--- 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](https://huggingface.co/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