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: 39
num_epochs: 3.0
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
1.4236	0.3	39	0.8751	50.5409
0.6855	0.6	78	0.5695	23.9592
0.5081	0.9	117	0.5313	22.8399
0.4893	1.2	156	0.5168	24.1559
0.4438	1.5	195	0.5044	23.7156
0.4488	1.8	234	0.4984	22.0765
0.4311	2.1	273	0.4919	21.8143
0.3966	2.4	312	0.4908	21.7908
0.4207	2.7	351	0.4884	21.7347
0.3881	3.0	390	0.4877	21.6035

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

whisper-new-nnat-10h-caps-xlsr