| { | |
| "best_metric": 18.301895430821354, | |
| "best_model_checkpoint": "./checkpoint-2000", | |
| "epoch": 39.119804400978, | |
| "eval_steps": 1000, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 39.86157989501953, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.0555, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.800000000000001e-07, | |
| "loss": 1.5219, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 6.197519779205322, | |
| "learning_rate": 1.48e-06, | |
| "loss": 1.0167, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 5.485505104064941, | |
| "learning_rate": 1.98e-06, | |
| "loss": 0.7299, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 5.534335613250732, | |
| "learning_rate": 2.4800000000000004e-06, | |
| "loss": 0.6317, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 4.898209095001221, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 0.5503, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 5.0602946281433105, | |
| "learning_rate": 3.48e-06, | |
| "loss": 0.4998, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 4.8069305419921875, | |
| "learning_rate": 3.980000000000001e-06, | |
| "loss": 0.4457, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 4.571457386016846, | |
| "learning_rate": 4.48e-06, | |
| "loss": 0.3778, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 4.707397937774658, | |
| "learning_rate": 4.980000000000001e-06, | |
| "loss": 0.333, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 4.317914009094238, | |
| "learning_rate": 5.480000000000001e-06, | |
| "loss": 0.2859, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.6433582305908203, | |
| "learning_rate": 5.98e-06, | |
| "loss": 0.2229, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 2.5579586029052734, | |
| "learning_rate": 6.480000000000001e-06, | |
| "loss": 0.1767, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 2.2651941776275635, | |
| "learning_rate": 6.98e-06, | |
| "loss": 0.1474, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 2.528773546218872, | |
| "learning_rate": 7.48e-06, | |
| "loss": 0.1493, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 2.422232151031494, | |
| "learning_rate": 7.980000000000002e-06, | |
| "loss": 0.142, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 2.209630012512207, | |
| "learning_rate": 8.48e-06, | |
| "loss": 0.1126, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.9945831298828125, | |
| "learning_rate": 8.98e-06, | |
| "loss": 0.0909, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 1.8972020149230957, | |
| "learning_rate": 9.48e-06, | |
| "loss": 0.0892, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 2.345607042312622, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.0909, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 1.3913277387619019, | |
| "learning_rate": 9.946666666666667e-06, | |
| "loss": 0.0713, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "grad_norm": 2.5306901931762695, | |
| "learning_rate": 9.891111111111113e-06, | |
| "loss": 0.0535, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 1.3689407110214233, | |
| "learning_rate": 9.835555555555556e-06, | |
| "loss": 0.051, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "grad_norm": 2.091756820678711, | |
| "learning_rate": 9.780000000000001e-06, | |
| "loss": 0.0573, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.11, | |
| "grad_norm": 1.5179609060287476, | |
| "learning_rate": 9.724444444444445e-06, | |
| "loss": 0.0454, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "grad_norm": 1.3439680337905884, | |
| "learning_rate": 9.66888888888889e-06, | |
| "loss": 0.029, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 1.279569387435913, | |
| "learning_rate": 9.613333333333335e-06, | |
| "loss": 0.032, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "grad_norm": 1.6171939373016357, | |
| "learning_rate": 9.557777777777777e-06, | |
| "loss": 0.0314, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "grad_norm": 0.8949939608573914, | |
| "learning_rate": 9.502222222222223e-06, | |
| "loss": 0.0245, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "grad_norm": 1.3610639572143555, | |
| "learning_rate": 9.446666666666667e-06, | |
| "loss": 0.0165, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 1.5446442365646362, | |
| "learning_rate": 9.391111111111111e-06, | |
| "loss": 0.0169, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "grad_norm": 1.2825671434402466, | |
| "learning_rate": 9.335555555555557e-06, | |
| "loss": 0.0193, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "grad_norm": 0.7063258290290833, | |
| "learning_rate": 9.280000000000001e-06, | |
| "loss": 0.0154, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 8.31, | |
| "grad_norm": 0.72287917137146, | |
| "learning_rate": 9.224444444444445e-06, | |
| "loss": 0.0102, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 8.56, | |
| "grad_norm": 1.2877657413482666, | |
| "learning_rate": 9.168888888888889e-06, | |
| "loss": 0.0093, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 1.6262348890304565, | |
| "learning_rate": 9.113333333333335e-06, | |
| "loss": 0.0117, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "grad_norm": 0.49374520778656006, | |
| "learning_rate": 9.057777777777779e-06, | |
| "loss": 0.0121, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "grad_norm": 0.8290354013442993, | |
| "learning_rate": 9.002222222222223e-06, | |
| "loss": 0.0063, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 9.54, | |
| "grad_norm": 0.9974227547645569, | |
| "learning_rate": 8.946666666666669e-06, | |
| "loss": 0.006, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "grad_norm": 0.8566415309906006, | |
| "learning_rate": 8.891111111111111e-06, | |
| "loss": 0.007, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.78, | |
| "eval_loss": 0.41129249334335327, | |
| "eval_runtime": 1450.197, | |
| "eval_samples_per_second": 1.996, | |
| "eval_steps_per_second": 0.499, | |
| "eval_wer": 18.926838201629888, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 1.7889363765716553, | |
| "learning_rate": 8.835555555555557e-06, | |
| "loss": 0.0067, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 10.27, | |
| "grad_norm": 0.9269456267356873, | |
| "learning_rate": 8.78e-06, | |
| "loss": 0.0047, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 10.51, | |
| "grad_norm": 0.7966364622116089, | |
| "learning_rate": 8.724444444444445e-06, | |
| "loss": 0.0047, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 10.76, | |
| "grad_norm": 0.6406449675559998, | |
| "learning_rate": 8.66888888888889e-06, | |
| "loss": 0.0049, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.4881477355957031, | |
| "learning_rate": 8.613333333333333e-06, | |
| "loss": 0.0049, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "grad_norm": 0.4843809902667999, | |
| "learning_rate": 8.557777777777778e-06, | |
| "loss": 0.0036, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "grad_norm": 0.8177527189254761, | |
| "learning_rate": 8.502222222222223e-06, | |
| "loss": 0.0028, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "grad_norm": 0.676511287689209, | |
| "learning_rate": 8.446666666666668e-06, | |
| "loss": 0.0031, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.98, | |
| "grad_norm": 0.9199188351631165, | |
| "learning_rate": 8.391111111111112e-06, | |
| "loss": 0.003, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "grad_norm": 0.248287171125412, | |
| "learning_rate": 8.335555555555556e-06, | |
| "loss": 0.0024, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 12.47, | |
| "grad_norm": 0.9288859963417053, | |
| "learning_rate": 8.28e-06, | |
| "loss": 0.0022, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 12.71, | |
| "grad_norm": 0.8308430910110474, | |
| "learning_rate": 8.224444444444444e-06, | |
| "loss": 0.0025, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.96, | |
| "grad_norm": 0.33064720034599304, | |
| "learning_rate": 8.16888888888889e-06, | |
| "loss": 0.0027, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "grad_norm": 0.5254776477813721, | |
| "learning_rate": 8.113333333333334e-06, | |
| "loss": 0.0019, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 13.45, | |
| "grad_norm": 0.4681105315685272, | |
| "learning_rate": 8.057777777777778e-06, | |
| "loss": 0.0021, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 13.69, | |
| "grad_norm": 0.352115273475647, | |
| "learning_rate": 8.002222222222222e-06, | |
| "loss": 0.0026, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.94, | |
| "grad_norm": 1.5301597118377686, | |
| "learning_rate": 7.946666666666666e-06, | |
| "loss": 0.0027, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 14.18, | |
| "grad_norm": 0.28577202558517456, | |
| "learning_rate": 7.891111111111112e-06, | |
| "loss": 0.0018, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 14.43, | |
| "grad_norm": 0.7940084338188171, | |
| "learning_rate": 7.835555555555556e-06, | |
| "loss": 0.002, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 14.67, | |
| "grad_norm": 1.031543493270874, | |
| "learning_rate": 7.78e-06, | |
| "loss": 0.0021, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.91, | |
| "grad_norm": 0.5695396661758423, | |
| "learning_rate": 7.724444444444446e-06, | |
| "loss": 0.0016, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 15.16, | |
| "grad_norm": 0.2285182625055313, | |
| "learning_rate": 7.66888888888889e-06, | |
| "loss": 0.0014, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 15.4, | |
| "grad_norm": 0.40289613604545593, | |
| "learning_rate": 7.613333333333334e-06, | |
| "loss": 0.0017, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "grad_norm": 0.5758986473083496, | |
| "learning_rate": 7.557777777777779e-06, | |
| "loss": 0.0012, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 15.89, | |
| "grad_norm": 0.5524174571037292, | |
| "learning_rate": 7.502222222222223e-06, | |
| "loss": 0.001, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 16.14, | |
| "grad_norm": 0.7031832933425903, | |
| "learning_rate": 7.446666666666668e-06, | |
| "loss": 0.0015, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 16.38, | |
| "grad_norm": 0.22649440169334412, | |
| "learning_rate": 7.3911111111111125e-06, | |
| "loss": 0.0012, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 16.63, | |
| "grad_norm": 0.46751469373703003, | |
| "learning_rate": 7.335555555555556e-06, | |
| "loss": 0.0012, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 16.87, | |
| "grad_norm": 0.3201611340045929, | |
| "learning_rate": 7.280000000000001e-06, | |
| "loss": 0.0009, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 17.11, | |
| "grad_norm": 0.05176452174782753, | |
| "learning_rate": 7.224444444444445e-06, | |
| "loss": 0.0007, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 17.36, | |
| "grad_norm": 0.5956466794013977, | |
| "learning_rate": 7.1688888888888895e-06, | |
| "loss": 0.0007, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "grad_norm": 0.1542888581752777, | |
| "learning_rate": 7.113333333333334e-06, | |
| "loss": 0.001, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 17.85, | |
| "grad_norm": 0.624476432800293, | |
| "learning_rate": 7.057777777777778e-06, | |
| "loss": 0.0016, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 18.09, | |
| "grad_norm": 0.3069753050804138, | |
| "learning_rate": 7.0022222222222225e-06, | |
| "loss": 0.0013, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 18.34, | |
| "grad_norm": 0.3797595500946045, | |
| "learning_rate": 6.946666666666667e-06, | |
| "loss": 0.0018, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 18.58, | |
| "grad_norm": 0.3482624590396881, | |
| "learning_rate": 6.891111111111111e-06, | |
| "loss": 0.0014, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 18.83, | |
| "grad_norm": 0.05667097494006157, | |
| "learning_rate": 6.835555555555556e-06, | |
| "loss": 0.0008, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 19.07, | |
| "grad_norm": 0.09759561717510223, | |
| "learning_rate": 6.780000000000001e-06, | |
| "loss": 0.0011, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 19.32, | |
| "grad_norm": 0.5752401947975159, | |
| "learning_rate": 6.724444444444444e-06, | |
| "loss": 0.0014, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 19.56, | |
| "grad_norm": 0.1028558760881424, | |
| "learning_rate": 6.668888888888889e-06, | |
| "loss": 0.0009, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 19.56, | |
| "eval_loss": 0.4927152395248413, | |
| "eval_runtime": 1461.0654, | |
| "eval_samples_per_second": 1.981, | |
| "eval_steps_per_second": 0.496, | |
| "eval_wer": 18.301895430821354, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "grad_norm": 0.28834694623947144, | |
| "learning_rate": 6.613333333333334e-06, | |
| "loss": 0.0012, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 20.05, | |
| "grad_norm": 0.38117802143096924, | |
| "learning_rate": 6.557777777777778e-06, | |
| "loss": 0.0017, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 20.29, | |
| "grad_norm": 0.6632546782493591, | |
| "learning_rate": 6.502222222222223e-06, | |
| "loss": 0.001, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 20.54, | |
| "grad_norm": 0.32943013310432434, | |
| "learning_rate": 6.446666666666668e-06, | |
| "loss": 0.0017, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 20.78, | |
| "grad_norm": 0.6536120772361755, | |
| "learning_rate": 6.391111111111111e-06, | |
| "loss": 0.001, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "grad_norm": 0.6113318204879761, | |
| "learning_rate": 6.335555555555556e-06, | |
| "loss": 0.0009, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 21.27, | |
| "grad_norm": 0.24811410903930664, | |
| "learning_rate": 6.280000000000001e-06, | |
| "loss": 0.0012, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 21.52, | |
| "grad_norm": 1.5321826934814453, | |
| "learning_rate": 6.224444444444445e-06, | |
| "loss": 0.0015, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 21.76, | |
| "grad_norm": 0.6304628252983093, | |
| "learning_rate": 6.16888888888889e-06, | |
| "loss": 0.002, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.4291835427284241, | |
| "learning_rate": 6.113333333333333e-06, | |
| "loss": 0.0017, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 22.25, | |
| "grad_norm": 1.250555396080017, | |
| "learning_rate": 6.057777777777778e-06, | |
| "loss": 0.0014, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 22.49, | |
| "grad_norm": 0.14119359850883484, | |
| "learning_rate": 6.002222222222223e-06, | |
| "loss": 0.0011, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 22.74, | |
| "grad_norm": 0.6686920523643494, | |
| "learning_rate": 5.946666666666668e-06, | |
| "loss": 0.0013, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 22.98, | |
| "grad_norm": 0.6255317330360413, | |
| "learning_rate": 5.891111111111112e-06, | |
| "loss": 0.001, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 23.23, | |
| "grad_norm": 0.08858698606491089, | |
| "learning_rate": 5.8355555555555565e-06, | |
| "loss": 0.0006, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "grad_norm": 0.08142836391925812, | |
| "learning_rate": 5.78e-06, | |
| "loss": 0.0005, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 23.72, | |
| "grad_norm": 1.0621262788772583, | |
| "learning_rate": 5.724444444444445e-06, | |
| "loss": 0.0005, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 23.96, | |
| "grad_norm": 0.04572203755378723, | |
| "learning_rate": 5.6688888888888895e-06, | |
| "loss": 0.0004, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 24.21, | |
| "grad_norm": 0.03155618906021118, | |
| "learning_rate": 5.613333333333334e-06, | |
| "loss": 0.0002, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 24.45, | |
| "grad_norm": 0.03207828477025032, | |
| "learning_rate": 5.557777777777778e-06, | |
| "loss": 0.0003, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 24.69, | |
| "grad_norm": 0.0497412383556366, | |
| "learning_rate": 5.5022222222222224e-06, | |
| "loss": 0.0003, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 24.94, | |
| "grad_norm": 0.21509559452533722, | |
| "learning_rate": 5.4466666666666665e-06, | |
| "loss": 0.0004, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 25.18, | |
| "grad_norm": 0.024903174489736557, | |
| "learning_rate": 5.391111111111111e-06, | |
| "loss": 0.0003, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 25.43, | |
| "grad_norm": 0.02149089053273201, | |
| "learning_rate": 5.335555555555556e-06, | |
| "loss": 0.0003, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 25.67, | |
| "grad_norm": 0.014418188482522964, | |
| "learning_rate": 5.28e-06, | |
| "loss": 0.0002, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 25.92, | |
| "grad_norm": 0.018349435180425644, | |
| "learning_rate": 5.224444444444445e-06, | |
| "loss": 0.0002, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 26.16, | |
| "grad_norm": 0.013832501135766506, | |
| "learning_rate": 5.168888888888889e-06, | |
| "loss": 0.0002, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 26.41, | |
| "grad_norm": 0.012281795963644981, | |
| "learning_rate": 5.113333333333333e-06, | |
| "loss": 0.0001, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 26.65, | |
| "grad_norm": 0.009847081266343594, | |
| "learning_rate": 5.057777777777778e-06, | |
| "loss": 0.0001, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 26.89, | |
| "grad_norm": 0.020844602957367897, | |
| "learning_rate": 5.002222222222223e-06, | |
| "loss": 0.0002, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 27.14, | |
| "grad_norm": 0.012561388313770294, | |
| "learning_rate": 4.946666666666667e-06, | |
| "loss": 0.0003, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 27.38, | |
| "grad_norm": 0.0130954310297966, | |
| "learning_rate": 4.891111111111111e-06, | |
| "loss": 0.0002, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 27.63, | |
| "grad_norm": 0.008950438350439072, | |
| "learning_rate": 4.835555555555556e-06, | |
| "loss": 0.0001, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 27.87, | |
| "grad_norm": 0.007355996407568455, | |
| "learning_rate": 4.78e-06, | |
| "loss": 0.0001, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 28.12, | |
| "grad_norm": 0.008460123091936111, | |
| "learning_rate": 4.724444444444445e-06, | |
| "loss": 0.0001, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 28.36, | |
| "grad_norm": 0.007149124052375555, | |
| "learning_rate": 4.66888888888889e-06, | |
| "loss": 0.0001, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 28.61, | |
| "grad_norm": 0.008197379298508167, | |
| "learning_rate": 4.613333333333334e-06, | |
| "loss": 0.0001, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 28.85, | |
| "grad_norm": 0.00648567546159029, | |
| "learning_rate": 4.557777777777778e-06, | |
| "loss": 0.0001, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 29.1, | |
| "grad_norm": 0.006952579598873854, | |
| "learning_rate": 4.502222222222223e-06, | |
| "loss": 0.0001, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 29.34, | |
| "grad_norm": 0.009140390902757645, | |
| "learning_rate": 4.446666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 29.34, | |
| "eval_loss": 0.535641610622406, | |
| "eval_runtime": 1473.7946, | |
| "eval_samples_per_second": 1.964, | |
| "eval_steps_per_second": 0.491, | |
| "eval_wer": 18.395751304825566, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 29.58, | |
| "grad_norm": 0.007503976579755545, | |
| "learning_rate": 4.391111111111112e-06, | |
| "loss": 0.0001, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 29.83, | |
| "grad_norm": 0.005778305232524872, | |
| "learning_rate": 4.3355555555555565e-06, | |
| "loss": 0.0001, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 30.07, | |
| "grad_norm": 0.007032153662294149, | |
| "learning_rate": 4.2800000000000005e-06, | |
| "loss": 0.0001, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 30.32, | |
| "grad_norm": 0.005656179040670395, | |
| "learning_rate": 4.2244444444444446e-06, | |
| "loss": 0.0001, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 30.56, | |
| "grad_norm": 0.006409101653844118, | |
| "learning_rate": 4.168888888888889e-06, | |
| "loss": 0.0001, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 30.81, | |
| "grad_norm": 0.006126554682850838, | |
| "learning_rate": 4.1133333333333335e-06, | |
| "loss": 0.0001, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 31.05, | |
| "grad_norm": 0.005458911880850792, | |
| "learning_rate": 4.057777777777778e-06, | |
| "loss": 0.0001, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 31.3, | |
| "grad_norm": 0.005242755636572838, | |
| "learning_rate": 4.002222222222222e-06, | |
| "loss": 0.0001, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 31.54, | |
| "grad_norm": 0.005507026333361864, | |
| "learning_rate": 3.946666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 31.78, | |
| "grad_norm": 0.006995031144469976, | |
| "learning_rate": 3.891111111111111e-06, | |
| "loss": 0.0001, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "grad_norm": 0.0054997107945382595, | |
| "learning_rate": 3.835555555555555e-06, | |
| "loss": 0.0001, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 32.27, | |
| "grad_norm": 0.004706221166998148, | |
| "learning_rate": 3.7800000000000002e-06, | |
| "loss": 0.0001, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 32.52, | |
| "grad_norm": 0.005071236286312342, | |
| "learning_rate": 3.724444444444445e-06, | |
| "loss": 0.0001, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 32.76, | |
| "grad_norm": 0.005186586640775204, | |
| "learning_rate": 3.668888888888889e-06, | |
| "loss": 0.0001, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 33.01, | |
| "grad_norm": 0.006826834753155708, | |
| "learning_rate": 3.6133333333333336e-06, | |
| "loss": 0.0001, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 33.25, | |
| "grad_norm": 0.005360448732972145, | |
| "learning_rate": 3.5577777777777785e-06, | |
| "loss": 0.0001, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 33.5, | |
| "grad_norm": 0.004380129277706146, | |
| "learning_rate": 3.5022222222222225e-06, | |
| "loss": 0.0001, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 33.74, | |
| "grad_norm": 0.006299168337136507, | |
| "learning_rate": 3.446666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 33.99, | |
| "grad_norm": 0.006263605318963528, | |
| "learning_rate": 3.391111111111111e-06, | |
| "loss": 0.0001, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 34.23, | |
| "grad_norm": 0.004375001415610313, | |
| "learning_rate": 3.335555555555556e-06, | |
| "loss": 0.0001, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 34.47, | |
| "grad_norm": 0.004684335086494684, | |
| "learning_rate": 3.2800000000000004e-06, | |
| "loss": 0.0001, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 34.72, | |
| "grad_norm": 0.004894171841442585, | |
| "learning_rate": 3.2244444444444444e-06, | |
| "loss": 0.0001, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 34.96, | |
| "grad_norm": 0.005006886553019285, | |
| "learning_rate": 3.1688888888888893e-06, | |
| "loss": 0.0001, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 35.21, | |
| "grad_norm": 0.004793678876012564, | |
| "learning_rate": 3.1133333333333337e-06, | |
| "loss": 0.0001, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 35.45, | |
| "grad_norm": 0.0059304991737008095, | |
| "learning_rate": 3.0577777777777778e-06, | |
| "loss": 0.0001, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 35.7, | |
| "grad_norm": 0.00519231241196394, | |
| "learning_rate": 3.0022222222222227e-06, | |
| "loss": 0.0001, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 35.94, | |
| "grad_norm": 0.003882919903844595, | |
| "learning_rate": 2.946666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 36.19, | |
| "grad_norm": 0.004180469550192356, | |
| "learning_rate": 2.891111111111111e-06, | |
| "loss": 0.0001, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 36.43, | |
| "grad_norm": 0.004309108946472406, | |
| "learning_rate": 2.835555555555556e-06, | |
| "loss": 0.0001, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 36.67, | |
| "grad_norm": 0.004415574017912149, | |
| "learning_rate": 2.7800000000000005e-06, | |
| "loss": 0.0001, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 36.92, | |
| "grad_norm": 0.004281037952750921, | |
| "learning_rate": 2.7244444444444445e-06, | |
| "loss": 0.0001, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 37.16, | |
| "grad_norm": 0.00387546862475574, | |
| "learning_rate": 2.6688888888888894e-06, | |
| "loss": 0.0001, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 37.41, | |
| "grad_norm": 0.004122734069824219, | |
| "learning_rate": 2.6133333333333334e-06, | |
| "loss": 0.0001, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 37.65, | |
| "grad_norm": 0.004634737502783537, | |
| "learning_rate": 2.557777777777778e-06, | |
| "loss": 0.0001, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 37.9, | |
| "grad_norm": 0.003926947247236967, | |
| "learning_rate": 2.5022222222222224e-06, | |
| "loss": 0.0001, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 38.14, | |
| "grad_norm": 0.0039747897535562515, | |
| "learning_rate": 2.446666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 38.39, | |
| "grad_norm": 0.003933820873498917, | |
| "learning_rate": 2.3911111111111113e-06, | |
| "loss": 0.0001, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 38.63, | |
| "grad_norm": 0.0039012329652905464, | |
| "learning_rate": 2.3355555555555557e-06, | |
| "loss": 0.0001, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 38.88, | |
| "grad_norm": 0.0034920210018754005, | |
| "learning_rate": 2.28e-06, | |
| "loss": 0.0001, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 39.12, | |
| "grad_norm": 0.004213281441479921, | |
| "learning_rate": 2.2244444444444447e-06, | |
| "loss": 0.0001, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 39.12, | |
| "eval_loss": 0.5614951848983765, | |
| "eval_runtime": 1476.654, | |
| "eval_samples_per_second": 1.96, | |
| "eval_steps_per_second": 0.49, | |
| "eval_wer": 18.48273967585386, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 1000, | |
| "total_flos": 7.383284315947008e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |