--- language: - he license: apache-2.0 base_model: openai/whisper-small 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-small](https://huggingface.co/openai/whisper-small) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.1845 - Wer: 12.0473 - Avg Precision Exact: 0.9003 - Avg Recall Exact: 0.8996 - Avg F1 Exact: 0.8996 - Avg Precision Letter Shift: 0.9202 - Avg Recall Letter Shift: 0.9196 - Avg F1 Letter Shift: 0.9195 - Avg Precision Word Level: 0.9230 - Avg Recall Word Level: 0.9222 - Avg F1 Word Level: 0.9223 - Avg Precision Word Shift: 0.9761 - Avg Recall Word Shift: 0.9759 - Avg F1 Word Shift: 0.9756 - 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: 0.0909 - Recall Min Word Shift: 0.125 - F1 Min Word Shift: 0.1053 ## 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: 1000 - training_steps: 100000 - 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 | 0.0 | 1 | 7.4353 | 114.3865 | 0.0009 | 0.0042 | 0.0014 | 0.0165 | 0.0189 | 0.0160 | 0.0059 | 0.0313 | 0.0094 | 0.0797 | 0.0892 | 0.0786 | 0.0 | 0.0 | 0.0 | 0.2 | 1.0 | 0.25 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | | 0.0187 | 0.8 | 10000 | 0.1393 | 16.0458 | 0.8664 | 0.8713 | 0.8683 | 0.8907 | 0.8958 | 0.8927 | 0.8947 | 0.8995 | 0.8965 | 0.9600 | 0.9642 | 0.9615 | 0.9286 | 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 | 0.1 | 0.1176 | | 0.0075 | 1.6 | 20000 | 0.1476 | 14.6009 | 0.8758 | 0.8790 | 0.8770 | 0.8983 | 0.9016 | 0.8994 | 0.9016 | 0.9054 | 0.9030 | 0.9659 | 0.9707 | 0.9678 | 0.9375 | 0.9444 | 0.9524 | 1.0 | 1.0 | 1.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.2222 | 0.1818 | 0.2000 | | 0.0095 | 2.4 | 30000 | 0.1548 | 13.7140 | 0.8840 | 0.8849 | 0.8841 | 0.9054 | 0.9064 | 0.9055 | 0.9085 | 0.9093 | 0.9085 | 0.9695 | 0.9710 | 0.9698 | 1.0 | 1.0 | 0.9630 | 1.0 | 1.0 | 1.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.2 | 0.1538 | 0.1739 | | 0.0025 | 3.2 | 40000 | 0.1539 | 13.0636 | 0.8895 | 0.8892 | 0.8890 | 0.9103 | 0.9102 | 0.9099 | 0.9129 | 0.9128 | 0.9125 | 0.9722 | 0.9729 | 0.9721 | 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 | 0.125 | 0.1333 | | 0.0019 | 4.0 | 50000 | 0.1636 | 12.6016 | 0.8922 | 0.8928 | 0.8922 | 0.9130 | 0.9137 | 0.9130 | 0.9158 | 0.9163 | 0.9157 | 0.9737 | 0.9746 | 0.9737 | 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 | 0.2222 | 0.1667 | 0.1905 | | 0.0009 | 4.8 | 60000 | 0.1743 | 12.5795 | 0.8969 | 0.8972 | 0.8967 | 0.9176 | 0.9180 | 0.9174 | 0.9210 | 0.9212 | 0.9207 | 0.9743 | 0.9745 | 0.9740 | 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 | 0.125 | 0.1053 | | 0.0011 | 5.6 | 70000 | 0.1819 | 12.5314 | 0.8986 | 0.8982 | 0.8980 | 0.9189 | 0.9186 | 0.9183 | 0.9219 | 0.9213 | 0.9212 | 0.9749 | 0.9753 | 0.9747 | 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 | | 0.0009 | 6.4 | 80000 | 0.1802 | 12.3023 | 0.8977 | 0.8974 | 0.8972 | 0.9179 | 0.9176 | 0.9174 | 0.9207 | 0.9202 | 0.9201 | 0.9755 | 0.9755 | 0.9750 | 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 | | 0.0 | 7.2 | 90000 | 0.1826 | 12.1064 | 0.9011 | 0.9001 | 0.9003 | 0.9212 | 0.9203 | 0.9204 | 0.9240 | 0.9230 | 0.9232 | 0.9765 | 0.9760 | 0.9758 | 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 | 0.1000 | | 0.0 | 8.0 | 100000 | 0.1845 | 12.0473 | 0.9003 | 0.8996 | 0.8996 | 0.9202 | 0.9196 | 0.9195 | 0.9230 | 0.9222 | 0.9223 | 0.9761 | 0.9759 | 0.9756 | 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 | 0.125 | 0.1053 | ### Framework versions - Transformers 4.39.0.dev0 - Pytorch 2.2.1+cu121 - Datasets 2.16.1 - Tokenizers 0.15.0