| { | |
| "best_metric": 0.2764733135700226, | |
| "best_model_checkpoint": "/models/hfhub/DewiBrynJones/wav2vec2-xlsr-53-ft-ccv-en-cy/checkpoint-9000", | |
| "epoch": 4.524886877828054, | |
| "eval_steps": 500, | |
| "global_step": 9000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.6815159320831299, | |
| "learning_rate": 0.00014774999999999999, | |
| "loss": 5.9898, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.3093085289001465, | |
| "eval_runtime": 3176.8808, | |
| "eval_samples_per_second": 8.256, | |
| "eval_steps_per_second": 0.258, | |
| "eval_wer": 0.7970769457237188, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.3292043209075928, | |
| "learning_rate": 0.00029775, | |
| "loss": 1.0749, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.5815957188606262, | |
| "eval_runtime": 1123.0217, | |
| "eval_samples_per_second": 23.355, | |
| "eval_steps_per_second": 0.73, | |
| "eval_wer": 0.4617458414821357, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.9652466773986816, | |
| "learning_rate": 0.0002855853658536585, | |
| "loss": 0.4332, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.48338082432746887, | |
| "eval_runtime": 1101.5414, | |
| "eval_samples_per_second": 23.81, | |
| "eval_steps_per_second": 0.744, | |
| "eval_wer": 0.4091476878430383, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.57124924659729, | |
| "learning_rate": 0.0002709512195121951, | |
| "loss": 0.3655, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.1706229448318481, | |
| "learning_rate": 0.0002563170731707317, | |
| "loss": 0.3303, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 0.42033129930496216, | |
| "eval_runtime": 1101.1368, | |
| "eval_samples_per_second": 23.819, | |
| "eval_steps_per_second": 0.745, | |
| "eval_wer": 0.3419174394885707, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.8928599953651428, | |
| "learning_rate": 0.0002416829268292683, | |
| "loss": 0.276, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 0.3909631669521332, | |
| "eval_runtime": 1098.4606, | |
| "eval_samples_per_second": 23.877, | |
| "eval_steps_per_second": 0.746, | |
| "eval_wer": 0.3186423569490884, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.6678148508071899, | |
| "learning_rate": 0.00022704878048780485, | |
| "loss": 0.2591, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_loss": 0.39008986949920654, | |
| "eval_runtime": 1093.6554, | |
| "eval_samples_per_second": 23.982, | |
| "eval_steps_per_second": 0.75, | |
| "eval_wer": 0.3067188190019557, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.7449674606323242, | |
| "learning_rate": 0.00021241463414634144, | |
| "loss": 0.2501, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_loss": 0.3645510971546173, | |
| "eval_runtime": 1101.42, | |
| "eval_samples_per_second": 23.813, | |
| "eval_steps_per_second": 0.744, | |
| "eval_wer": 0.2895379891910079, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.994420051574707, | |
| "learning_rate": 0.00019778048780487803, | |
| "loss": 0.2332, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.632382333278656, | |
| "learning_rate": 0.00018314634146341462, | |
| "loss": 0.224, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "eval_loss": 0.35174447298049927, | |
| "eval_runtime": 1113.1837, | |
| "eval_samples_per_second": 23.561, | |
| "eval_steps_per_second": 0.737, | |
| "eval_wer": 0.2805501230206296, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.5861485600471497, | |
| "learning_rate": 0.00016851219512195123, | |
| "loss": 0.182, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_loss": 0.33475443720817566, | |
| "eval_runtime": 1111.5845, | |
| "eval_samples_per_second": 23.595, | |
| "eval_steps_per_second": 0.738, | |
| "eval_wer": 0.2655689441255021, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.585738480091095, | |
| "learning_rate": 0.0001538780487804878, | |
| "loss": 0.1777, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "eval_loss": 0.32769647240638733, | |
| "eval_runtime": 1109.5932, | |
| "eval_samples_per_second": 23.637, | |
| "eval_steps_per_second": 0.739, | |
| "eval_wer": 0.2611948772948079, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 0.6404664516448975, | |
| "learning_rate": 0.00013924390243902438, | |
| "loss": 0.1734, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_loss": 0.33233708143234253, | |
| "eval_runtime": 1114.1252, | |
| "eval_samples_per_second": 23.541, | |
| "eval_steps_per_second": 0.736, | |
| "eval_wer": 0.2643113999116775, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 1.567084550857544, | |
| "learning_rate": 0.00012460975609756097, | |
| "loss": 0.1704, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 1.35818612575531, | |
| "learning_rate": 0.00010997560975609755, | |
| "loss": 0.1629, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "eval_loss": 0.31713536381721497, | |
| "eval_runtime": 1084.7842, | |
| "eval_samples_per_second": 24.178, | |
| "eval_steps_per_second": 0.756, | |
| "eval_wer": 0.24851428931928585, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 1.0975894927978516, | |
| "learning_rate": 9.534146341463413e-05, | |
| "loss": 0.1338, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "eval_loss": 0.310283362865448, | |
| "eval_runtime": 1090.9879, | |
| "eval_samples_per_second": 24.041, | |
| "eval_steps_per_second": 0.752, | |
| "eval_wer": 0.23984186066073643, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 1.2747470140457153, | |
| "learning_rate": 8.070731707317072e-05, | |
| "loss": 0.1292, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "eval_loss": 0.2933865785598755, | |
| "eval_runtime": 1076.7354, | |
| "eval_samples_per_second": 24.359, | |
| "eval_steps_per_second": 0.762, | |
| "eval_wer": 0.22680798267196603, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.5606548190116882, | |
| "learning_rate": 6.607317073170731e-05, | |
| "loss": 0.1264, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "eval_loss": 0.29226595163345337, | |
| "eval_runtime": 1074.899, | |
| "eval_samples_per_second": 24.4, | |
| "eval_steps_per_second": 0.763, | |
| "eval_wer": 0.22483965259815364, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 1.5185168981552124, | |
| "learning_rate": 5.14390243902439e-05, | |
| "loss": 0.1241, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.7815582156181335, | |
| "learning_rate": 3.680487804878048e-05, | |
| "loss": 0.118, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "eval_loss": 0.28800907731056213, | |
| "eval_runtime": 1083.1518, | |
| "eval_samples_per_second": 24.215, | |
| "eval_steps_per_second": 0.757, | |
| "eval_wer": 0.21931739322440225, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 1.339690089225769, | |
| "learning_rate": 2.217073170731707e-05, | |
| "loss": 0.0996, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "eval_loss": 0.2792861759662628, | |
| "eval_runtime": 1078.7477, | |
| "eval_samples_per_second": 24.313, | |
| "eval_steps_per_second": 0.76, | |
| "eval_wer": 0.21242403213256786, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 1.0311238765716553, | |
| "learning_rate": 7.536585365853659e-06, | |
| "loss": 0.0969, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "eval_loss": 0.2764733135700226, | |
| "eval_runtime": 1074.1619, | |
| "eval_samples_per_second": 24.417, | |
| "eval_steps_per_second": 0.763, | |
| "eval_wer": 0.21145248459613483, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "step": 9000, | |
| "total_flos": 8.933861078537978e+19, | |
| "train_loss": 0.49683192168341744, | |
| "train_runtime": 52483.3074, | |
| "train_samples_per_second": 10.975, | |
| "train_steps_per_second": 0.171 | |
| } | |
| ], | |
| "logging_steps": 400, | |
| "max_steps": 9000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 8.933861078537978e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |