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: 50
num_epochs: 3.0
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
1.0713	0.2994	50	0.7510	28.8063
0.5554	0.5988	100	0.5267	22.8446
0.5012	0.8982	150	0.4931	21.7019
0.4774	1.1976	200	0.4769	20.6856
0.4678	1.4970	250	0.4682	20.4327
0.4621	1.7964	300	0.4607	20.1845
0.4558	2.0958	350	0.4570	20.4749
0.4012	2.3952	400	0.4545	20.1236
0.4248	2.6946	450	0.4524	19.9363
0.4217	2.9940	500	0.4516	19.8473

greenw0lf/whisper-new-non-natives

whisper-new-non-natives