| { | |
| "best_metric": 0.923943661971831, | |
| "best_model_checkpoint": "./ssw-finetune/checkpoint-1150", | |
| "epoch": 115.0, | |
| "eval_steps": 25, | |
| "global_step": 1150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.561491012573242, | |
| "learning_rate": 2.9999999999999997e-06, | |
| "loss": 7.7799, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.18166971206665, | |
| "learning_rate": 6.749999999999999e-06, | |
| "loss": 7.4713, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 6.806884288787842, | |
| "learning_rate": 1.05e-05, | |
| "loss": 7.696, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.3499999999999998e-05, | |
| "loss": 7.9462, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.6499999999999998e-05, | |
| "loss": 8.0165, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 7.467132091522217, | |
| "eval_runtime": 1.054, | |
| "eval_samples_per_second": 22.77, | |
| "eval_steps_per_second": 0.949, | |
| "eval_wer": 1.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.7979044914245605, | |
| "learning_rate": 2.025e-05, | |
| "loss": 6.5204, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 9.716986656188965, | |
| "learning_rate": 2.3999999999999997e-05, | |
| "loss": 7.6715, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.3519299030303955, | |
| "learning_rate": 2.7749999999999997e-05, | |
| "loss": 7.0161, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 5.6749138832092285, | |
| "learning_rate": 3.149999999999999e-05, | |
| "loss": 8.0617, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 8.150848388671875, | |
| "learning_rate": 3.5249999999999996e-05, | |
| "loss": 6.3142, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 6.626723766326904, | |
| "eval_runtime": 1.071, | |
| "eval_samples_per_second": 22.409, | |
| "eval_steps_per_second": 0.934, | |
| "eval_wer": 1.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 6.339476585388184, | |
| "learning_rate": 3.9e-05, | |
| "loss": 6.2643, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 13.012835502624512, | |
| "learning_rate": 4.2749999999999996e-05, | |
| "loss": 7.1655, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 11.24893569946289, | |
| "learning_rate": 4.65e-05, | |
| "loss": 5.8178, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 24.677473068237305, | |
| "learning_rate": 5.025e-05, | |
| "loss": 5.5684, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 8.214367866516113, | |
| "learning_rate": 5.399999999999999e-05, | |
| "loss": 4.3185, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "eval_loss": 3.72790789604187, | |
| "eval_runtime": 1.0391, | |
| "eval_samples_per_second": 23.097, | |
| "eval_steps_per_second": 0.962, | |
| "eval_wer": 1.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 8.643641471862793, | |
| "learning_rate": 5.7749999999999994e-05, | |
| "loss": 4.1807, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 10.54008674621582, | |
| "learning_rate": 6.149999999999999e-05, | |
| "loss": 3.7552, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.332289934158325, | |
| "learning_rate": 6.525e-05, | |
| "loss": 3.7053, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 4.925398349761963, | |
| "learning_rate": 6.9e-05, | |
| "loss": 3.3661, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 5.291933536529541, | |
| "learning_rate": 7.274999999999999e-05, | |
| "loss": 3.1777, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 3.050647735595703, | |
| "eval_runtime": 1.0273, | |
| "eval_samples_per_second": 23.362, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 1.6660319566726685, | |
| "learning_rate": 7.649999999999999e-05, | |
| "loss": 3.0435, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.681082546710968, | |
| "learning_rate": 8.025e-05, | |
| "loss": 3.254, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 2.713016986846924, | |
| "learning_rate": 8.4e-05, | |
| "loss": 2.972, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 7.211615085601807, | |
| "learning_rate": 8.774999999999999e-05, | |
| "loss": 3.1145, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 3.3372182846069336, | |
| "learning_rate": 9.149999999999999e-05, | |
| "loss": 3.0587, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "eval_loss": 2.936924695968628, | |
| "eval_runtime": 1.0424, | |
| "eval_samples_per_second": 23.023, | |
| "eval_steps_per_second": 0.959, | |
| "eval_wer": 1.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.895374059677124, | |
| "learning_rate": 9.525e-05, | |
| "loss": 2.9096, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 8.356375694274902, | |
| "learning_rate": 9.9e-05, | |
| "loss": 3.3159, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.6825320720672607, | |
| "learning_rate": 0.00010275, | |
| "loss": 2.9022, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.5, | |
| "grad_norm": 0.7314967513084412, | |
| "learning_rate": 0.00010649999999999999, | |
| "loss": 2.9058, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.183772563934326, | |
| "learning_rate": 0.00011024999999999998, | |
| "loss": 3.0633, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 2.9296257495880127, | |
| "eval_runtime": 1.0347, | |
| "eval_samples_per_second": 23.194, | |
| "eval_steps_per_second": 0.966, | |
| "eval_wer": 1.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.5, | |
| "grad_norm": 1.4891362190246582, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 2.9901, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 5.58284854888916, | |
| "learning_rate": 0.00011774999999999999, | |
| "loss": 2.9861, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "grad_norm": 1.3804000616073608, | |
| "learning_rate": 0.0001215, | |
| "loss": 2.9584, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.6562563180923462, | |
| "learning_rate": 0.00012524999999999998, | |
| "loss": 3.0194, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 0.653541088104248, | |
| "learning_rate": 0.000129, | |
| "loss": 2.9639, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "eval_loss": 2.926556348800659, | |
| "eval_runtime": 1.0442, | |
| "eval_samples_per_second": 22.985, | |
| "eval_steps_per_second": 0.958, | |
| "eval_wer": 1.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 9.456038475036621, | |
| "learning_rate": 0.00013275, | |
| "loss": 2.8944, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 18.5, | |
| "grad_norm": 0.3759576082229614, | |
| "learning_rate": 0.00013649999999999998, | |
| "loss": 2.9149, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 3.0567305088043213, | |
| "learning_rate": 0.00014025, | |
| "loss": 3.0321, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 19.5, | |
| "grad_norm": 8.436885833740234, | |
| "learning_rate": 0.00014399999999999998, | |
| "loss": 2.9683, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.9778860807418823, | |
| "learning_rate": 0.00014774999999999999, | |
| "loss": 2.9576, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 2.9644908905029297, | |
| "eval_runtime": 1.0268, | |
| "eval_samples_per_second": 23.374, | |
| "eval_steps_per_second": 0.974, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 20.5, | |
| "grad_norm": 0.6856608390808105, | |
| "learning_rate": 0.00014976923076923077, | |
| "loss": 2.9374, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 1.157402515411377, | |
| "learning_rate": 0.00014919230769230767, | |
| "loss": 2.875, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 21.5, | |
| "grad_norm": 0.42920613288879395, | |
| "learning_rate": 0.0001486153846153846, | |
| "loss": 2.9796, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 4.603660583496094, | |
| "learning_rate": 0.00014803846153846152, | |
| "loss": 2.9233, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "grad_norm": 1.3661619424819946, | |
| "learning_rate": 0.00014746153846153845, | |
| "loss": 2.8708, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "eval_loss": 2.9085776805877686, | |
| "eval_runtime": 1.0387, | |
| "eval_samples_per_second": 23.106, | |
| "eval_steps_per_second": 0.963, | |
| "eval_wer": 1.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.7445681691169739, | |
| "learning_rate": 0.00014688461538461537, | |
| "loss": 2.933, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 23.5, | |
| "grad_norm": 1.2040903568267822, | |
| "learning_rate": 0.0001463076923076923, | |
| "loss": 2.9217, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 4.538419246673584, | |
| "learning_rate": 0.00014573076923076923, | |
| "loss": 2.9043, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 24.5, | |
| "grad_norm": 0.36169031262397766, | |
| "learning_rate": 0.00014515384615384615, | |
| "loss": 2.8554, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.2133870124816895, | |
| "learning_rate": 0.00014457692307692305, | |
| "loss": 2.943, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 2.900446653366089, | |
| "eval_runtime": 1.0279, | |
| "eval_samples_per_second": 23.348, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 1.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 25.5, | |
| "grad_norm": 1.1455128192901611, | |
| "learning_rate": 0.00014399999999999998, | |
| "loss": 2.8775, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 3.7162177562713623, | |
| "learning_rate": 0.0001434230769230769, | |
| "loss": 2.9401, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 26.5, | |
| "grad_norm": 4.095553398132324, | |
| "learning_rate": 0.00014284615384615383, | |
| "loss": 2.9053, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 2.0302634239196777, | |
| "learning_rate": 0.00014226923076923075, | |
| "loss": 2.975, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "grad_norm": 3.123234510421753, | |
| "learning_rate": 0.00014169230769230768, | |
| "loss": 2.9225, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "eval_loss": 2.9469966888427734, | |
| "eval_runtime": 1.023, | |
| "eval_samples_per_second": 23.46, | |
| "eval_steps_per_second": 0.978, | |
| "eval_wer": 1.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.886202871799469, | |
| "learning_rate": 0.0001411153846153846, | |
| "loss": 2.8783, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 28.5, | |
| "grad_norm": 0.48980531096458435, | |
| "learning_rate": 0.00014053846153846153, | |
| "loss": 2.8977, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 2.4499869346618652, | |
| "learning_rate": 0.00013996153846153843, | |
| "loss": 2.9178, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 29.5, | |
| "grad_norm": 3.5155863761901855, | |
| "learning_rate": 0.00013938461538461536, | |
| "loss": 2.8955, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 3.8240697383880615, | |
| "learning_rate": 0.00013880769230769228, | |
| "loss": 2.9897, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_loss": 2.9530646800994873, | |
| "eval_runtime": 1.0334, | |
| "eval_samples_per_second": 23.224, | |
| "eval_steps_per_second": 0.968, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 30.5, | |
| "grad_norm": 1.4881560802459717, | |
| "learning_rate": 0.0001382307692307692, | |
| "loss": 2.8732, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 5.950206756591797, | |
| "learning_rate": 0.00013765384615384613, | |
| "loss": 2.9688, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 31.5, | |
| "grad_norm": 0.8825148940086365, | |
| "learning_rate": 0.00013707692307692306, | |
| "loss": 2.869, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 1.6368755102157593, | |
| "learning_rate": 0.00013649999999999998, | |
| "loss": 2.8843, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "grad_norm": 1.556404709815979, | |
| "learning_rate": 0.0001359230769230769, | |
| "loss": 2.8514, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "eval_loss": 2.911478042602539, | |
| "eval_runtime": 1.021, | |
| "eval_samples_per_second": 23.505, | |
| "eval_steps_per_second": 0.979, | |
| "eval_wer": 1.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.6802976131439209, | |
| "learning_rate": 0.00013534615384615384, | |
| "loss": 2.8542, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 33.5, | |
| "grad_norm": 0.5035978555679321, | |
| "learning_rate": 0.00013476923076923076, | |
| "loss": 2.9064, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 1.6443456411361694, | |
| "learning_rate": 0.0001341923076923077, | |
| "loss": 2.8498, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 34.5, | |
| "grad_norm": 0.6262179017066956, | |
| "learning_rate": 0.0001336153846153846, | |
| "loss": 2.8368, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.8266497850418091, | |
| "learning_rate": 0.00013303846153846154, | |
| "loss": 2.8681, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_loss": 2.9094789028167725, | |
| "eval_runtime": 1.0369, | |
| "eval_samples_per_second": 23.145, | |
| "eval_steps_per_second": 0.964, | |
| "eval_wer": 1.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 35.5, | |
| "grad_norm": 0.33677324652671814, | |
| "learning_rate": 0.00013246153846153846, | |
| "loss": 2.8163, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 0.6221341490745544, | |
| "learning_rate": 0.0001318846153846154, | |
| "loss": 2.8746, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 36.5, | |
| "grad_norm": 0.5015878677368164, | |
| "learning_rate": 0.00013130769230769232, | |
| "loss": 2.8477, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "grad_norm": 0.6005992889404297, | |
| "learning_rate": 0.00013073076923076921, | |
| "loss": 2.838, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "grad_norm": 0.4997330605983734, | |
| "learning_rate": 0.00013015384615384614, | |
| "loss": 2.8431, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "eval_loss": 2.90104603767395, | |
| "eval_runtime": 1.017, | |
| "eval_samples_per_second": 23.599, | |
| "eval_steps_per_second": 0.983, | |
| "eval_wer": 1.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 1.342210292816162, | |
| "learning_rate": 0.00012957692307692307, | |
| "loss": 2.8672, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 38.5, | |
| "grad_norm": 1.2935914993286133, | |
| "learning_rate": 0.000129, | |
| "loss": 2.848, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "grad_norm": 0.41487249732017517, | |
| "learning_rate": 0.00012842307692307692, | |
| "loss": 2.8244, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 39.5, | |
| "grad_norm": 1.1988450288772583, | |
| "learning_rate": 0.00012784615384615384, | |
| "loss": 2.8328, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 2.2671468257904053, | |
| "learning_rate": 0.00012726923076923077, | |
| "loss": 2.8843, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_loss": 2.9156665802001953, | |
| "eval_runtime": 1.0579, | |
| "eval_samples_per_second": 22.686, | |
| "eval_steps_per_second": 0.945, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 40.5, | |
| "grad_norm": 1.003772497177124, | |
| "learning_rate": 0.0001266923076923077, | |
| "loss": 2.8312, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "grad_norm": 1.2402571439743042, | |
| "learning_rate": 0.00012611538461538462, | |
| "loss": 2.8291, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 41.5, | |
| "grad_norm": 0.29388442635536194, | |
| "learning_rate": 0.00012553846153846152, | |
| "loss": 2.8275, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 0.9477460980415344, | |
| "learning_rate": 0.00012496153846153844, | |
| "loss": 2.8384, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 42.5, | |
| "grad_norm": 1.4519686698913574, | |
| "learning_rate": 0.00012438461538461537, | |
| "loss": 2.9357, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 42.5, | |
| "eval_loss": 2.902658462524414, | |
| "eval_runtime": 1.0363, | |
| "eval_samples_per_second": 23.158, | |
| "eval_steps_per_second": 0.965, | |
| "eval_wer": 1.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "grad_norm": 0.4391646087169647, | |
| "learning_rate": 0.0001238076923076923, | |
| "loss": 2.8395, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 43.5, | |
| "grad_norm": 2.1784377098083496, | |
| "learning_rate": 0.00012323076923076922, | |
| "loss": 2.8599, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 0.9729048609733582, | |
| "learning_rate": 0.00012265384615384615, | |
| "loss": 2.8489, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 44.5, | |
| "grad_norm": 0.5243009328842163, | |
| "learning_rate": 0.00012207692307692307, | |
| "loss": 2.83, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 0.7081323862075806, | |
| "learning_rate": 0.0001215, | |
| "loss": 2.8236, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "eval_loss": 2.901521682739258, | |
| "eval_runtime": 1.0318, | |
| "eval_samples_per_second": 23.261, | |
| "eval_steps_per_second": 0.969, | |
| "eval_wer": 1.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 45.5, | |
| "grad_norm": 0.3105088770389557, | |
| "learning_rate": 0.00012092307692307691, | |
| "loss": 2.8189, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "grad_norm": 0.6120209097862244, | |
| "learning_rate": 0.00012034615384615384, | |
| "loss": 2.8075, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 46.5, | |
| "grad_norm": 0.996507465839386, | |
| "learning_rate": 0.00011976923076923076, | |
| "loss": 2.8318, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "grad_norm": 7.280458927154541, | |
| "learning_rate": 0.00011919230769230767, | |
| "loss": 2.871, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 47.5, | |
| "grad_norm": 0.8332684636116028, | |
| "learning_rate": 0.0001186153846153846, | |
| "loss": 2.8376, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 47.5, | |
| "eval_loss": 2.900068998336792, | |
| "eval_runtime": 1.0322, | |
| "eval_samples_per_second": 23.251, | |
| "eval_steps_per_second": 0.969, | |
| "eval_wer": 1.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 0.6555355191230774, | |
| "learning_rate": 0.00011803846153846153, | |
| "loss": 2.7954, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 48.5, | |
| "grad_norm": 1.127866268157959, | |
| "learning_rate": 0.00011746153846153845, | |
| "loss": 2.8494, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "grad_norm": 0.7961714863777161, | |
| "learning_rate": 0.00011688461538461538, | |
| "loss": 2.8446, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 49.5, | |
| "grad_norm": 1.9832100868225098, | |
| "learning_rate": 0.00011630769230769229, | |
| "loss": 2.8353, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.9229313731193542, | |
| "learning_rate": 0.00011573076923076922, | |
| "loss": 2.8148, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 2.8878333568573, | |
| "eval_runtime": 1.0279, | |
| "eval_samples_per_second": 23.349, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 50.5, | |
| "grad_norm": 2.113555669784546, | |
| "learning_rate": 0.00011515384615384614, | |
| "loss": 2.816, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "grad_norm": 2.10042667388916, | |
| "learning_rate": 0.00011457692307692307, | |
| "loss": 2.8544, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 51.5, | |
| "grad_norm": 0.48272839188575745, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 2.8207, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "grad_norm": 0.9009172320365906, | |
| "learning_rate": 0.00011342307692307692, | |
| "loss": 2.8008, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 52.5, | |
| "grad_norm": 1.0341640710830688, | |
| "learning_rate": 0.00011284615384615384, | |
| "loss": 2.8057, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 52.5, | |
| "eval_loss": 2.8624706268310547, | |
| "eval_runtime": 1.037, | |
| "eval_samples_per_second": 23.144, | |
| "eval_steps_per_second": 0.964, | |
| "eval_wer": 1.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "grad_norm": 1.3395497798919678, | |
| "learning_rate": 0.00011226923076923077, | |
| "loss": 2.7866, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 53.5, | |
| "grad_norm": 0.3619355261325836, | |
| "learning_rate": 0.00011169230769230768, | |
| "loss": 2.7779, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "grad_norm": 1.4029289484024048, | |
| "learning_rate": 0.0001111153846153846, | |
| "loss": 2.789, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 54.5, | |
| "grad_norm": 0.29736635088920593, | |
| "learning_rate": 0.00011053846153846152, | |
| "loss": 2.7452, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "grad_norm": 1.7570823431015015, | |
| "learning_rate": 0.00010996153846153845, | |
| "loss": 2.7268, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "eval_loss": 2.819674253463745, | |
| "eval_runtime": 1.0343, | |
| "eval_samples_per_second": 23.205, | |
| "eval_steps_per_second": 0.967, | |
| "eval_wer": 1.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 55.5, | |
| "grad_norm": 0.3762887418270111, | |
| "learning_rate": 0.00010938461538461537, | |
| "loss": 2.7224, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "grad_norm": 1.0835281610488892, | |
| "learning_rate": 0.0001088076923076923, | |
| "loss": 2.7022, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 56.5, | |
| "grad_norm": 1.721433401107788, | |
| "learning_rate": 0.00010823076923076922, | |
| "loss": 2.6927, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 57.0, | |
| "grad_norm": 2.9872403144836426, | |
| "learning_rate": 0.00010765384615384615, | |
| "loss": 2.7924, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 57.5, | |
| "grad_norm": 0.5493649840354919, | |
| "learning_rate": 0.00010707692307692306, | |
| "loss": 2.6252, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 57.5, | |
| "eval_loss": 2.807591676712036, | |
| "eval_runtime": 1.0323, | |
| "eval_samples_per_second": 23.25, | |
| "eval_steps_per_second": 0.969, | |
| "eval_wer": 1.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "grad_norm": 1.2353851795196533, | |
| "learning_rate": 0.00010649999999999999, | |
| "loss": 2.6458, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 58.5, | |
| "grad_norm": 0.7240511775016785, | |
| "learning_rate": 0.00010592307692307691, | |
| "loss": 2.5911, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 59.0, | |
| "grad_norm": 0.9982340335845947, | |
| "learning_rate": 0.00010534615384615384, | |
| "loss": 2.6489, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 59.5, | |
| "grad_norm": 0.6784680485725403, | |
| "learning_rate": 0.00010476923076923076, | |
| "loss": 2.5169, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 1.9756778478622437, | |
| "learning_rate": 0.00010419230769230769, | |
| "loss": 2.5511, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_loss": 2.615316152572632, | |
| "eval_runtime": 1.0274, | |
| "eval_samples_per_second": 23.361, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 1.0056338028169014, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 60.5, | |
| "grad_norm": 1.3284317255020142, | |
| "learning_rate": 0.00010361538461538462, | |
| "loss": 2.4731, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "grad_norm": 1.3110464811325073, | |
| "learning_rate": 0.00010303846153846154, | |
| "loss": 2.4817, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 61.5, | |
| "grad_norm": 1.003812551498413, | |
| "learning_rate": 0.00010246153846153844, | |
| "loss": 2.3945, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "grad_norm": 1.148573398590088, | |
| "learning_rate": 0.00010188461538461537, | |
| "loss": 2.399, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "grad_norm": 0.5585479736328125, | |
| "learning_rate": 0.00010130769230769229, | |
| "loss": 2.323, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "eval_loss": 2.4444546699523926, | |
| "eval_runtime": 1.0272, | |
| "eval_samples_per_second": 23.365, | |
| "eval_steps_per_second": 0.974, | |
| "eval_wer": 1.0169014084507042, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 63.0, | |
| "grad_norm": 2.2142958641052246, | |
| "learning_rate": 0.00010073076923076922, | |
| "loss": 2.2927, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 63.5, | |
| "grad_norm": 1.0168890953063965, | |
| "learning_rate": 0.00010015384615384614, | |
| "loss": 2.2108, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "grad_norm": 1.312639832496643, | |
| "learning_rate": 9.957692307692307e-05, | |
| "loss": 2.1866, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 64.5, | |
| "grad_norm": 0.5699294209480286, | |
| "learning_rate": 9.9e-05, | |
| "loss": 2.1114, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "grad_norm": 1.4273818731307983, | |
| "learning_rate": 9.842307692307692e-05, | |
| "loss": 2.1119, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "eval_loss": 2.2476181983947754, | |
| "eval_runtime": 1.0519, | |
| "eval_samples_per_second": 22.815, | |
| "eval_steps_per_second": 0.951, | |
| "eval_wer": 1.1183098591549296, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 65.5, | |
| "grad_norm": 0.5214980244636536, | |
| "learning_rate": 9.784615384615383e-05, | |
| "loss": 2.0414, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "grad_norm": 2.480297803878784, | |
| "learning_rate": 9.726923076923076e-05, | |
| "loss": 2.0609, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 66.5, | |
| "grad_norm": 3.5270726680755615, | |
| "learning_rate": 9.669230769230768e-05, | |
| "loss": 1.9963, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 67.0, | |
| "grad_norm": 14.827882766723633, | |
| "learning_rate": 9.611538461538461e-05, | |
| "loss": 1.9333, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 67.5, | |
| "grad_norm": 1.1005451679229736, | |
| "learning_rate": 9.553846153846153e-05, | |
| "loss": 1.8514, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 67.5, | |
| "eval_loss": 2.173093318939209, | |
| "eval_runtime": 1.033, | |
| "eval_samples_per_second": 23.233, | |
| "eval_steps_per_second": 0.968, | |
| "eval_wer": 1.095774647887324, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "grad_norm": 1.5897767543792725, | |
| "learning_rate": 9.496153846153846e-05, | |
| "loss": 1.9986, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 68.5, | |
| "grad_norm": 0.8863438963890076, | |
| "learning_rate": 9.438461538461539e-05, | |
| "loss": 1.8067, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 69.0, | |
| "grad_norm": 1.305874228477478, | |
| "learning_rate": 9.380769230769231e-05, | |
| "loss": 1.7975, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 69.5, | |
| "grad_norm": 0.6541560292243958, | |
| "learning_rate": 9.323076923076921e-05, | |
| "loss": 1.7655, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 1.056104063987732, | |
| "learning_rate": 9.265384615384614e-05, | |
| "loss": 1.7094, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_loss": 2.0642001628875732, | |
| "eval_runtime": 1.0377, | |
| "eval_samples_per_second": 23.129, | |
| "eval_steps_per_second": 0.964, | |
| "eval_wer": 1.0309859154929577, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 70.5, | |
| "grad_norm": 0.5228053331375122, | |
| "learning_rate": 9.207692307692306e-05, | |
| "loss": 1.6764, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 71.0, | |
| "grad_norm": 6.9655256271362305, | |
| "learning_rate": 9.149999999999999e-05, | |
| "loss": 1.7414, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 71.5, | |
| "grad_norm": 0.6360809206962585, | |
| "learning_rate": 9.092307692307691e-05, | |
| "loss": 1.6232, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "grad_norm": 1.2141180038452148, | |
| "learning_rate": 9.034615384615384e-05, | |
| "loss": 1.6497, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 72.5, | |
| "grad_norm": 0.874902606010437, | |
| "learning_rate": 8.976923076923077e-05, | |
| "loss": 1.6069, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 72.5, | |
| "eval_loss": 2.0792412757873535, | |
| "eval_runtime": 1.0243, | |
| "eval_samples_per_second": 23.431, | |
| "eval_steps_per_second": 0.976, | |
| "eval_wer": 1.0788732394366196, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 73.0, | |
| "grad_norm": 0.9335172176361084, | |
| "learning_rate": 8.919230769230769e-05, | |
| "loss": 1.4947, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 73.5, | |
| "grad_norm": 1.299177885055542, | |
| "learning_rate": 8.861538461538462e-05, | |
| "loss": 1.5304, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "grad_norm": 1.6317135095596313, | |
| "learning_rate": 8.803846153846153e-05, | |
| "loss": 1.5218, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 74.5, | |
| "grad_norm": 0.8083561062812805, | |
| "learning_rate": 8.746153846153845e-05, | |
| "loss": 1.5259, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "grad_norm": 1.805677890777588, | |
| "learning_rate": 8.688461538461538e-05, | |
| "loss": 1.4663, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "eval_loss": 2.0323963165283203, | |
| "eval_runtime": 1.0407, | |
| "eval_samples_per_second": 23.062, | |
| "eval_steps_per_second": 0.961, | |
| "eval_wer": 1.036619718309859, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 75.5, | |
| "grad_norm": 0.8463692665100098, | |
| "learning_rate": 8.63076923076923e-05, | |
| "loss": 1.4244, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "grad_norm": 2.091686248779297, | |
| "learning_rate": 8.573076923076923e-05, | |
| "loss": 1.3791, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 76.5, | |
| "grad_norm": 0.7040625810623169, | |
| "learning_rate": 8.515384615384614e-05, | |
| "loss": 1.3495, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 77.0, | |
| "grad_norm": 1.7725024223327637, | |
| "learning_rate": 8.457692307692307e-05, | |
| "loss": 1.3497, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 77.5, | |
| "grad_norm": 0.808942437171936, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.288, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 77.5, | |
| "eval_loss": 2.0642640590667725, | |
| "eval_runtime": 1.0443, | |
| "eval_samples_per_second": 22.982, | |
| "eval_steps_per_second": 0.958, | |
| "eval_wer": 1.0929577464788733, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "grad_norm": 3.843997001647949, | |
| "learning_rate": 8.342307692307691e-05, | |
| "loss": 1.2597, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 78.5, | |
| "grad_norm": 0.9082187414169312, | |
| "learning_rate": 8.284615384615383e-05, | |
| "loss": 1.2702, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 79.0, | |
| "grad_norm": 1.4159339666366577, | |
| "learning_rate": 8.226923076923076e-05, | |
| "loss": 1.2833, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 79.5, | |
| "grad_norm": 1.0848701000213623, | |
| "learning_rate": 8.169230769230768e-05, | |
| "loss": 1.2117, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 2.275663137435913, | |
| "learning_rate": 8.111538461538461e-05, | |
| "loss": 1.262, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_loss": 2.084003210067749, | |
| "eval_runtime": 1.0408, | |
| "eval_samples_per_second": 23.059, | |
| "eval_steps_per_second": 0.961, | |
| "eval_wer": 1.076056338028169, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 80.5, | |
| "grad_norm": 0.9842613339424133, | |
| "learning_rate": 8.053846153846154e-05, | |
| "loss": 1.2799, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 81.0, | |
| "grad_norm": 20.336593627929688, | |
| "learning_rate": 7.996153846153846e-05, | |
| "loss": 1.2903, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 81.5, | |
| "grad_norm": 0.8291641473770142, | |
| "learning_rate": 7.938461538461539e-05, | |
| "loss": 1.1215, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "grad_norm": 1.6971830129623413, | |
| "learning_rate": 7.88076923076923e-05, | |
| "loss": 1.1435, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 82.5, | |
| "grad_norm": 0.69861900806427, | |
| "learning_rate": 7.823076923076923e-05, | |
| "loss": 1.043, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 82.5, | |
| "eval_loss": 2.149214506149292, | |
| "eval_runtime": 1.0296, | |
| "eval_samples_per_second": 23.311, | |
| "eval_steps_per_second": 0.971, | |
| "eval_wer": 1.0901408450704226, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 83.0, | |
| "grad_norm": 1.7208884954452515, | |
| "learning_rate": 7.776923076923076e-05, | |
| "loss": 1.203, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 83.5, | |
| "grad_norm": 0.8559800982475281, | |
| "learning_rate": 7.719230769230768e-05, | |
| "loss": 1.0825, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "grad_norm": 1.6605381965637207, | |
| "learning_rate": 7.661538461538461e-05, | |
| "loss": 1.1121, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 84.5, | |
| "grad_norm": 1.077573537826538, | |
| "learning_rate": 7.603846153846154e-05, | |
| "loss": 1.0145, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "grad_norm": 2.7091293334960938, | |
| "learning_rate": 7.546153846153846e-05, | |
| "loss": 1.0501, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "eval_loss": 2.177476644515991, | |
| "eval_runtime": 1.0552, | |
| "eval_samples_per_second": 22.744, | |
| "eval_steps_per_second": 0.948, | |
| "eval_wer": 1.0591549295774647, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 85.5, | |
| "grad_norm": 1.3562541007995605, | |
| "learning_rate": 7.488461538461539e-05, | |
| "loss": 1.1098, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 86.0, | |
| "grad_norm": 2.6526386737823486, | |
| "learning_rate": 7.43076923076923e-05, | |
| "loss": 0.8642, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 86.5, | |
| "grad_norm": 1.1710244417190552, | |
| "learning_rate": 7.373076923076922e-05, | |
| "loss": 0.9004, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 87.0, | |
| "grad_norm": 2.9008164405822754, | |
| "learning_rate": 7.315384615384615e-05, | |
| "loss": 1.037, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "grad_norm": 0.6306678056716919, | |
| "learning_rate": 7.257692307692308e-05, | |
| "loss": 0.9726, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "eval_loss": 2.176731586456299, | |
| "eval_runtime": 1.0313, | |
| "eval_samples_per_second": 23.271, | |
| "eval_steps_per_second": 0.97, | |
| "eval_wer": 1.028169014084507, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 88.0, | |
| "grad_norm": 1.6984366178512573, | |
| "learning_rate": 7.199999999999999e-05, | |
| "loss": 1.1201, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 88.5, | |
| "grad_norm": 0.803970992565155, | |
| "learning_rate": 7.142307692307691e-05, | |
| "loss": 0.908, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 89.0, | |
| "grad_norm": 2.103391408920288, | |
| "learning_rate": 7.084615384615384e-05, | |
| "loss": 0.8684, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 89.5, | |
| "grad_norm": 0.9575273990631104, | |
| "learning_rate": 7.026923076923077e-05, | |
| "loss": 0.9791, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 3.000880479812622, | |
| "learning_rate": 6.969230769230768e-05, | |
| "loss": 0.8079, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "eval_loss": 2.1965668201446533, | |
| "eval_runtime": 1.0433, | |
| "eval_samples_per_second": 23.003, | |
| "eval_steps_per_second": 0.958, | |
| "eval_wer": 0.9943661971830986, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 90.5, | |
| "grad_norm": 0.6576473712921143, | |
| "learning_rate": 6.91153846153846e-05, | |
| "loss": 0.846, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 91.0, | |
| "grad_norm": 2.2526416778564453, | |
| "learning_rate": 6.853846153846153e-05, | |
| "loss": 0.8868, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 91.5, | |
| "grad_norm": 0.5678216814994812, | |
| "learning_rate": 6.796153846153845e-05, | |
| "loss": 0.8925, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 92.0, | |
| "grad_norm": 2.549266815185547, | |
| "learning_rate": 6.738461538461538e-05, | |
| "loss": 1.0163, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 92.5, | |
| "grad_norm": 0.7736966013908386, | |
| "learning_rate": 6.68076923076923e-05, | |
| "loss": 0.7198, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 92.5, | |
| "eval_loss": 2.2433066368103027, | |
| "eval_runtime": 1.0523, | |
| "eval_samples_per_second": 22.808, | |
| "eval_steps_per_second": 0.95, | |
| "eval_wer": 1.0028169014084507, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 93.0, | |
| "grad_norm": 3.742175817489624, | |
| "learning_rate": 6.623076923076923e-05, | |
| "loss": 1.011, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 93.5, | |
| "grad_norm": 0.748150110244751, | |
| "learning_rate": 6.565384615384616e-05, | |
| "loss": 0.7659, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "grad_norm": 2.121845006942749, | |
| "learning_rate": 6.507692307692307e-05, | |
| "loss": 0.7862, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 94.5, | |
| "grad_norm": 0.7966519594192505, | |
| "learning_rate": 6.45e-05, | |
| "loss": 0.8271, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "grad_norm": 1.6206731796264648, | |
| "learning_rate": 6.392307692307692e-05, | |
| "loss": 0.6312, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "eval_loss": 2.309884786605835, | |
| "eval_runtime": 1.062, | |
| "eval_samples_per_second": 22.599, | |
| "eval_steps_per_second": 0.942, | |
| "eval_wer": 0.9971830985915493, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 95.5, | |
| "grad_norm": 3.240893602371216, | |
| "learning_rate": 6.334615384615385e-05, | |
| "loss": 0.723, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "grad_norm": 1.4926756620407104, | |
| "learning_rate": 6.276923076923076e-05, | |
| "loss": 0.7344, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 96.5, | |
| "grad_norm": 0.8542086482048035, | |
| "learning_rate": 6.219230769230769e-05, | |
| "loss": 0.7649, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 97.0, | |
| "grad_norm": 2.2014851570129395, | |
| "learning_rate": 6.161538461538461e-05, | |
| "loss": 0.6969, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 97.5, | |
| "grad_norm": 0.6612327694892883, | |
| "learning_rate": 6.103846153846154e-05, | |
| "loss": 0.6336, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 97.5, | |
| "eval_loss": 2.3546626567840576, | |
| "eval_runtime": 1.0484, | |
| "eval_samples_per_second": 22.893, | |
| "eval_steps_per_second": 0.954, | |
| "eval_wer": 0.9971830985915493, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 98.0, | |
| "grad_norm": 2.117011547088623, | |
| "learning_rate": 6.0461538461538456e-05, | |
| "loss": 0.7537, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 98.5, | |
| "grad_norm": 8.142460823059082, | |
| "learning_rate": 5.988461538461538e-05, | |
| "loss": 0.6593, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 99.0, | |
| "grad_norm": 2.6468851566314697, | |
| "learning_rate": 5.93076923076923e-05, | |
| "loss": 0.8069, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 99.5, | |
| "grad_norm": 1.392821192741394, | |
| "learning_rate": 5.8730769230769226e-05, | |
| "loss": 0.746, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 2.0805888175964355, | |
| "learning_rate": 5.8153846153846145e-05, | |
| "loss": 0.9073, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 2.350856304168701, | |
| "eval_runtime": 1.0707, | |
| "eval_samples_per_second": 22.414, | |
| "eval_steps_per_second": 0.934, | |
| "eval_wer": 0.9943661971830986, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 100.5, | |
| "grad_norm": 18.686534881591797, | |
| "learning_rate": 5.757692307692307e-05, | |
| "loss": 0.7907, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 101.0, | |
| "grad_norm": 1.7688676118850708, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 0.5693, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 101.5, | |
| "grad_norm": 0.9006216526031494, | |
| "learning_rate": 5.642307692307692e-05, | |
| "loss": 0.6408, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 102.0, | |
| "grad_norm": 2.382704496383667, | |
| "learning_rate": 5.584615384615384e-05, | |
| "loss": 0.7203, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 102.5, | |
| "grad_norm": 0.8852857351303101, | |
| "learning_rate": 5.526923076923076e-05, | |
| "loss": 0.6431, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 102.5, | |
| "eval_loss": 2.4202942848205566, | |
| "eval_runtime": 1.0529, | |
| "eval_samples_per_second": 22.794, | |
| "eval_steps_per_second": 0.95, | |
| "eval_wer": 1.0056338028169014, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 103.0, | |
| "grad_norm": 3.3610403537750244, | |
| "learning_rate": 5.4692307692307686e-05, | |
| "loss": 0.6476, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 103.5, | |
| "grad_norm": 0.8738270401954651, | |
| "learning_rate": 5.411538461538461e-05, | |
| "loss": 0.5492, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 104.0, | |
| "grad_norm": 2.4251339435577393, | |
| "learning_rate": 5.353846153846153e-05, | |
| "loss": 0.6005, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 104.5, | |
| "grad_norm": 0.7935536503791809, | |
| "learning_rate": 5.2961538461538456e-05, | |
| "loss": 0.5855, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 105.0, | |
| "grad_norm": 2.805385112762451, | |
| "learning_rate": 5.238461538461538e-05, | |
| "loss": 0.62, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 105.0, | |
| "eval_loss": 2.3933348655700684, | |
| "eval_runtime": 1.0674, | |
| "eval_samples_per_second": 22.485, | |
| "eval_steps_per_second": 0.937, | |
| "eval_wer": 0.9746478873239437, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 105.5, | |
| "grad_norm": 1.2249245643615723, | |
| "learning_rate": 5.180769230769231e-05, | |
| "loss": 0.652, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 106.0, | |
| "grad_norm": 1.2247533798217773, | |
| "learning_rate": 5.123076923076922e-05, | |
| "loss": 0.6108, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 106.5, | |
| "grad_norm": 0.8812918663024902, | |
| "learning_rate": 5.0653846153846146e-05, | |
| "loss": 0.6453, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 107.0, | |
| "grad_norm": 2.7638535499572754, | |
| "learning_rate": 5.007692307692307e-05, | |
| "loss": 0.568, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 107.5, | |
| "grad_norm": 1.3182368278503418, | |
| "learning_rate": 4.95e-05, | |
| "loss": 0.708, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 107.5, | |
| "eval_loss": 2.4381346702575684, | |
| "eval_runtime": 1.061, | |
| "eval_samples_per_second": 22.619, | |
| "eval_steps_per_second": 0.942, | |
| "eval_wer": 0.9690140845070423, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 108.0, | |
| "grad_norm": 2.4760406017303467, | |
| "learning_rate": 4.8923076923076916e-05, | |
| "loss": 0.6171, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 108.5, | |
| "grad_norm": 0.5409008264541626, | |
| "learning_rate": 4.834615384615384e-05, | |
| "loss": 0.5542, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 109.0, | |
| "grad_norm": 1.675410509109497, | |
| "learning_rate": 4.776923076923077e-05, | |
| "loss": 0.6491, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 109.5, | |
| "grad_norm": 0.8941754698753357, | |
| "learning_rate": 4.719230769230769e-05, | |
| "loss": 0.7266, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 110.0, | |
| "grad_norm": 1.9851211309432983, | |
| "learning_rate": 4.6615384615384605e-05, | |
| "loss": 0.6729, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 110.0, | |
| "eval_loss": 2.474308967590332, | |
| "eval_runtime": 1.0636, | |
| "eval_samples_per_second": 22.564, | |
| "eval_steps_per_second": 0.94, | |
| "eval_wer": 1.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 110.5, | |
| "grad_norm": 0.677306592464447, | |
| "learning_rate": 4.603846153846153e-05, | |
| "loss": 0.7625, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 111.0, | |
| "grad_norm": 2.572356700897217, | |
| "learning_rate": 4.546153846153846e-05, | |
| "loss": 0.5146, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 111.5, | |
| "grad_norm": 1.2789101600646973, | |
| "learning_rate": 4.488461538461538e-05, | |
| "loss": 0.5504, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 112.0, | |
| "grad_norm": 2.3920390605926514, | |
| "learning_rate": 4.430769230769231e-05, | |
| "loss": 0.4821, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 112.5, | |
| "grad_norm": 1.219436764717102, | |
| "learning_rate": 4.373076923076923e-05, | |
| "loss": 0.5779, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 112.5, | |
| "eval_loss": 2.492933988571167, | |
| "eval_runtime": 1.0274, | |
| "eval_samples_per_second": 23.36, | |
| "eval_steps_per_second": 0.973, | |
| "eval_wer": 0.9549295774647887, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 113.0, | |
| "grad_norm": 3.558155059814453, | |
| "learning_rate": 4.315384615384615e-05, | |
| "loss": 0.4743, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 113.5, | |
| "grad_norm": 0.9398171901702881, | |
| "learning_rate": 4.257692307692307e-05, | |
| "loss": 0.493, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 114.0, | |
| "grad_norm": 4.514529705047607, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.4341, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 114.5, | |
| "grad_norm": 1.015120029449463, | |
| "learning_rate": 4.142307692307692e-05, | |
| "loss": 0.5069, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 115.0, | |
| "grad_norm": 2.043063163757324, | |
| "learning_rate": 4.084615384615384e-05, | |
| "loss": 0.6303, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 115.0, | |
| "eval_loss": 2.5056331157684326, | |
| "eval_runtime": 1.0408, | |
| "eval_samples_per_second": 23.06, | |
| "eval_steps_per_second": 0.961, | |
| "eval_wer": 0.923943661971831, | |
| "step": 1150 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 150, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.771505223996499e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |