| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.23076923076923, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09615384615384616, | |
| "grad_norm": 14.387816429138184, | |
| "learning_rate": 4.6000000000000004e-07, | |
| "loss": 1.0674, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 11.297602653503418, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 0.9654, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "grad_norm": 8.031381607055664, | |
| "learning_rate": 1.46e-06, | |
| "loss": 0.7994, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 6.947944641113281, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 0.5243, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "grad_norm": 5.577055931091309, | |
| "learning_rate": 2.46e-06, | |
| "loss": 0.4378, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 5.237943649291992, | |
| "learning_rate": 2.96e-06, | |
| "loss": 0.39, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "grad_norm": 5.476370811462402, | |
| "learning_rate": 3.46e-06, | |
| "loss": 0.3755, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 6.1683244705200195, | |
| "learning_rate": 3.96e-06, | |
| "loss": 0.3851, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "grad_norm": 5.542613506317139, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 0.3461, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 7.256478786468506, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 0.3369, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0576923076923077, | |
| "grad_norm": 5.407237529754639, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 0.2975, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 5.3987579345703125, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.2739, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 4.13232946395874, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 0.2519, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 4.822619438171387, | |
| "learning_rate": 6.96e-06, | |
| "loss": 0.2663, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4423076923076923, | |
| "grad_norm": 4.323586463928223, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.2482, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 4.278296947479248, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.2525, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.6346153846153846, | |
| "grad_norm": 5.396093368530273, | |
| "learning_rate": 8.46e-06, | |
| "loss": 0.2575, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 4.330433368682861, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.2502, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.8269230769230769, | |
| "grad_norm": 4.134779930114746, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.2615, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 4.702469348907471, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.2542, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.019230769230769, | |
| "grad_norm": 2.5487704277038574, | |
| "learning_rate": 9.94888888888889e-06, | |
| "loss": 0.2502, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "grad_norm": 3.5824356079101562, | |
| "learning_rate": 9.893333333333334e-06, | |
| "loss": 0.1373, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.2115384615384617, | |
| "grad_norm": 3.7268402576446533, | |
| "learning_rate": 9.837777777777778e-06, | |
| "loss": 0.1367, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 3.243474245071411, | |
| "learning_rate": 9.782222222222222e-06, | |
| "loss": 0.1298, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4038461538461537, | |
| "grad_norm": 3.1370737552642822, | |
| "learning_rate": 9.726666666666668e-06, | |
| "loss": 0.135, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.788139581680298, | |
| "learning_rate": 9.671111111111112e-06, | |
| "loss": 0.1239, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.5961538461538463, | |
| "grad_norm": 3.4849274158477783, | |
| "learning_rate": 9.615555555555558e-06, | |
| "loss": 0.1226, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 3.447479009628296, | |
| "learning_rate": 9.56e-06, | |
| "loss": 0.1281, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.7884615384615383, | |
| "grad_norm": 3.737367868423462, | |
| "learning_rate": 9.504444444444446e-06, | |
| "loss": 0.121, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "grad_norm": 3.7365550994873047, | |
| "learning_rate": 9.44888888888889e-06, | |
| "loss": 0.1284, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.980769230769231, | |
| "grad_norm": 4.205287933349609, | |
| "learning_rate": 9.393333333333334e-06, | |
| "loss": 0.1393, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 2.6460373401641846, | |
| "learning_rate": 9.33777777777778e-06, | |
| "loss": 0.0834, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.173076923076923, | |
| "grad_norm": 2.1553828716278076, | |
| "learning_rate": 9.282222222222222e-06, | |
| "loss": 0.0541, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.269230769230769, | |
| "grad_norm": 1.773935079574585, | |
| "learning_rate": 9.226666666666668e-06, | |
| "loss": 0.0559, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.3653846153846154, | |
| "grad_norm": 2.047421455383301, | |
| "learning_rate": 9.171111111111112e-06, | |
| "loss": 0.0618, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 2.5151689052581787, | |
| "learning_rate": 9.115555555555556e-06, | |
| "loss": 0.0604, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.5576923076923075, | |
| "grad_norm": 1.7141085863113403, | |
| "learning_rate": 9.060000000000001e-06, | |
| "loss": 0.0567, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.6538461538461537, | |
| "grad_norm": 2.3757855892181396, | |
| "learning_rate": 9.004444444444445e-06, | |
| "loss": 0.0567, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 2.048637866973877, | |
| "learning_rate": 8.94888888888889e-06, | |
| "loss": 0.0595, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 2.78825306892395, | |
| "learning_rate": 8.893333333333333e-06, | |
| "loss": 0.0547, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.2570088803768158, | |
| "eval_runtime": 418.585, | |
| "eval_samples_per_second": 8.698, | |
| "eval_steps_per_second": 0.545, | |
| "eval_wer": 0.16873172769038006, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.9423076923076925, | |
| "grad_norm": 2.7837064266204834, | |
| "learning_rate": 8.83777777777778e-06, | |
| "loss": 0.0566, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.038461538461538, | |
| "grad_norm": 1.4186588525772095, | |
| "learning_rate": 8.782222222222223e-06, | |
| "loss": 0.0472, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.134615384615385, | |
| "grad_norm": 1.4111158847808838, | |
| "learning_rate": 8.726666666666667e-06, | |
| "loss": 0.026, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.8355940580368042, | |
| "learning_rate": 8.671111111111113e-06, | |
| "loss": 0.025, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.326923076923077, | |
| "grad_norm": 2.8829832077026367, | |
| "learning_rate": 8.615555555555555e-06, | |
| "loss": 0.0281, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.423076923076923, | |
| "grad_norm": 1.8035304546356201, | |
| "learning_rate": 8.560000000000001e-06, | |
| "loss": 0.0275, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.519230769230769, | |
| "grad_norm": 2.558990240097046, | |
| "learning_rate": 8.504444444444445e-06, | |
| "loss": 0.0284, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 3.0063929557800293, | |
| "learning_rate": 8.448888888888889e-06, | |
| "loss": 0.0262, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.711538461538462, | |
| "grad_norm": 1.181315541267395, | |
| "learning_rate": 8.393333333333335e-06, | |
| "loss": 0.0224, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "grad_norm": 1.4534687995910645, | |
| "learning_rate": 8.337777777777777e-06, | |
| "loss": 0.026, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.903846153846154, | |
| "grad_norm": 1.941493034362793, | |
| "learning_rate": 8.282222222222223e-06, | |
| "loss": 0.0276, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.048115849494934, | |
| "learning_rate": 8.226666666666667e-06, | |
| "loss": 0.0278, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.096153846153846, | |
| "grad_norm": 1.3306798934936523, | |
| "learning_rate": 8.171111111111113e-06, | |
| "loss": 0.0136, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 5.1923076923076925, | |
| "grad_norm": 1.5832746028900146, | |
| "learning_rate": 8.115555555555557e-06, | |
| "loss": 0.0128, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.288461538461538, | |
| "grad_norm": 1.175854206085205, | |
| "learning_rate": 8.06e-06, | |
| "loss": 0.0115, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 1.0961897373199463, | |
| "learning_rate": 8.004444444444445e-06, | |
| "loss": 0.0128, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.480769230769231, | |
| "grad_norm": 1.1296463012695312, | |
| "learning_rate": 7.948888888888889e-06, | |
| "loss": 0.0116, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 5.576923076923077, | |
| "grad_norm": 1.3500127792358398, | |
| "learning_rate": 7.893333333333335e-06, | |
| "loss": 0.0127, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.673076923076923, | |
| "grad_norm": 1.6575067043304443, | |
| "learning_rate": 7.837777777777779e-06, | |
| "loss": 0.0126, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 2.729482412338257, | |
| "learning_rate": 7.782222222222223e-06, | |
| "loss": 0.0125, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.865384615384615, | |
| "grad_norm": 1.4455863237380981, | |
| "learning_rate": 7.726666666666667e-06, | |
| "loss": 0.0122, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 5.961538461538462, | |
| "grad_norm": 1.5944123268127441, | |
| "learning_rate": 7.67111111111111e-06, | |
| "loss": 0.0147, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 6.0576923076923075, | |
| "grad_norm": 0.8240692615509033, | |
| "learning_rate": 7.6155555555555564e-06, | |
| "loss": 0.0094, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.41078874468803406, | |
| "learning_rate": 7.5600000000000005e-06, | |
| "loss": 0.0063, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 0.2937760651111603, | |
| "learning_rate": 7.504444444444445e-06, | |
| "loss": 0.008, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 6.346153846153846, | |
| "grad_norm": 0.2572005093097687, | |
| "learning_rate": 7.44888888888889e-06, | |
| "loss": 0.006, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.4423076923076925, | |
| "grad_norm": 0.5730597376823425, | |
| "learning_rate": 7.393333333333333e-06, | |
| "loss": 0.0083, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 6.538461538461538, | |
| "grad_norm": 0.9147136211395264, | |
| "learning_rate": 7.337777777777778e-06, | |
| "loss": 0.0066, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 6.634615384615385, | |
| "grad_norm": 0.45330244302749634, | |
| "learning_rate": 7.282222222222222e-06, | |
| "loss": 0.0065, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 6.730769230769231, | |
| "grad_norm": 0.6328800916671753, | |
| "learning_rate": 7.226666666666667e-06, | |
| "loss": 0.0078, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 6.826923076923077, | |
| "grad_norm": 0.2872777581214905, | |
| "learning_rate": 7.171111111111112e-06, | |
| "loss": 0.0079, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 1.7963801622390747, | |
| "learning_rate": 7.115555555555557e-06, | |
| "loss": 0.0078, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.019230769230769, | |
| "grad_norm": 0.2437676042318344, | |
| "learning_rate": 7.06e-06, | |
| "loss": 0.0068, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 7.115384615384615, | |
| "grad_norm": 0.20286525785923004, | |
| "learning_rate": 7.004444444444445e-06, | |
| "loss": 0.0039, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.211538461538462, | |
| "grad_norm": 0.16271603107452393, | |
| "learning_rate": 6.948888888888889e-06, | |
| "loss": 0.0038, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 7.3076923076923075, | |
| "grad_norm": 0.12690773606300354, | |
| "learning_rate": 6.893333333333334e-06, | |
| "loss": 0.0037, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.403846153846154, | |
| "grad_norm": 0.41975799202919006, | |
| "learning_rate": 6.837777777777779e-06, | |
| "loss": 0.0042, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.2555113732814789, | |
| "learning_rate": 6.782222222222222e-06, | |
| "loss": 0.0033, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 7.596153846153846, | |
| "grad_norm": 1.6203404664993286, | |
| "learning_rate": 6.726666666666667e-06, | |
| "loss": 0.0049, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 0.16367614269256592, | |
| "learning_rate": 6.671111111111112e-06, | |
| "loss": 0.0034, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "eval_loss": 0.31435152888298035, | |
| "eval_runtime": 414.9289, | |
| "eval_samples_per_second": 8.775, | |
| "eval_steps_per_second": 0.549, | |
| "eval_wer": 0.15889368416167804, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.788461538461538, | |
| "grad_norm": 0.1244119256734848, | |
| "learning_rate": 6.615555555555556e-06, | |
| "loss": 0.0035, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 7.884615384615385, | |
| "grad_norm": 0.7467623353004456, | |
| "learning_rate": 6.560000000000001e-06, | |
| "loss": 0.0042, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 7.980769230769231, | |
| "grad_norm": 0.26792147755622864, | |
| "learning_rate": 6.504444444444446e-06, | |
| "loss": 0.0043, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 8.076923076923077, | |
| "grad_norm": 0.0888567715883255, | |
| "learning_rate": 6.448888888888889e-06, | |
| "loss": 0.0029, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.173076923076923, | |
| "grad_norm": 0.22752462327480316, | |
| "learning_rate": 6.393333333333334e-06, | |
| "loss": 0.003, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 8.26923076923077, | |
| "grad_norm": 0.11363103985786438, | |
| "learning_rate": 6.3377777777777786e-06, | |
| "loss": 0.0023, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 8.365384615384615, | |
| "grad_norm": 0.4889778196811676, | |
| "learning_rate": 6.282222222222223e-06, | |
| "loss": 0.0024, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 0.08544881641864777, | |
| "learning_rate": 6.2266666666666675e-06, | |
| "loss": 0.0023, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.557692307692308, | |
| "grad_norm": 0.1377352476119995, | |
| "learning_rate": 6.171111111111112e-06, | |
| "loss": 0.0032, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 8.653846153846153, | |
| "grad_norm": 0.49440711736679077, | |
| "learning_rate": 6.1155555555555555e-06, | |
| "loss": 0.0022, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 0.10170517861843109, | |
| "learning_rate": 6.0600000000000004e-06, | |
| "loss": 0.0029, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 8.846153846153847, | |
| "grad_norm": 0.0798933133482933, | |
| "learning_rate": 6.004444444444445e-06, | |
| "loss": 0.0025, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 8.942307692307692, | |
| "grad_norm": 0.06800112873315811, | |
| "learning_rate": 5.948888888888889e-06, | |
| "loss": 0.0023, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 9.038461538461538, | |
| "grad_norm": 0.05597899481654167, | |
| "learning_rate": 5.893333333333334e-06, | |
| "loss": 0.0019, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 9.134615384615385, | |
| "grad_norm": 0.055818960070610046, | |
| "learning_rate": 5.837777777777777e-06, | |
| "loss": 0.0016, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 0.07546485215425491, | |
| "learning_rate": 5.782222222222222e-06, | |
| "loss": 0.002, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.326923076923077, | |
| "grad_norm": 0.06564825028181076, | |
| "learning_rate": 5.726666666666667e-06, | |
| "loss": 0.0016, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 9.423076923076923, | |
| "grad_norm": 0.0399746373295784, | |
| "learning_rate": 5.671111111111112e-06, | |
| "loss": 0.0016, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 9.51923076923077, | |
| "grad_norm": 0.059612762182950974, | |
| "learning_rate": 5.615555555555556e-06, | |
| "loss": 0.0017, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 9.615384615384615, | |
| "grad_norm": 0.06947290152311325, | |
| "learning_rate": 5.560000000000001e-06, | |
| "loss": 0.0016, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.711538461538462, | |
| "grad_norm": 0.989651083946228, | |
| "learning_rate": 5.504444444444444e-06, | |
| "loss": 0.0017, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 9.807692307692308, | |
| "grad_norm": 0.07500547170639038, | |
| "learning_rate": 5.448888888888889e-06, | |
| "loss": 0.0016, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 9.903846153846153, | |
| "grad_norm": 0.09042783081531525, | |
| "learning_rate": 5.393333333333334e-06, | |
| "loss": 0.0017, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.3791869282722473, | |
| "learning_rate": 5.337777777777779e-06, | |
| "loss": 0.0017, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.096153846153847, | |
| "grad_norm": 0.043280228972435, | |
| "learning_rate": 5.282222222222223e-06, | |
| "loss": 0.0013, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 10.192307692307692, | |
| "grad_norm": 0.04278066009283066, | |
| "learning_rate": 5.226666666666667e-06, | |
| "loss": 0.0014, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 10.288461538461538, | |
| "grad_norm": 0.043745964765548706, | |
| "learning_rate": 5.171111111111111e-06, | |
| "loss": 0.0013, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 10.384615384615385, | |
| "grad_norm": 0.05145561695098877, | |
| "learning_rate": 5.115555555555556e-06, | |
| "loss": 0.0013, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 10.48076923076923, | |
| "grad_norm": 0.04810141772031784, | |
| "learning_rate": 5.060000000000001e-06, | |
| "loss": 0.0013, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 10.576923076923077, | |
| "grad_norm": 0.046568796038627625, | |
| "learning_rate": 5.004444444444445e-06, | |
| "loss": 0.002, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 10.673076923076923, | |
| "grad_norm": 0.05206083878874779, | |
| "learning_rate": 4.94888888888889e-06, | |
| "loss": 0.0013, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 10.76923076923077, | |
| "grad_norm": 0.04524935781955719, | |
| "learning_rate": 4.893333333333334e-06, | |
| "loss": 0.0013, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 10.865384615384615, | |
| "grad_norm": 0.05039019510149956, | |
| "learning_rate": 4.837777777777778e-06, | |
| "loss": 0.0012, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 10.961538461538462, | |
| "grad_norm": 0.048185594379901886, | |
| "learning_rate": 4.7822222222222226e-06, | |
| "loss": 0.0024, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 11.057692307692308, | |
| "grad_norm": 0.04220257326960564, | |
| "learning_rate": 4.7266666666666674e-06, | |
| "loss": 0.0012, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 11.153846153846153, | |
| "grad_norm": 0.04200848937034607, | |
| "learning_rate": 4.6711111111111115e-06, | |
| "loss": 0.0011, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "grad_norm": 0.04376620426774025, | |
| "learning_rate": 4.6155555555555555e-06, | |
| "loss": 0.0011, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 11.346153846153847, | |
| "grad_norm": 0.03870835155248642, | |
| "learning_rate": 4.56e-06, | |
| "loss": 0.0011, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 11.442307692307692, | |
| "grad_norm": 0.03991697356104851, | |
| "learning_rate": 4.504444444444444e-06, | |
| "loss": 0.0011, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 11.538461538461538, | |
| "grad_norm": 0.03659369796514511, | |
| "learning_rate": 4.448888888888889e-06, | |
| "loss": 0.0011, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.538461538461538, | |
| "eval_loss": 0.33577749133110046, | |
| "eval_runtime": 414.076, | |
| "eval_samples_per_second": 8.793, | |
| "eval_steps_per_second": 0.551, | |
| "eval_wer": 0.16153881850665924, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.634615384615385, | |
| "grad_norm": 0.035982731729745865, | |
| "learning_rate": 4.393333333333334e-06, | |
| "loss": 0.0011, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 11.73076923076923, | |
| "grad_norm": 0.04060469567775726, | |
| "learning_rate": 4.337777777777778e-06, | |
| "loss": 0.0011, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 11.826923076923077, | |
| "grad_norm": 0.03556708246469498, | |
| "learning_rate": 4.282222222222222e-06, | |
| "loss": 0.0011, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 11.923076923076923, | |
| "grad_norm": 0.037598781287670135, | |
| "learning_rate": 4.226666666666667e-06, | |
| "loss": 0.001, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.01923076923077, | |
| "grad_norm": 0.03427188843488693, | |
| "learning_rate": 4.171111111111111e-06, | |
| "loss": 0.0011, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 12.115384615384615, | |
| "grad_norm": 0.03573931008577347, | |
| "learning_rate": 4.115555555555556e-06, | |
| "loss": 0.0009, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 12.211538461538462, | |
| "grad_norm": 0.03210974112153053, | |
| "learning_rate": 4.060000000000001e-06, | |
| "loss": 0.0009, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 12.307692307692308, | |
| "grad_norm": 0.036103058606386185, | |
| "learning_rate": 4.004444444444445e-06, | |
| "loss": 0.001, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 12.403846153846153, | |
| "grad_norm": 0.036252211779356, | |
| "learning_rate": 3.948888888888889e-06, | |
| "loss": 0.001, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.03289297968149185, | |
| "learning_rate": 3.893333333333333e-06, | |
| "loss": 0.0009, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 12.596153846153847, | |
| "grad_norm": 0.035281531512737274, | |
| "learning_rate": 3.837777777777778e-06, | |
| "loss": 0.001, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 12.692307692307692, | |
| "grad_norm": 0.036654986441135406, | |
| "learning_rate": 3.782222222222223e-06, | |
| "loss": 0.0009, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 12.788461538461538, | |
| "grad_norm": 0.028989532962441444, | |
| "learning_rate": 3.726666666666667e-06, | |
| "loss": 0.0009, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 12.884615384615385, | |
| "grad_norm": 0.033068809658288956, | |
| "learning_rate": 3.6711111111111113e-06, | |
| "loss": 0.0009, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 12.98076923076923, | |
| "grad_norm": 0.035067297518253326, | |
| "learning_rate": 3.615555555555556e-06, | |
| "loss": 0.0009, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 13.076923076923077, | |
| "grad_norm": 0.03311024233698845, | |
| "learning_rate": 3.5600000000000002e-06, | |
| "loss": 0.0008, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 13.173076923076923, | |
| "grad_norm": 0.0331607460975647, | |
| "learning_rate": 3.5044444444444447e-06, | |
| "loss": 0.0009, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 13.26923076923077, | |
| "grad_norm": 0.031973276287317276, | |
| "learning_rate": 3.4488888888888896e-06, | |
| "loss": 0.0015, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 13.365384615384615, | |
| "grad_norm": 0.031224718317389488, | |
| "learning_rate": 3.3933333333333336e-06, | |
| "loss": 0.0008, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 13.461538461538462, | |
| "grad_norm": 0.030508123338222504, | |
| "learning_rate": 3.337777777777778e-06, | |
| "loss": 0.0008, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 13.557692307692308, | |
| "grad_norm": 0.03208326920866966, | |
| "learning_rate": 3.282222222222223e-06, | |
| "loss": 0.0009, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 13.653846153846153, | |
| "grad_norm": 0.02819848246872425, | |
| "learning_rate": 3.226666666666667e-06, | |
| "loss": 0.0008, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "grad_norm": 0.02973184920847416, | |
| "learning_rate": 3.1711111111111114e-06, | |
| "loss": 0.0009, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 13.846153846153847, | |
| "grad_norm": 0.026373526081442833, | |
| "learning_rate": 3.1155555555555555e-06, | |
| "loss": 0.0008, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 13.942307692307692, | |
| "grad_norm": 0.030721429735422134, | |
| "learning_rate": 3.0600000000000003e-06, | |
| "loss": 0.0008, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 14.038461538461538, | |
| "grad_norm": 0.027194401249289513, | |
| "learning_rate": 3.004444444444445e-06, | |
| "loss": 0.0008, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 14.134615384615385, | |
| "grad_norm": 0.02467515505850315, | |
| "learning_rate": 2.948888888888889e-06, | |
| "loss": 0.0008, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 14.23076923076923, | |
| "grad_norm": 0.02789357490837574, | |
| "learning_rate": 2.8933333333333337e-06, | |
| "loss": 0.0008, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 14.326923076923077, | |
| "grad_norm": 0.02929651364684105, | |
| "learning_rate": 2.837777777777778e-06, | |
| "loss": 0.0008, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 14.423076923076923, | |
| "grad_norm": 0.02698478475213051, | |
| "learning_rate": 2.7822222222222222e-06, | |
| "loss": 0.0008, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 14.51923076923077, | |
| "grad_norm": 0.026549380272626877, | |
| "learning_rate": 2.726666666666667e-06, | |
| "loss": 0.0008, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 14.615384615384615, | |
| "grad_norm": 0.025898369029164314, | |
| "learning_rate": 2.6711111111111116e-06, | |
| "loss": 0.0008, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 14.711538461538462, | |
| "grad_norm": 0.024109981954097748, | |
| "learning_rate": 2.6155555555555556e-06, | |
| "loss": 0.0008, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 14.807692307692308, | |
| "grad_norm": 0.028975041583180428, | |
| "learning_rate": 2.56e-06, | |
| "loss": 0.0007, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 14.903846153846153, | |
| "grad_norm": 0.029065443202853203, | |
| "learning_rate": 2.504444444444445e-06, | |
| "loss": 0.0008, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.03024737536907196, | |
| "learning_rate": 2.448888888888889e-06, | |
| "loss": 0.0013, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 15.096153846153847, | |
| "grad_norm": 0.020746273919939995, | |
| "learning_rate": 2.3933333333333334e-06, | |
| "loss": 0.0007, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 15.192307692307692, | |
| "grad_norm": 0.022498076781630516, | |
| "learning_rate": 2.337777777777778e-06, | |
| "loss": 0.0007, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 15.288461538461538, | |
| "grad_norm": 0.025430675595998764, | |
| "learning_rate": 2.2822222222222223e-06, | |
| "loss": 0.0007, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "grad_norm": 0.024518582969903946, | |
| "learning_rate": 2.226666666666667e-06, | |
| "loss": 0.0007, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "eval_loss": 0.34604960680007935, | |
| "eval_runtime": 414.6061, | |
| "eval_samples_per_second": 8.782, | |
| "eval_steps_per_second": 0.55, | |
| "eval_wer": 0.1614460067752564, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 15.48076923076923, | |
| "grad_norm": 0.028795845806598663, | |
| "learning_rate": 2.1711111111111113e-06, | |
| "loss": 0.0007, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 15.576923076923077, | |
| "grad_norm": 0.02770942822098732, | |
| "learning_rate": 2.1155555555555557e-06, | |
| "loss": 0.0007, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 15.673076923076923, | |
| "grad_norm": 0.026291629299521446, | |
| "learning_rate": 2.06e-06, | |
| "loss": 0.0008, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 15.76923076923077, | |
| "grad_norm": 0.024974102154374123, | |
| "learning_rate": 2.0044444444444446e-06, | |
| "loss": 0.0007, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 15.865384615384615, | |
| "grad_norm": 0.02725864201784134, | |
| "learning_rate": 1.948888888888889e-06, | |
| "loss": 0.0007, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 15.961538461538462, | |
| "grad_norm": 0.027727024629712105, | |
| "learning_rate": 1.8933333333333333e-06, | |
| "loss": 0.0007, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 16.057692307692307, | |
| "grad_norm": 0.023184822872281075, | |
| "learning_rate": 1.837777777777778e-06, | |
| "loss": 0.0007, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 16.153846153846153, | |
| "grad_norm": 0.020365368574857712, | |
| "learning_rate": 1.7822222222222225e-06, | |
| "loss": 0.0007, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "grad_norm": 0.023292195051908493, | |
| "learning_rate": 1.7266666666666667e-06, | |
| "loss": 0.0006, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 16.346153846153847, | |
| "grad_norm": 0.02580893039703369, | |
| "learning_rate": 1.6711111111111112e-06, | |
| "loss": 0.0006, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 16.442307692307693, | |
| "grad_norm": 0.02440100722014904, | |
| "learning_rate": 1.6155555555555559e-06, | |
| "loss": 0.0007, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 16.53846153846154, | |
| "grad_norm": 0.02583976835012436, | |
| "learning_rate": 1.56e-06, | |
| "loss": 0.0007, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 16.634615384615383, | |
| "grad_norm": 0.024059804156422615, | |
| "learning_rate": 1.5044444444444446e-06, | |
| "loss": 0.0007, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 16.73076923076923, | |
| "grad_norm": 0.023279106244444847, | |
| "learning_rate": 1.4488888888888892e-06, | |
| "loss": 0.0007, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 16.826923076923077, | |
| "grad_norm": 0.022877462208271027, | |
| "learning_rate": 1.3933333333333335e-06, | |
| "loss": 0.0007, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 16.923076923076923, | |
| "grad_norm": 0.02401687018573284, | |
| "learning_rate": 1.337777777777778e-06, | |
| "loss": 0.0006, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 17.01923076923077, | |
| "grad_norm": 0.021333323791623116, | |
| "learning_rate": 1.2822222222222222e-06, | |
| "loss": 0.0006, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 17.115384615384617, | |
| "grad_norm": 0.019655166193842888, | |
| "learning_rate": 1.2266666666666666e-06, | |
| "loss": 0.0006, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 17.21153846153846, | |
| "grad_norm": 0.02106211706995964, | |
| "learning_rate": 1.171111111111111e-06, | |
| "loss": 0.0006, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 17.307692307692307, | |
| "grad_norm": 0.022571563720703125, | |
| "learning_rate": 1.1155555555555558e-06, | |
| "loss": 0.0006, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 17.403846153846153, | |
| "grad_norm": 0.023800240829586983, | |
| "learning_rate": 1.06e-06, | |
| "loss": 0.0006, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 0.0228471290320158, | |
| "learning_rate": 1.0044444444444445e-06, | |
| "loss": 0.0006, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 17.596153846153847, | |
| "grad_norm": 0.021810224279761314, | |
| "learning_rate": 9.488888888888889e-07, | |
| "loss": 0.0006, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 17.692307692307693, | |
| "grad_norm": 0.02282741293311119, | |
| "learning_rate": 8.933333333333334e-07, | |
| "loss": 0.0008, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 17.78846153846154, | |
| "grad_norm": 0.024272125214338303, | |
| "learning_rate": 8.37777777777778e-07, | |
| "loss": 0.0006, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 17.884615384615383, | |
| "grad_norm": 0.026012342423200607, | |
| "learning_rate": 7.822222222222223e-07, | |
| "loss": 0.0006, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 17.98076923076923, | |
| "grad_norm": 0.02298705466091633, | |
| "learning_rate": 7.266666666666668e-07, | |
| "loss": 0.0006, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 18.076923076923077, | |
| "grad_norm": 0.021737568080425262, | |
| "learning_rate": 6.711111111111111e-07, | |
| "loss": 0.0006, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 18.173076923076923, | |
| "grad_norm": 0.021294375881552696, | |
| "learning_rate": 6.155555555555556e-07, | |
| "loss": 0.0006, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 18.26923076923077, | |
| "grad_norm": 0.0215039923787117, | |
| "learning_rate": 5.6e-07, | |
| "loss": 0.0006, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 18.365384615384617, | |
| "grad_norm": 0.02031785622239113, | |
| "learning_rate": 5.044444444444445e-07, | |
| "loss": 0.0006, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 18.46153846153846, | |
| "grad_norm": 0.02249339409172535, | |
| "learning_rate": 4.488888888888889e-07, | |
| "loss": 0.0006, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 18.557692307692307, | |
| "grad_norm": 0.022298617288470268, | |
| "learning_rate": 3.9333333333333336e-07, | |
| "loss": 0.0006, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 18.653846153846153, | |
| "grad_norm": 0.019332217052578926, | |
| "learning_rate": 3.3777777777777777e-07, | |
| "loss": 0.0006, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "grad_norm": 0.02518359012901783, | |
| "learning_rate": 2.822222222222222e-07, | |
| "loss": 0.0006, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 18.846153846153847, | |
| "grad_norm": 0.022782141342759132, | |
| "learning_rate": 2.266666666666667e-07, | |
| "loss": 0.0006, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 18.942307692307693, | |
| "grad_norm": 0.023741982877254486, | |
| "learning_rate": 1.7111111111111114e-07, | |
| "loss": 0.0006, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 19.03846153846154, | |
| "grad_norm": 0.02083681896328926, | |
| "learning_rate": 1.1555555555555556e-07, | |
| "loss": 0.0006, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 19.134615384615383, | |
| "grad_norm": 0.020423738285899162, | |
| "learning_rate": 6.000000000000001e-08, | |
| "loss": 0.0006, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "grad_norm": 0.0213633980602026, | |
| "learning_rate": 4.444444444444445e-09, | |
| "loss": 0.0006, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "eval_loss": 0.35056987404823303, | |
| "eval_runtime": 416.1932, | |
| "eval_samples_per_second": 8.748, | |
| "eval_steps_per_second": 0.548, | |
| "eval_wer": 0.16172444196946495, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "step": 5000, | |
| "total_flos": 4.616818092859392e+19, | |
| "train_loss": 0.05540717707388103, | |
| "train_runtime": 18195.7149, | |
| "train_samples_per_second": 8.793, | |
| "train_steps_per_second": 0.275 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.616818092859392e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |