metadata
language:
- he
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- hf-asr-leaderboard
- generated_from_trainer
metrics:
- wer
model-index:
- name: he-cantillation
results: []
he-cantillation
This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1635
- Wer: 15.8869
- Avg Precision Exact: 0.8555
- Avg Recall Exact: 0.8550
- Avg F1 Exact: 0.8548
- Avg Precision Letter Shift: 0.8805
- Avg Recall Letter Shift: 0.8801
- Avg F1 Letter Shift: 0.8799
- Avg Precision Word Level: 0.8850
- Avg Recall Word Level: 0.8845
- Avg F1 Word Level: 0.8842
- Avg Precision Word Shift: 0.9658
- Avg Recall Word Shift: 0.9665
- Avg F1 Word Shift: 0.9656
- Precision Median Exact: 0.9333
- Recall Median Exact: 0.9333
- F1 Median Exact: 0.9375
- 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: 0.1111
- F1 Min Word Shift: 0.125
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: 0.0001
- 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: 50
- training_steps: 20000
- 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 | 8e-05 | 1 | 8.6830 | 101.5891 | 0.0009 | 0.0016 | 0.0010 | 0.0100 | 0.0077 | 0.0077 | 0.0054 | 0.0173 | 0.0076 | 0.0465 | 0.0380 | 0.0371 | 0.0 | 0.0 | 0.0 | 0.125 | 0.3333 | 0.1667 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
0.0958 | 0.32 | 4000 | 0.1924 | 25.9534 | 0.7565 | 0.7583 | 0.7566 | 0.7917 | 0.7936 | 0.7919 | 0.7983 | 0.8005 | 0.7986 | 0.9281 | 0.9324 | 0.9293 | 0.8571 | 0.8571 | 0.8571 | 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 |
0.0539 | 0.64 | 8000 | 0.1813 | 21.8625 | 0.7983 | 0.7994 | 0.7983 | 0.8279 | 0.8291 | 0.8279 | 0.8342 | 0.8359 | 0.8344 | 0.9437 | 0.9468 | 0.9445 | 0.9091 | 0.9091 | 0.9032 | 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 |
0.0353 | 0.96 | 12000 | 0.1755 | 19.0909 | 0.8271 | 0.8284 | 0.8273 | 0.8558 | 0.8572 | 0.8560 | 0.8610 | 0.8624 | 0.8611 | 0.9537 | 0.9561 | 0.9542 | 0.9167 | 0.9167 | 0.9231 | 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 |
0.0197 | 1.28 | 16000 | 0.1664 | 17.0288 | 0.8452 | 0.8444 | 0.8444 | 0.8711 | 0.8705 | 0.8703 | 0.8762 | 0.8756 | 0.8754 | 0.9609 | 0.9625 | 0.9610 | 0.9286 | 0.9286 | 0.9286 | 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 |
0.0088 | 1.6 | 20000 | 0.1635 | 15.8869 | 0.8555 | 0.8550 | 0.8548 | 0.8805 | 0.8801 | 0.8799 | 0.8850 | 0.8845 | 0.8842 | 0.9658 | 0.9665 | 0.9656 | 0.9333 | 0.9333 | 0.9375 | 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 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.2.1
- Datasets 2.20.0
- Tokenizers 0.19.1