
|
|
"mean_token_accuracy": 0.11885289028286934,
|
|
"num_tokens": 344045.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.439525604248047,
|
|
"epoch": 0.009273017752801456,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 0.000207,
|
|
"loss": 6.9127,
|
|
"mean_token_accuracy": 0.1201499916613102,
|
|
"num_tokens": 348726.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.421639108657837,
|
|
"epoch": 0.009384740858256897,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 6.8973,
|
|
"mean_token_accuracy": 0.11855003610253334,
|
|
"num_tokens": 353228.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.533071041107178,
|
|
"epoch": 0.009496463963712336,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 0.000212,
|
|
"loss": 6.8732,
|
|
"mean_token_accuracy": 0.12777991741895675,
|
|
"num_tokens": 357379.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.404365539550781,
|
|
"epoch": 0.009608187069167775,
|
|
"grad_norm": 1.78125,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 6.8237,
|
|
"mean_token_accuracy": 0.11873817816376686,
|
|
"num_tokens": 361347.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.3829724311828615,
|
|
"epoch": 0.009719910174623213,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 6.8001,
|
|
"mean_token_accuracy": 0.12097886875271797,
|
|
"num_tokens": 365364.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.4662988662719725,
|
|
"epoch": 0.009831633280078652,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 6.8435,
|
|
"mean_token_accuracy": 0.1184915505349636,
|
|
"num_tokens": 369648.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.550151824951172,
|
|
"epoch": 0.009943356385534093,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.000222,
|
|
"loss": 6.8211,
|
|
"mean_token_accuracy": 0.1224476382136345,
|
|
"num_tokens": 373734.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.477874803543091,
|
|
"epoch": 0.010055079490989532,
|
|
"grad_norm": 1.6328125,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 6.8174,
|
|
"mean_token_accuracy": 0.12541017457842826,
|
|
"num_tokens": 378184.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.454969882965088,
|
|
"epoch": 0.01016680259644497,
|
|
"grad_norm": 1.3828125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.9275,
|
|
"mean_token_accuracy": 0.11591533869504929,
|
|
"num_tokens": 382686.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.344128131866455,
|
|
"epoch": 0.01027852570190041,
|
|
"grad_norm": 1.6640625,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.8348,
|
|
"mean_token_accuracy": 0.12774229273200036,
|
|
"num_tokens": 387210.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.421209812164307,
|
|
"epoch": 0.01039024880735585,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.8216,
|
|
"mean_token_accuracy": 0.12031811997294425,
|
|
"num_tokens": 391247.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.465836143493652,
|
|
"epoch": 0.010501971912811289,
|
|
"grad_norm": 1.4921875,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.6907,
|
|
"mean_token_accuracy": 0.13249015137553216,
|
|
"num_tokens": 395188.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.280446100234985,
|
|
"epoch": 0.010613695018266728,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.7666,
|
|
"mean_token_accuracy": 0.12789778113365174,
|
|
"num_tokens": 399513.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.51575026512146,
|
|
"epoch": 0.010725418123722167,
|
|
"grad_norm": 1.59375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 6.8438,
|
|
"mean_token_accuracy": 0.12305677309632301,
|
|
"num_tokens": 404052.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.260741233825684,
|
|
"epoch": 0.010837141229177606,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.000242,
|
|
"loss": 6.6724,
|
|
"mean_token_accuracy": 0.13464880287647246,
|
|
"num_tokens": 408299.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.3084382057189945,
|
|
"epoch": 0.010948864334633046,
|
|
"grad_norm": 1.5859375,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.6808,
|
|
"mean_token_accuracy": 0.12868183255195617,
|
|
"num_tokens": 412376.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.318713665008545,
|
|
"epoch": 0.011060587440088485,
|
|
"grad_norm": 1.484375,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.7365,
|
|
"mean_token_accuracy": 0.12830623611807823,
|
|
"num_tokens": 417010.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.343731689453125,
|
|
"epoch": 0.011172310545543924,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.615,
|
|
"mean_token_accuracy": 0.1360365241765976,
|
|
"num_tokens": 421167.0,
|
|
"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": 91074473164800.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|