learning_rate: 0.0001
train_batch_size: 48
eval_batch_size: 32
seed: 42
optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 9
num_epochs: 3.0
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
1.1871	0.3	9	1.4259	47.6046
1.0885	0.6	18	1.2025	47.4313
0.8458	0.9	27	0.9784	44.6588
0.6932	1.2	36	0.8203	34.5619
0.594	1.5	45	0.7399	32.3233
0.5737	1.8	54	0.6830	29.3917
0.5994	2.1	63	0.6463	28.3426
0.5363	2.4	72	0.6283	29.1013
0.5238	2.7	81	0.6203	27.5980
0.5176	3.0	90	0.6178	26.1275

greenw0lf/whisper-new-nnat-5h-mmr-xlsr

whisper-new-nnat-5h-mmr-xlsr