| { | |
| "best_metric": 0.21356959640979767, | |
| "best_model_checkpoint": "/scratch/skscla001/results/w2v-bert-bem-bl/checkpoint-5500", | |
| "epoch": 9.99297259311314, | |
| "eval_steps": 500, | |
| "global_step": 7110, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.35137034434293746, | |
| "grad_norm": 6.841489791870117, | |
| "learning_rate": 0.00014879999999999998, | |
| "loss": 1.1986, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7027406886858749, | |
| "grad_norm": 5.7157793045043945, | |
| "learning_rate": 0.0002988, | |
| "loss": 0.5344, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7027406886858749, | |
| "eval_loss": 0.5448043942451477, | |
| "eval_runtime": 58.822, | |
| "eval_samples_per_second": 25.365, | |
| "eval_steps_per_second": 3.179, | |
| "eval_wer": 0.7378665974565274, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0541110330288124, | |
| "grad_norm": 1.5800367593765259, | |
| "learning_rate": 0.00028874432677760965, | |
| "loss": 0.4918, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4054813773717498, | |
| "grad_norm": 1.6567301750183105, | |
| "learning_rate": 0.0002774432677760968, | |
| "loss": 0.4242, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4054813773717498, | |
| "eval_loss": 0.3054925501346588, | |
| "eval_runtime": 56.7685, | |
| "eval_samples_per_second": 26.282, | |
| "eval_steps_per_second": 3.294, | |
| "eval_wer": 0.6024742624794532, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7568517217146873, | |
| "grad_norm": 1.094335675239563, | |
| "learning_rate": 0.0002660968229954614, | |
| "loss": 0.3958, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.1082220660576247, | |
| "grad_norm": 1.0848636627197266, | |
| "learning_rate": 0.000254750378214826, | |
| "loss": 0.3603, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1082220660576247, | |
| "eval_loss": 0.26925167441368103, | |
| "eval_runtime": 56.9762, | |
| "eval_samples_per_second": 26.186, | |
| "eval_steps_per_second": 3.282, | |
| "eval_wer": 0.5385413963145601, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.459592410400562, | |
| "grad_norm": 0.9221932888031006, | |
| "learning_rate": 0.00024340393343419062, | |
| "loss": 0.3228, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.8109627547434997, | |
| "grad_norm": 1.12798273563385, | |
| "learning_rate": 0.0002320574886535552, | |
| "loss": 0.3144, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.8109627547434997, | |
| "eval_loss": 0.2683100700378418, | |
| "eval_runtime": 56.9393, | |
| "eval_samples_per_second": 26.203, | |
| "eval_steps_per_second": 3.284, | |
| "eval_wer": 0.5529025002162817, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.162333099086437, | |
| "grad_norm": 0.6111273169517517, | |
| "learning_rate": 0.0002207110438729198, | |
| "loss": 0.2931, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.5137034434293746, | |
| "grad_norm": 0.5683824419975281, | |
| "learning_rate": 0.0002093645990922844, | |
| "loss": 0.2656, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.5137034434293746, | |
| "eval_loss": 0.24722729623317719, | |
| "eval_runtime": 57.2449, | |
| "eval_samples_per_second": 26.063, | |
| "eval_steps_per_second": 3.267, | |
| "eval_wer": 0.5258240332208669, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.865073787772312, | |
| "grad_norm": 0.7263462543487549, | |
| "learning_rate": 0.000198018154311649, | |
| "loss": 0.2574, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.2164441321152495, | |
| "grad_norm": 0.422783762216568, | |
| "learning_rate": 0.0001866717095310136, | |
| "loss": 0.2311, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.2164441321152495, | |
| "eval_loss": 0.2352277785539627, | |
| "eval_runtime": 56.81, | |
| "eval_samples_per_second": 26.263, | |
| "eval_steps_per_second": 3.292, | |
| "eval_wer": 0.5025521238861493, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.567814476458187, | |
| "grad_norm": 0.5190473794937134, | |
| "learning_rate": 0.00017532526475037822, | |
| "loss": 0.2133, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 4.919184820801124, | |
| "grad_norm": 0.7750300765037537, | |
| "learning_rate": 0.0001639788199697428, | |
| "loss": 0.2106, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.919184820801124, | |
| "eval_loss": 0.2326769232749939, | |
| "eval_runtime": 56.8929, | |
| "eval_samples_per_second": 26.225, | |
| "eval_steps_per_second": 3.287, | |
| "eval_wer": 0.5003027943593736, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.270555165144062, | |
| "grad_norm": 0.37189096212387085, | |
| "learning_rate": 0.0001526323751891074, | |
| "loss": 0.1835, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 5.621925509486999, | |
| "grad_norm": 0.5811314582824707, | |
| "learning_rate": 0.000141285930408472, | |
| "loss": 0.1816, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.621925509486999, | |
| "eval_loss": 0.22979679703712463, | |
| "eval_runtime": 56.9781, | |
| "eval_samples_per_second": 26.185, | |
| "eval_steps_per_second": 3.282, | |
| "eval_wer": 0.49865905355134527, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.973295853829937, | |
| "grad_norm": 1.2501654624938965, | |
| "learning_rate": 0.0001299394856278366, | |
| "loss": 0.1804, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 6.324666198172874, | |
| "grad_norm": 0.40958958864212036, | |
| "learning_rate": 0.0001185930408472012, | |
| "loss": 0.1432, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.324666198172874, | |
| "eval_loss": 0.21775686740875244, | |
| "eval_runtime": 57.4498, | |
| "eval_samples_per_second": 25.97, | |
| "eval_steps_per_second": 3.255, | |
| "eval_wer": 0.46855264296219395, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.676036542515812, | |
| "grad_norm": 0.5354466438293457, | |
| "learning_rate": 0.0001072465960665658, | |
| "loss": 0.1475, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 7.027406886858749, | |
| "grad_norm": 0.44123438000679016, | |
| "learning_rate": 9.59001512859304e-05, | |
| "loss": 0.1431, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.027406886858749, | |
| "eval_loss": 0.21715782582759857, | |
| "eval_runtime": 57.0515, | |
| "eval_samples_per_second": 26.152, | |
| "eval_steps_per_second": 3.278, | |
| "eval_wer": 0.47469504282377367, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.378777231201687, | |
| "grad_norm": 0.29582667350769043, | |
| "learning_rate": 8.4553706505295e-05, | |
| "loss": 0.1111, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 7.730147575544624, | |
| "grad_norm": 0.5549105405807495, | |
| "learning_rate": 7.32072617246596e-05, | |
| "loss": 0.1069, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 7.730147575544624, | |
| "eval_loss": 0.21356959640979767, | |
| "eval_runtime": 56.8619, | |
| "eval_samples_per_second": 26.239, | |
| "eval_steps_per_second": 3.289, | |
| "eval_wer": 0.45393200103815207, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 8.081517919887562, | |
| "grad_norm": 0.632247269153595, | |
| "learning_rate": 6.18608169440242e-05, | |
| "loss": 0.1045, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 8.432888264230499, | |
| "grad_norm": 0.3406793475151062, | |
| "learning_rate": 5.05143721633888e-05, | |
| "loss": 0.0767, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 8.432888264230499, | |
| "eval_loss": 0.22697694599628448, | |
| "eval_runtime": 56.8774, | |
| "eval_samples_per_second": 26.232, | |
| "eval_steps_per_second": 3.288, | |
| "eval_wer": 0.4403495112033913, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 8.784258608573436, | |
| "grad_norm": 0.4026772379875183, | |
| "learning_rate": 3.91679273827534e-05, | |
| "loss": 0.0781, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 9.135628952916374, | |
| "grad_norm": 0.4836815595626831, | |
| "learning_rate": 2.7821482602118e-05, | |
| "loss": 0.0667, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 9.135628952916374, | |
| "eval_loss": 0.23749949038028717, | |
| "eval_runtime": 56.671, | |
| "eval_samples_per_second": 26.327, | |
| "eval_steps_per_second": 3.3, | |
| "eval_wer": 0.4385327450471494, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 9.486999297259311, | |
| "grad_norm": 0.17178182303905487, | |
| "learning_rate": 1.64750378214826e-05, | |
| "loss": 0.0502, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 9.838369641602249, | |
| "grad_norm": 0.3071761131286621, | |
| "learning_rate": 5.128593040847201e-06, | |
| "loss": 0.0468, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 9.838369641602249, | |
| "eval_loss": 0.2403486669063568, | |
| "eval_runtime": 56.6826, | |
| "eval_samples_per_second": 26.322, | |
| "eval_steps_per_second": 3.299, | |
| "eval_wer": 0.4353317761051994, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 9.99297259311314, | |
| "step": 7110, | |
| "total_flos": 2.094794810533582e+19, | |
| "train_loss": 0.25149581422282674, | |
| "train_runtime": 9293.6781, | |
| "train_samples_per_second": 12.242, | |
| "train_steps_per_second": 0.765 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 7110, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.094794810533582e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |