learning_rate: 0.0001
train_batch_size: 16
eval_batch_size: 8
seed: 42
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: 1000
training_steps: 20000
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
1.4911	0.3047	500	1.4207
0.6276	0.6094	1000	0.6385
0.6078	0.9141	1500	0.5878
0.5866	1.2188	2000	0.5650
0.5387	1.5235	2500	0.5517
0.5724	1.8282	3000	0.5410
0.5471	2.1328	3500	0.5330
0.5038	2.4375	4000	0.5261
0.5224	2.7422	4500	0.5212
0.5406	3.0469	5000	0.5168
0.5113	3.3516	5500	0.5139
0.5041	3.6563	6000	0.5111
0.5208	3.9610	6500	0.5080
0.5075	4.2657	7000	0.5068
0.4422	4.5704	7500	0.5055
0.4999	4.8751	8000	0.5025
0.4665	5.1798	8500	0.5028
0.4713	5.4845	9000	0.5016
0.4626	5.7892	9500	0.5004
0.4744	6.0938	10000	0.4988
0.4589	6.3985	10500	0.4989
0.474	6.7032	11000	0.4983
0.4609	7.0079	11500	0.4971
0.4686	7.3126	12000	0.4973
0.4432	7.6173	12500	0.4966
0.4311	7.9220	13000	0.4967
0.4566	8.2267	13500	0.4968
0.4517	8.5314	14000	0.4970
0.4735	8.8361	14500	0.4960
0.4209	9.1408	15000	0.4963
0.4729	9.4455	15500	0.4959
0.452	9.7502	16000	0.4952
0.4845	10.0548	16500	0.4960
0.4335	10.3595	17000	0.4965
0.4014	10.6642	17500	0.4960
0.4874	10.9689	18000	0.4958
0.4057	11.2736	18500	0.4961

HouraMor/whisper-large-adalora-r8-lr1e4-st20k

whisper-large-adalora-r8-lr1e4-st20k