--- language: - he license: apache-2.0 base_model: openai/whisper-large-v2 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-large-v2](https://huggingface.co/openai/whisper-large-v2) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.1039 - Wer: 11.7812 - Avg Precision Exact: 0.9060 - Avg Recall Exact: 0.9051 - Avg F1 Exact: 0.9052 - Avg Precision Letter Shift: 0.9250 - Avg Recall Letter Shift: 0.9242 - Avg F1 Letter Shift: 0.9242 - Avg Precision Word Level: 0.9276 - Avg Recall Word Level: 0.9270 - Avg F1 Word Level: 0.9269 - Avg Precision Word Shift: 0.9735 - Avg Recall Word Shift: 0.9738 - Avg F1 Word Shift: 0.9732 - 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: 50 - training_steps: 8000 - 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 | 5.7633 | 117.0584 | 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 | | 0.1301 | 0.08 | 1000 | 0.1621 | 22.8271 | 0.8134 | 0.8185 | 0.8150 | 0.8421 | 0.8478 | 0.8440 | 0.8463 | 0.8533 | 0.8488 | 0.9240 | 0.9351 | 0.9284 | 0.9 | 0.9 | 0.8966 | 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 | | 0.0926 | 0.16 | 2000 | 0.1298 | 17.9675 | 0.8521 | 0.8555 | 0.8532 | 0.8767 | 0.8802 | 0.8778 | 0.8803 | 0.8838 | 0.8814 | 0.9476 | 0.9528 | 0.9495 | 0.9231 | 0.9231 | 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 | 0.0 | 0.0 | 0.0 | | 0.0393 | 0.24 | 3000 | 0.1162 | 15.4250 | 0.8854 | 0.8838 | 0.8841 | 0.9067 | 0.9055 | 0.9055 | 0.9096 | 0.9089 | 0.9087 | 0.9643 | 0.9651 | 0.9641 | 0.9375 | 0.9375 | 0.9565 | 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 | | 0.0345 | 0.32 | 4000 | 0.1122 | 14.0946 | 0.8904 | 0.8905 | 0.8900 | 0.9124 | 0.9125 | 0.9120 | 0.9155 | 0.9164 | 0.9155 | 0.9657 | 0.9678 | 0.9662 | 0.9412 | 0.9412 | 0.9565 | 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 | | 0.0264 | 0.4 | 5000 | 0.1077 | 12.7827 | 0.9002 | 0.9017 | 0.9005 | 0.9194 | 0.9210 | 0.9197 | 0.9216 | 0.9237 | 0.9222 | 0.9692 | 0.9720 | 0.9701 | 1.0 | 1.0 | 0.9677 | 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 | 0.125 | 0.1333 | | 0.0247 | 0.48 | 6000 | 0.1036 | 12.6829 | 0.9002 | 0.9013 | 0.9004 | 0.9198 | 0.9211 | 0.9200 | 0.9222 | 0.9238 | 0.9226 | 0.9703 | 0.9731 | 0.9712 | 1.0 | 1.0 | 0.9677 | 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 | | 0.0115 | 0.56 | 7000 | 0.1045 | 12.0103 | 0.9035 | 0.9036 | 0.9032 | 0.9230 | 0.9231 | 0.9226 | 0.9256 | 0.9262 | 0.9255 | 0.9721 | 0.9732 | 0.9722 | 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 | | 0.0123 | 0.64 | 8000 | 0.1039 | 11.7812 | 0.9060 | 0.9051 | 0.9052 | 0.9250 | 0.9242 | 0.9242 | 0.9276 | 0.9270 | 0.9269 | 0.9735 | 0.9738 | 0.9732 | 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