| { | |
| "best_global_step": 1900, | |
| "best_metric": 0.21757769584655762, | |
| "best_model_checkpoint": "/scratch/skscla001/experiments/datasets/results/mms-1b-all-bemgen-combined-fusion/checkpoint-1600", | |
| "epoch": 5.0, | |
| "eval_steps": 100, | |
| "global_step": 1970, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.25380710659898476, | |
| "grad_norm": 29.435340881347656, | |
| "learning_rate": 0.00029099999999999997, | |
| "loss": 7.7297, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25380710659898476, | |
| "eval_loss": 5.600469589233398, | |
| "eval_runtime": 68.6711, | |
| "eval_samples_per_second": 24.406, | |
| "eval_steps_per_second": 6.102, | |
| "eval_wer": 0.9995995194233079, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5076142131979695, | |
| "grad_norm": 12.57145881652832, | |
| "learning_rate": 0.0002845989304812834, | |
| "loss": 4.8876, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5076142131979695, | |
| "eval_loss": 5.189175605773926, | |
| "eval_runtime": 69.1166, | |
| "eval_samples_per_second": 24.249, | |
| "eval_steps_per_second": 6.062, | |
| "eval_wer": 1.1228140435188894, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7614213197969543, | |
| "grad_norm": 5.573763847351074, | |
| "learning_rate": 0.00026855614973262027, | |
| "loss": 4.4438, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7614213197969543, | |
| "eval_loss": 4.348100662231445, | |
| "eval_runtime": 68.2882, | |
| "eval_samples_per_second": 24.543, | |
| "eval_steps_per_second": 6.136, | |
| "eval_wer": 1.1233480176211454, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.015228426395939, | |
| "grad_norm": 2.898040294647217, | |
| "learning_rate": 0.00025267379679144383, | |
| "loss": 3.7194, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.015228426395939, | |
| "eval_loss": 3.3745944499969482, | |
| "eval_runtime": 68.4183, | |
| "eval_samples_per_second": 24.496, | |
| "eval_steps_per_second": 6.124, | |
| "eval_wer": 1.0120144173007608, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2690355329949239, | |
| "grad_norm": 2.0177152156829834, | |
| "learning_rate": 0.00023663101604278074, | |
| "loss": 3.3173, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2690355329949239, | |
| "eval_loss": 3.256007671356201, | |
| "eval_runtime": 68.7372, | |
| "eval_samples_per_second": 24.383, | |
| "eval_steps_per_second": 6.096, | |
| "eval_wer": 1.0060072086503804, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5228426395939088, | |
| "grad_norm": 3.2718145847320557, | |
| "learning_rate": 0.00022058823529411765, | |
| "loss": 3.1776, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5228426395939088, | |
| "eval_loss": 2.9593822956085205, | |
| "eval_runtime": 67.7629, | |
| "eval_samples_per_second": 24.733, | |
| "eval_steps_per_second": 6.183, | |
| "eval_wer": 1.0037378187157924, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.7766497461928934, | |
| "grad_norm": 1.4685205221176147, | |
| "learning_rate": 0.0002045454545454545, | |
| "loss": 1.1178, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7766497461928934, | |
| "eval_loss": 0.2583008110523224, | |
| "eval_runtime": 68.257, | |
| "eval_samples_per_second": 24.554, | |
| "eval_steps_per_second": 6.139, | |
| "eval_wer": 0.48344680283006275, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.030456852791878, | |
| "grad_norm": 0.8484176993370056, | |
| "learning_rate": 0.00018850267379679142, | |
| "loss": 0.4382, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.030456852791878, | |
| "eval_loss": 0.23566803336143494, | |
| "eval_runtime": 68.4145, | |
| "eval_samples_per_second": 24.498, | |
| "eval_steps_per_second": 6.124, | |
| "eval_wer": 0.42964891202776667, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.284263959390863, | |
| "grad_norm": 0.4717855155467987, | |
| "learning_rate": 0.0001724598930481283, | |
| "loss": 0.422, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.284263959390863, | |
| "eval_loss": 0.23215773701667786, | |
| "eval_runtime": 69.371, | |
| "eval_samples_per_second": 24.16, | |
| "eval_steps_per_second": 6.04, | |
| "eval_wer": 0.43038312641836873, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.5380710659898478, | |
| "grad_norm": 1.4339512586593628, | |
| "learning_rate": 0.00015641711229946522, | |
| "loss": 0.4101, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.5380710659898478, | |
| "eval_loss": 0.23107033967971802, | |
| "eval_runtime": 68.1286, | |
| "eval_samples_per_second": 24.601, | |
| "eval_steps_per_second": 6.15, | |
| "eval_wer": 0.4306501134694967, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.7918781725888326, | |
| "grad_norm": 0.7119982838630676, | |
| "learning_rate": 0.00014037433155080213, | |
| "loss": 0.3923, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.7918781725888326, | |
| "eval_loss": 0.23034636676311493, | |
| "eval_runtime": 67.7918, | |
| "eval_samples_per_second": 24.723, | |
| "eval_steps_per_second": 6.181, | |
| "eval_wer": 0.4159658256574556, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.045685279187817, | |
| "grad_norm": 0.410969614982605, | |
| "learning_rate": 0.00012433155080213902, | |
| "loss": 0.382, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.045685279187817, | |
| "eval_loss": 0.22485551238059998, | |
| "eval_runtime": 67.7668, | |
| "eval_samples_per_second": 24.732, | |
| "eval_steps_per_second": 6.183, | |
| "eval_wer": 0.40608730476571886, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.299492385786802, | |
| "grad_norm": 0.44749805331230164, | |
| "learning_rate": 0.00010828877005347593, | |
| "loss": 0.3799, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.299492385786802, | |
| "eval_loss": 0.22400638461112976, | |
| "eval_runtime": 69.3331, | |
| "eval_samples_per_second": 24.173, | |
| "eval_steps_per_second": 6.043, | |
| "eval_wer": 0.40174876518488856, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.553299492385787, | |
| "grad_norm": 0.48391538858413696, | |
| "learning_rate": 9.224598930481283e-05, | |
| "loss": 0.3733, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.553299492385787, | |
| "eval_loss": 0.2204192578792572, | |
| "eval_runtime": 68.018, | |
| "eval_samples_per_second": 24.641, | |
| "eval_steps_per_second": 6.16, | |
| "eval_wer": 0.3992123881991723, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.8071065989847717, | |
| "grad_norm": 0.7022409439086914, | |
| "learning_rate": 7.620320855614973e-05, | |
| "loss": 0.3757, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.8071065989847717, | |
| "eval_loss": 0.22058387100696564, | |
| "eval_runtime": 68.8893, | |
| "eval_samples_per_second": 24.329, | |
| "eval_steps_per_second": 6.082, | |
| "eval_wer": 0.40308370044052866, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.060913705583756, | |
| "grad_norm": 0.5409083366394043, | |
| "learning_rate": 6.016042780748663e-05, | |
| "loss": 0.3789, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.060913705583756, | |
| "eval_loss": 0.2191571146249771, | |
| "eval_runtime": 68.0976, | |
| "eval_samples_per_second": 24.612, | |
| "eval_steps_per_second": 6.153, | |
| "eval_wer": 0.4031504472033106, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.314720812182741, | |
| "grad_norm": 0.6483948230743408, | |
| "learning_rate": 4.4117647058823526e-05, | |
| "loss": 0.3635, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.314720812182741, | |
| "eval_loss": 0.21890130639076233, | |
| "eval_runtime": 68.9885, | |
| "eval_samples_per_second": 24.294, | |
| "eval_steps_per_second": 6.073, | |
| "eval_wer": 0.40515285008677077, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.568527918781726, | |
| "grad_norm": 0.45765629410743713, | |
| "learning_rate": 2.8074866310160424e-05, | |
| "loss": 0.3583, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.568527918781726, | |
| "eval_loss": 0.21793432533740997, | |
| "eval_runtime": 69.0917, | |
| "eval_samples_per_second": 24.258, | |
| "eval_steps_per_second": 6.064, | |
| "eval_wer": 0.3988786543852623, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.822335025380711, | |
| "grad_norm": 1.3359663486480713, | |
| "learning_rate": 1.2032085561497326e-05, | |
| "loss": 0.3733, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.822335025380711, | |
| "eval_loss": 0.21757769584655762, | |
| "eval_runtime": 68.6773, | |
| "eval_samples_per_second": 24.404, | |
| "eval_steps_per_second": 6.101, | |
| "eval_wer": 0.39680950473902016, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1970, | |
| "total_flos": 1.9392924694598373e+19, | |
| "train_loss": 1.6902474505042062, | |
| "train_runtime": 4165.473, | |
| "train_samples_per_second": 7.561, | |
| "train_steps_per_second": 0.473 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1970, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.9392924694598373e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |