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: 44
num_epochs: 3.0
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
0.8161	0.2993	44	0.8121	39.6197
0.4394	0.5986	88	0.5509	23.8187
0.4197	0.8980	132	0.5137	21.9032
0.3837	1.1973	176	0.4968	21.7768
0.3694	1.4966	220	0.4862	21.3694
0.3711	1.7959	264	0.4769	20.9760
0.3426	2.0952	308	0.4718	20.7793
0.3321	2.3946	352	0.4702	20.7840
0.3128	2.6939	396	0.4684	20.7559
0.3295	2.9932	440	0.4671	20.6201

greenw0lf/whisper-new-nnat-20h-random-seed42

whisper-new-nnat-20h-random-seed42