learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 50
num_epochs: 2
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.1636	0.2213	500	0.3490	0.4499	0.1493
0.1416	0.4427	1000	0.3499	0.4498	0.1495
0.1492	0.6640	1500	0.3489	0.4538	0.1505
0.1385	0.8853	2000	0.3522	0.4515	0.149
0.1281	1.1067	2500	0.3531	0.4486	0.1496
0.1525	1.3280	3000	0.3575	0.4502	0.1495
0.1312	1.5494	3500	0.3591	0.4546	0.1515
0.1374	1.7707	4000	0.3587	0.4562	0.1525
0.1292	1.9920	4500	0.3590	0.456	0.1536

Tsedee/whisper-large-v2-mn-commercial-v2

whisper-large-v2-mn-commercial-v2