whisper-synthesized-turkish-8-hour

This model is a fine-tuned version of openai/whisper-small on the None 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
1.2682	0.52	100	0.5845	99.7901
0.4591	1.04	200	0.3895	21.4541
0.2482	1.56	300	0.2241	12.2145
0.1554	2.08	400	0.2092	11.7825
0.096	2.6	500	0.2035	13.9057
0.0765	3.12	600	0.2052	11.2517
0.0424	3.65	700	0.2024	13.4490
0.0403	4.17	800	0.2094	12.0849
0.0216	4.69	900	0.2049	13.1959
0.0201	5.21	1000	0.2079	12.1034
0.0101	5.73	1100	0.2073	12.5663
0.0131	6.25	1200	0.2093	16.7757
0.0088	6.77	1300	0.2121	16.5165
0.0073	7.29	1400	0.2142	15.3314
0.0036	7.81	1500	0.2183	13.7020
0.0047	8.33	1600	0.2159	16.1647
0.0038	8.85	1700	0.2166	13.7514
0.0027	9.38	1800	0.2172	19.9975
0.0028	9.9	1900	0.2183	18.2385
0.0015	10.42	2000	0.2196	17.4238
0.0023	10.94	2100	0.2192	14.7019
0.0012	11.46	2200	0.2216	15.9919
0.0017	11.98	2300	0.2215	19.6334
0.001	12.5	2400	0.2219	20.5160
0.0014	13.02	2500	0.2236	21.7813
0.0011	13.54	2600	0.2242	23.0897
0.0009	14.06	2700	0.2276	25.0401
0.001	14.58	2800	0.2269	18.7014
0.001	15.1	2900	0.2265	20.8554
0.0008	15.62	3000	0.2272	19.7013
0.0009	16.15	3100	0.2277	26.5831
0.0007	16.67	3200	0.2290	24.3427
0.0008	17.19	3300	0.2285	20.7011
0.0007	17.71	3400	0.2288	21.8738
0.0007	18.23	3500	0.2290	20.7258
0.0006	18.75	3600	0.2295	21.1641
0.0006	19.27	3700	0.2297	23.7625
0.0007	19.79	3800	0.2301	24.4044
0.0006	20.31	3900	0.2299	22.9786
0.0006	20.83	4000	0.2300	23.0527