whispertiny_buah25sil_v1_5

This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

Training Loss	Epoch	Step	Validation Loss	Wer
0.017	0.0333	5000	0.0843	6.0049
0.0118	0.0667	10000	0.0669	4.7978
0.0076	1.0236	15000	0.0709	4.2980
0.0063	1.0569	20000	0.1040	5.4744
0.0054	2.0139	25000	0.0818	4.7209
0.0043	2.0472	30000	0.1085	5.1899
0.0034	3.0042	35000	0.1015	4.4364
0.0033	3.0375	40000	0.1228	4.9900
0.0018	0.0333	45000	0.1195	4.6825
0.0025	0.0667	50000	0.1266	4.9516
0.0018	1.0236	55000	0.1063	4.2288
0.0017	1.0569	60000	0.1253	4.8209
0.0014	2.0139	65000	0.1267	4.4749
0.0013	2.0472	70000	0.1549	5.2437
0.0007	3.0042	75000	0.1242	4.2673
0.0008	3.0375	80000	0.1937	5.4129
0.0008	3.0708	85000	0.1457	4.5979
0.0003	4.0278	90000	0.1606	4.7132
0.0006	4.0611	95000	0.1753	5.0361
0.0003	5.0180	100000	0.1772	4.8901
0.0007	5.0514	105000	0.1768	4.6594
0.0001	6.0083	110000	0.1813	4.6363
0.0001	6.0416	115000	0.1720	4.3903
0.0001	6.0750	120000	0.1702	4.4441
0.0	7.0319	125000	0.1703	4.4133
0.0001	7.0652	130000	0.1706	4.3288
0.0	8.0222	135000	0.1779	4.4595
0.0	8.0555	140000	0.1752	4.3211
0.0	9.0125	145000	0.1818	4.3903
0.0	9.0458	150000	0.1780	4.3211