he-cantillation
This model is a fine-tuned version of ivrit-ai/whisper-v2-pd1-e1 on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0952
- Wer: 7.8511
- Avg Precision Exact: 0.9335
- Avg Recall Exact: 0.9352
- Avg F1 Exact: 0.9340
- Avg Precision Letter Shift: 0.9440
- Avg Recall Letter Shift: 0.9458
- Avg F1 Letter Shift: 0.9446
- Avg Precision Word Level: 0.9462
- Avg Recall Word Level: 0.9479
- Avg F1 Word Level: 0.9467
- Avg Precision Word Shift: 0.9714
- Avg Recall Word Shift: 0.9736
- Avg F1 Word Shift: 0.9721
- 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: 0.125
- F1 Min Word Shift: 0.1333
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: 80000
- 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 |
5.0835 |
121.5079 |
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 |
0.0 |
0.0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
| 0.0423 |
0.5167 |
10000 |
0.1010 |
13.7858 |
0.8705 |
0.8797 |
0.8745 |
0.8866 |
0.8961 |
0.8908 |
0.8899 |
0.8991 |
0.8939 |
0.9426 |
0.9519 |
0.9466 |
0.9286 |
0.9412 |
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 |
| 0.0139 |
1.0334 |
20000 |
0.0950 |
10.6832 |
0.9090 |
0.9076 |
0.9079 |
0.9219 |
0.9205 |
0.9208 |
0.9251 |
0.9237 |
0.9239 |
0.9610 |
0.9611 |
0.9605 |
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 |
| 0.0089 |
1.5501 |
30000 |
0.0914 |
10.2458 |
0.9091 |
0.9077 |
0.9081 |
0.9208 |
0.9196 |
0.9198 |
0.9231 |
0.9220 |
0.9222 |
0.9596 |
0.9590 |
0.9589 |
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 |
0.125 |
0.1111 |
| 0.0032 |
2.0668 |
40000 |
0.0922 |
9.3269 |
0.9163 |
0.9159 |
0.9157 |
0.9282 |
0.9279 |
0.9277 |
0.9307 |
0.9303 |
0.9301 |
0.9666 |
0.9676 |
0.9667 |
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 |
0.0909 |
0.1111 |
0.1176 |
| 0.0025 |
2.5834 |
50000 |
0.0924 |
9.0500 |
0.9171 |
0.9179 |
0.9172 |
0.9283 |
0.9292 |
0.9284 |
0.9307 |
0.9314 |
0.9307 |
0.9656 |
0.9669 |
0.9659 |
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 |
0.125 |
0.1333 |
| 0.0022 |
3.1001 |
60000 |
0.0933 |
8.3137 |
0.9272 |
0.9266 |
0.9266 |
0.9377 |
0.9371 |
0.9371 |
0.9399 |
0.9393 |
0.9393 |
0.9702 |
0.9702 |
0.9698 |
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 |
0.125 |
0.1333 |
| 0.0006 |
3.6168 |
70000 |
0.0947 |
8.0682 |
0.9287 |
0.9302 |
0.9291 |
0.9393 |
0.9409 |
0.9398 |
0.9417 |
0.9430 |
0.9420 |
0.9706 |
0.9723 |
0.9710 |
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 |
0.125 |
0.1333 |
| 0.0002 |
4.1335 |
80000 |
0.0952 |
7.8511 |
0.9335 |
0.9352 |
0.9340 |
0.9440 |
0.9458 |
0.9446 |
0.9462 |
0.9479 |
0.9467 |
0.9714 |
0.9736 |
0.9721 |
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 |
0.125 |
0.1333 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.2.1
- Datasets 2.20.0
- Tokenizers 0.19.1
