learning_rate: 1e-05
train_batch_size: 4
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 8
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
0.4373	1.0	576	0.4561	29.4118
0.1991	2.0	1152	0.4733	18.8694
0.0774	3.0	1728	0.5345	30.7105
0.0318	4.0	2304	0.6535	20.1681
0.0187	5.0	2880	0.6712	19.4805
0.0093	6.0	3456	0.6984	17.4943
0.0072	7.0	4032	0.7575	17.7998
0.0053	8.0	4608	0.7468	27.8075
0.0062	9.0	5184	0.7547	19.7097
0.0035	10.0	5760	0.7824	18.1054
0.0044	11.0	6336	0.7867	17.1887
0.0019	12.0	6912	0.8744	20.0153
0.0083	13.0	7488	0.8182	18.7166
0.0028	14.0	8064	0.7884	19.4805
0.0034	15.0	8640	0.8374	18.7166
0.0015	16.0	9216	0.8528	20.6264
0.0016	17.0	9792	0.8610	20.7028
0.0014	18.0	10368	0.8203	17.7235
0.0012	19.0	10944	0.8641	18.6402
0.0003	20.0	11520	0.8803	18.7166
0.0001	21.0	12096	0.8996	17.9526
0.0	22.0	12672	0.9051	17.6471
0.0005	23.0	13248	0.9239	18.5638
0.0	24.0	13824	0.8996	18.0290
0.0007	25.0	14400	0.8951	16.8831
0.0012	26.0	14976	0.9201	18.6402
0.0015	27.0	15552	0.9008	18.0290
0.0	28.0	16128	0.9032	18.7166
0.0	29.0	16704	0.9202	18.9458
0.0	30.0	17280	0.9211	19.0985

thevan2404/whisper-medium-mini-media-long-ft-30epochs-2

whisper-medium-mini-media-long-ft-30epochs-2