whisper-large-v3-mixed-pt

This model is a fine-tuned version of openai/whisper-large-v3 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:

learning_rate: 1e-05
train_batch_size: 256
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10

Training Loss	Epoch	Step	Validation Loss
0.2809	0.3106	50	0.1335
0.1703	0.6211	100	0.1115
0.1533	0.9317	150	0.1090
0.1145	1.2422	200	0.1084
0.1084	1.5528	250	0.1093
0.1192	1.8634	300	0.1054
0.0687	2.1739	350	0.1124
0.07	2.4845	400	0.1136
0.0687	2.7950	450	0.1128
0.0548	3.1056	500	0.1282
0.0438	3.4161	550	0.1263
0.0486	3.7267	600	0.1259
0.0428	4.0373	650	0.1297
0.0321	4.3478	700	0.1399
0.0329	4.6584	750	0.1366
0.0358	4.9689	800	0.1404
0.0224	5.2795	850	0.1586
0.0241	5.5901	900	0.1615
0.0219	5.9006	950	0.1599
0.0141	6.2112	1000	0.1701
0.0166	6.5217	1050	0.1745
0.0169	6.8323	1100	0.1734
0.0118	7.1429	1150	0.1843
0.0123	7.4534	1200	0.1836
0.0128	7.7640	1250	0.1869
0.0103	8.0745	1300	0.1881
0.01	8.3851	1350	0.1927
0.0107	8.6957	1400	0.1907
0.009	9.0062	1450	0.1923
0.0088	9.3168	1500	0.1970
0.0087	9.6273	1550	0.1974
0.0083	9.9379	1600	0.1967