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: 0.2146
Wer: 12.3134
Avg Precision Exact: 0.8990
Avg Recall Exact: 0.9012
Avg F1 Exact: 0.8997
Avg Precision Letter Shift: 0.9196
Avg Recall Letter Shift: 0.9219
Avg F1 Letter Shift: 0.9204
Avg Precision Word Level: 0.9216
Avg Recall Word Level: 0.9240
Avg F1 Word Level: 0.9224
Avg Precision Word Shift: 0.9730
Avg Recall Word Shift: 0.9761
Avg F1 Word Shift: 0.9741
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 Word Shift	Recall Min Word Shift	F1 Min Word Shift
No log	8e-05	1	7.1551	102.7088	0.0004	0.0034	0.0008	0.0169	0.0167	0.0165	0.0045	0.0485	0.0082	0.0865	0.0873	0.0863	0.0	0.0	0.0	0.1111	1.0	0.2000	0.0	0.0	0.0
0.0271	0.8	10000	0.1443	15.7539	0.8698	0.8704	0.8696	0.8921	0.8928	0.8920	0.8951	0.8962	0.8951	0.9611	0.9641	0.9620	0.9333	0.9333	0.9474	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0138	1.6	20000	0.1523	14.3570	0.8784	0.8793	0.8784	0.9010	0.9020	0.9010	0.9042	0.9053	0.9043	0.9672	0.9691	0.9676	0.9737	1.0	0.9565	1.0	1.0	1.0	0.0769	0.0769	0.0769
0.0059	2.4	30000	0.1584	17.4242	0.8838	0.8843	0.8835	0.9054	0.9061	0.9051	0.9077	0.9090	0.9076	0.9658	0.9681	0.9661	1.0	1.0	0.9655	1.0	1.0	1.0	0.0	0.0	0.0
0.0038	3.2	40000	0.1696	13.3925	0.8917	0.8933	0.8921	0.9129	0.9147	0.9134	0.9155	0.9171	0.9159	0.9724	0.9743	0.9729	1.0	1.0	0.9677	1.0	1.0	1.0	0.1429	0.1	0.1176
0.0039	4.0	50000	0.1695	13.2779	0.8819	0.8816	0.8814	0.9029	0.9027	0.9024	0.9052	0.9052	0.9048	0.9707	0.9714	0.9706	1.0	1.0	0.9697	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0027	4.8	60000	0.1823	13.4294	0.8837	0.8852	0.8841	0.9047	0.9066	0.9052	0.9073	0.9094	0.9078	0.9699	0.9729	0.9708	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1	0.1176
0.0008	5.6	70000	0.1762	13.0340	0.8870	0.8902	0.8883	0.9071	0.9104	0.9084	0.9097	0.9131	0.9110	0.9710	0.9751	0.9726	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1111	0.125
0.0006	6.4	80000	0.1845	12.8012	0.8926	0.8943	0.8931	0.9131	0.9150	0.9137	0.9155	0.9174	0.9161	0.9729	0.9753	0.9737	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1	0.1176
0.0009	7.2	90000	0.1883	12.6497	0.8960	0.8943	0.8948	0.9173	0.9158	0.9162	0.9197	0.9184	0.9187	0.9745	0.9747	0.9742	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1	0.1176
0.0003	8.0	100000	0.1888	12.3688	0.8960	0.8977	0.8965	0.9169	0.9186	0.9174	0.9192	0.9211	0.9198	0.9745	0.9772	0.9755	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0023	8.8	110000	0.1934	12.5240	0.8962	0.8948	0.8951	0.9162	0.9148	0.9151	0.9184	0.9172	0.9174	0.9737	0.9740	0.9734	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1	0.1176
0.0005	9.6	120000	0.2024	12.6534	0.8919	0.8935	0.8924	0.9122	0.9139	0.9127	0.9146	0.9163	0.9151	0.9720	0.9745	0.9728	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1111	0.125
0.0001	10.4	130000	0.1999	13.0229	0.8901	0.8925	0.8909	0.9109	0.9135	0.9118	0.9134	0.9160	0.9143	0.9714	0.9749	0.9727	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0002	11.2	140000	0.2010	12.9675	0.8873	0.8920	0.8892	0.9085	0.9135	0.9106	0.9112	0.9163	0.9133	0.9698	0.9752	0.9720	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0001	12.0	150000	0.2078	12.8455	0.8932	0.8961	0.8943	0.9149	0.9181	0.9161	0.9170	0.9202	0.9182	0.9724	0.9762	0.9739	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1111	0.125
0.0001	12.8	160000	0.2063	12.6718	0.8950	0.8962	0.8952	0.9161	0.9175	0.9164	0.9183	0.9198	0.9187	0.9731	0.9757	0.9740	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1111	0.125
0.0005	13.6	170000	0.2137	12.5203	0.8941	0.8960	0.8947	0.9143	0.9164	0.9150	0.9165	0.9186	0.9172	0.9712	0.9742	0.9723	1.0	1.0	1.0	1.0	1.0	1.0	0.0909	0.1	0.1000
0.0002	14.4	180000	0.2121	12.5979	0.8945	0.8964	0.8951	0.9155	0.9176	0.9162	0.9178	0.9199	0.9185	0.9731	0.9760	0.9741	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1111	0.125
0.0	15.2	190000	0.2143	12.3282	0.8985	0.9008	0.8993	0.9193	0.9218	0.9202	0.9215	0.9240	0.9224	0.9734	0.9766	0.9746	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0	16.0	200000	0.2146	12.3134	0.8990	0.9012	0.8997	0.9196	0.9219	0.9204	0.9216	0.9240	0.9224	0.9730	0.9761	0.9741	1.0	1.0	1.0	1.0	1.0	1.0	0.1429	0.1	0.1176

Framework versions

Transformers 4.41.2
Pytorch 2.2.1
Datasets 2.20.0
Tokenizers 0.19.1

cantillation
/

Teamim-small_Random_WeightDecay-0.05_Augmented_New-Data_date-02-08-2024

he-cantillation

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Model tree for cantillation/Teamim-small_Random_WeightDecay-0.05_Augmented_New-Data_date-02-08-2024

Evaluation results