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: 12
num_epochs: 3.0
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
1.0785	0.3	12	1.3913	46.9442
0.8148	0.6	24	1.0996	47.7404
0.7067	0.9	36	0.8460	27.9258
0.5998	1.2	48	0.7309	28.4035
0.5892	1.5	60	0.6516	25.6451
0.503	1.8	72	0.6187	26.2024
0.5696	2.1	84	0.6043	22.9897
0.5179	2.4	96	0.5966	22.7275
0.5116	2.7	108	0.5921	22.7462
0.5041	3.0	120	0.5906	22.7462

greenw0lf/whisper-new-nnat-10h-maxcos-xlsr

whisper-new-nnat-10h-maxcos-xlsr