learning_rate: 1e-05
train_batch_size: 6
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 12
optimizer: Use adamw_torch 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: 500
num_epochs: 30
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer
0.4346	1.0	222	0.3659	36.2651
0.2031	2.0	444	0.2588	16.9853
0.0971	3.0	666	0.2747	14.7865
0.0386	4.0	888	0.3092	12.7788
0.0137	5.0	1110	0.3421	12.0459
0.0044	6.0	1332	0.3698	12.2690
0.0016	7.0	1554	0.3963	12.2371
0.0012	8.0	1776	0.4160	11.9184
0.0007	9.0	1998	0.4351	12.0140
0.0005	10.0	2220	0.4393	12.0459
0.0011	11.0	2442	0.4429	11.0580
0.0006	12.0	2664	0.4672	12.6832
0.0003	13.0	2886	0.4761	11.9503
0.0013	14.0	3108	0.4584	11.9503
0.0009	15.0	3330	0.4660	11.8866
0.0004	16.0	3552	0.4634	11.9822
0.0004	17.0	3774	0.4795	11.8228
0.0004	18.0	3996	0.4760	11.9822
0.0003	19.0	4218	0.4850	12.4602
0.0002	20.0	4440	0.4885	12.4283
0.0002	21.0	4662	0.5012	10.5800
0.0005	22.0	4884	0.4906	12.0778
0.0009	23.0	5106	0.4521	14.2447
0.0012	24.0	5328	0.4314	13.3843
0.0005	25.0	5550	0.4719	12.0778
0.0009	26.0	5772	0.4642	10.9624
0.0004	27.0	5994	0.4740	10.8987
0.0003	28.0	6216	0.4898	10.7075
0.0002	29.0	6438	0.4931	10.7393
0.0001	30.0	6660	0.4948	10.7393

thevan2404/whisper-medium-ft-30epochs-MC-long

whisper-medium-ft-30epochs-MC-long