cantillation/Teamim-small_WeightDecay-0.05_Combined-Data_date-17-07-2024_10-08

he-cantillation

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

• Loss: 0.1496
• Wer: 9.8776
• Avg Precision Exact: 0.9087
• Avg Recall Exact: 0.9107
• Avg F1 Exact: 0.9093
• Avg Precision Letter Shift: 0.9225
• Avg Recall Letter Shift: 0.9246
• Avg F1 Letter Shift: 0.9232
• Avg Precision Word Level: 0.9249
• Avg Recall Word Level: 0.9267
• Avg F1 Word Level: 0.9254
• Avg Precision Word Shift: 0.9697
• Avg Recall Word Shift: 0.9724
• Avg F1 Word Shift: 0.9706
• Precision Median Exact: 1.0
• Recall Median Exact: 1.0
• F1 Median Exact: 1.0
• Precision Max Exact: 1.0
• Recall Max Exact: 1.0
• F1 Max Exact: 1.0
• Precision Min Exact: 0.0
• Recall Min Exact: 0.0
• F1 Min Exact: 0.0
• Precision Min Letter Shift: 0.0
• Recall Min Letter Shift: 0.0
• F1 Min Letter Shift: 0.0
• Precision Min Word Level: 0.0
• Recall Min Word Level: 0.0
• F1 Min Word Level: 0.0
• Precision Min Word Shift: 0.1429
• Recall Min Word Shift: 0.1
• F1 Min Word Shift: 0.1176

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 8
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 200000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Avg Precision Exact	Avg Recall Exact	Avg F1 Exact	Avg Precision Letter Shift	Avg Recall Letter Shift	Avg F1 Letter Shift	Avg Precision Word Level	Avg Recall Word Level	Avg F1 Word Level	Avg Precision Word Shift	Avg Recall Word Shift	Avg F1 Word Shift	Precision Median Exact	Recall Median Exact	F1 Median Exact	Precision Max Exact	Recall Max Exact	F1 Max Exact	Precision Min Exact	Recall Min Exact	F1 Min Exact	Precision Min Letter Shift	Recall Min Letter Shift	F1 Min Letter Shift	Precision Min Word Level	Recall Min Word Level	F1 Min Word Level	Precision Min Word Shift	Recall Min Word Shift	F1 Min Word Shift
No log	0.0001	1	7.3549	108.0965	0.0006	0.0037	0.0009	0.0204	0.0217	0.0204	0.0078	0.0760	0.0140	0.1364	0.1552	0.1388	0.0	0.0	0.0	0.2	1.0	0.3077	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
0.0527	0.5167	10000	0.1247	17.8011	0.8390	0.8441	0.8410	0.8596	0.8648	0.8616	0.8632	0.8682	0.8651	0.9344	0.9409	0.9369	0.9167	0.9167	0.9167	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0769	0.0667	0.0714
0.0181	1.0334	20000	0.1178	14.7550	0.8697	0.8697	0.8690	0.8884	0.8887	0.8878	0.8911	0.8918	0.8907	0.9502	0.9517	0.9501	0.9333	0.9333	0.9474	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
0.0121	1.5501	30000	0.1202	13.1219	0.8866	0.8899	0.8878	0.9039	0.9074	0.9052	0.9067	0.9103	0.9080	0.9551	0.9595	0.9568	1.0	1.0	0.9630	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.125	0.1333
0.0041	2.0668	40000	0.1301	12.2219	0.8897	0.8925	0.8907	0.9058	0.9087	0.9068	0.9089	0.9114	0.9097	0.9579	0.9612	0.9591	1.0	1.0	0.9655	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.125	0.1	0.1176
0.0053	2.5834	50000	0.1294	12.3635	0.8863	0.8873	0.8864	0.9030	0.9042	0.9031	0.9055	0.9068	0.9057	0.9594	0.9623	0.9603	1.0	1.0	0.9630	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0909	0.1	0.0952
0.01	3.1001	60000	0.1385	12.5586	0.8824	0.8860	0.8838	0.8984	0.9023	0.8999	0.9010	0.9049	0.9025	0.9557	0.9609	0.9578	1.0	1.0	0.9630	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0037	3.6168	70000	0.1374	12.4076	0.8905	0.8936	0.8916	0.9062	0.9093	0.9073	0.9085	0.9115	0.9095	0.9561	0.9598	0.9574	1.0	1.0	0.9630	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1111	0.125
0.0031	4.1335	80000	0.1370	11.7562	0.8959	0.8971	0.8961	0.9115	0.9128	0.9117	0.9137	0.9150	0.9139	0.9610	0.9639	0.9620	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
0.0012	4.6502	90000	0.1371	11.4573	0.8972	0.8963	0.8964	0.9126	0.9119	0.9118	0.9150	0.9146	0.9144	0.9655	0.9665	0.9655	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0021	5.1669	100000	0.1448	11.6492	0.8926	0.8952	0.8935	0.9075	0.9103	0.9085	0.9102	0.9128	0.9111	0.9601	0.9637	0.9614	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1	0.1	0.1053
0.0013	5.6836	110000	0.1440	11.0199	0.9003	0.9033	0.9014	0.9156	0.9188	0.9168	0.9179	0.9211	0.9191	0.9645	0.9685	0.9660	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0004	6.2003	120000	0.1427	10.7901	0.9033	0.9043	0.9035	0.9183	0.9194	0.9185	0.9205	0.9215	0.9206	0.9667	0.9686	0.9672	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0005	6.7170	130000	0.1434	10.7178	0.9012	0.9039	0.9021	0.9161	0.9189	0.9171	0.9185	0.9212	0.9194	0.9648	0.9684	0.9662	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1111	0.125
0.0006	7.2336	140000	0.1452	10.7776	0.9014	0.9041	0.9023	0.9162	0.9190	0.9172	0.9187	0.9211	0.9195	0.9661	0.9699	0.9676	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1111	0.125
0.0005	7.7503	150000	0.1471	10.6297	0.9037	0.9039	0.9034	0.9177	0.9180	0.9175	0.9198	0.9199	0.9195	0.9656	0.9673	0.9660	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0001	8.2670	160000	0.1484	10.3905	0.9035	0.9068	0.9048	0.9177	0.9212	0.9191	0.9200	0.9231	0.9212	0.9667	0.9703	0.9680	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0019	8.7837	170000	0.1468	10.1136	0.9078	0.9099	0.9085	0.9220	0.9242	0.9227	0.9244	0.9262	0.9249	0.9679	0.9704	0.9687	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0	9.3004	180000	0.1495	10.1167	0.9051	0.9073	0.9058	0.9194	0.9217	0.9202	0.9219	0.9241	0.9226	0.9681	0.9709	0.9691	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0	9.8171	190000	0.1497	10.0412	0.9078	0.9104	0.9087	0.9218	0.9245	0.9227	0.9243	0.9266	0.9251	0.9680	0.9713	0.9692	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176
0.0	10.3338	200000	0.1496	9.8776	0.9087	0.9107	0.9093	0.9225	0.9246	0.9232	0.9249	0.9267	0.9254	0.9697	0.9724	0.9706	1.0	1.0	1.0	1.0	1.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.1429	0.1	0.1176

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.1
• Datasets 2.20.0
• Tokenizers 0.19.1