Hugging Face's logo

cantillation
/
Teamim-tiny_DropOut-0.5_Augmented_Combined-Data_date-07-07-2024_08-34

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Hebrew
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Teamim-tiny_DropOut-0.5_Augmented_Combined-Data_date-07-07-2024_08-34 / README.md
cantillation's picture
cantillation
End of training
6a234db verified
preview code
|
raw
history blame contribute delete
22.9 kB
metadata
language:
  - he
license: apache-2.0
base_model: >-
  cantillation/Teamim-tiny_DropOut-0.5_Augmented_Combined-Data_date-06-07-2024_20-19
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 cantillation/Teamim-tiny_DropOut-0.5_Augmented_Combined-Data_date-06-07-2024_20-19 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 3.9592
  • Wer: 100.0
  • Avg Precision Exact: 0.0053
  • Avg Recall Exact: 0.0482
  • Avg F1 Exact: 0.0094
  • Avg Precision Letter Shift: 0.0129
  • Avg Recall Letter Shift: 0.1183
  • Avg F1 Letter Shift: 0.0228
  • Avg Precision Word Level: 0.0173
  • Avg Recall Word Level: 0.1573
  • Avg F1 Word Level: 0.0306
  • Avg Precision Word Shift: 0.0334
  • Avg Recall Word Shift: 0.2899
  • Avg F1 Word Shift: 0.0588
  • Precision Median Exact: 0.0
  • Recall Median Exact: 0.0
  • F1 Median Exact: 0.0
  • Precision Max Exact: 0.1667
  • Recall Max Exact: 1.0
  • F1 Max Exact: 0.25
  • 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: 0.0
  • Recall Min Word Shift: 0.0
  • F1 Min Word Shift: 0.0

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-06
  • 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: 50000
  • 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 0.2806 16.9452 0.8385 0.8403 0.8390 0.8609 0.8628 0.8613 0.8657 0.8679 0.8663 0.9468 0.9505 0.9481 0.9231 0.9231 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 0.0833 0.0769 0.08
6.379 0.1033 2000 1.1953 120.7433 0.1765 0.2759 0.1896 0.2338 0.3541 0.2490 0.2533 0.3945 0.2751 0.4039 0.6029 0.4412 0.1538 0.2 0.1667 1.0 1.0 0.9412 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5.4428 0.2067 4000 4.7348 100.0 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
5.1556 0.3100 6000 4.6051 100.0 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
4.935 0.4134 8000 4.4172 100.0 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
4.7728 0.5167 10000 4.1229 100.0 0.0 0.0 0.0 0.0003 0.0026 0.0005 0.0008 0.0092 0.0015 0.0012 0.0141 0.0022 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
4.636 0.6200 12000 4.0112 100.0378 0.0071 0.0423 0.0116 0.0155 0.0901 0.0251 0.0260 0.1528 0.0422 0.0567 0.3257 0.0922 0.0 0.0 0.0 0.2222 1.0 0.3333 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.4949 0.7234 14000 4.0080 100.0 0.0040 0.0390 0.0072 0.0097 0.0899 0.0171 0.0180 0.1677 0.0317 0.0349 0.3220 0.0615 0.0 0.0 0.0 0.1818 1.0 0.3077 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.4359 0.8267 16000 4.0252 100.0 0.0031 0.0347 0.0057 0.0075 0.0849 0.0138 0.0144 0.1589 0.0262 0.0250 0.2827 0.0457 0.0 0.0 0.0 0.1429 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
4.3834 0.9300 18000 4.0642 100.0 0.0025 0.0273 0.0045 0.0063 0.0700 0.0115 0.0125 0.1372 0.0228 0.0225 0.2534 0.0412 0.0 0.0 0.0 0.1429 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
4.317 1.0334 20000 4.0775 100.0 0.0017 0.0183 0.0031 0.0043 0.0470 0.0079 0.0080 0.0860 0.0146 0.0150 0.1663 0.0274 0.0 0.0 0.0 0.1429 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
4.269 1.1367 22000 4.1070 100.0 0.0010 0.0108 0.0018 0.0025 0.0266 0.0046 0.0044 0.0447 0.0079 0.0080 0.0834 0.0144 0.0 0.0 0.0 0.125 1.0 0.2222 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.2529 1.2401 24000 4.0914 100.0 0.0008 0.0086 0.0015 0.0021 0.0215 0.0038 0.0033 0.0339 0.0060 0.0064 0.0658 0.0115 0.0 0.0 0.0 0.125 1.0 0.2222 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.1986 1.3434 26000 4.0861 100.0 0.0012 0.0131 0.0022 0.0025 0.0255 0.0044 0.0038 0.0404 0.0070 0.0081 0.0865 0.0146 0.0 0.0 0.0 0.1429 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
4.1805 1.4467 28000 4.0716 100.0 0.0016 0.0171 0.0029 0.0034 0.0362 0.0062 0.0055 0.0579 0.0099 0.0103 0.1118 0.0188 0.0 0.0 0.0 0.1429 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
4.1489 1.5501 30000 4.0526 100.0 0.0027 0.0306 0.0050 0.0065 0.0701 0.0117 0.0096 0.1033 0.0174 0.0174 0.1901 0.0316 0.0 0.0 0.0 0.1429 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
4.0979 1.6534 32000 4.0399 100.0 0.0034 0.0389 0.0062 0.0082 0.0924 0.0149 0.0121 0.1331 0.0220 0.0217 0.2389 0.0394 0.0 0.0 0.0 0.2 1.0 0.3333 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.0667 1.7567 34000 4.0224 100.0 0.0042 0.0473 0.0076 0.0099 0.1091 0.0179 0.0143 0.1549 0.0259 0.0258 0.2749 0.0465 0.0 0.0 0.0 0.2 1.0 0.3333 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.0602 1.8601 36000 4.0042 100.0 0.0040 0.0466 0.0073 0.0103 0.1182 0.0187 0.0142 0.1606 0.0259 0.0254 0.2818 0.0461 0.0 0.0 0.0 0.2 1.0 0.3333 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.0584 1.9634 38000 3.9922 100.0 0.0038 0.0441 0.0069 0.0104 0.1215 0.0190 0.0140 0.1623 0.0256 0.0244 0.2807 0.0446 0.0 0.0 0.0 0.2 1.0 0.3333 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.0015 2.0668 40000 3.9825 100.0 0.0042 0.0506 0.0077 0.0111 0.1293 0.0203 0.0149 0.1726 0.0272 0.0254 0.2923 0.0464 0.0 0.0 0.0 0.1429 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
4.0073 2.1701 42000 3.9722 100.0 0.0044 0.0508 0.0080 0.0112 0.1285 0.0205 0.0149 0.1704 0.0272 0.0255 0.2891 0.0465 0.0 0.0 0.0 0.1429 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
4.0284 2.2734 44000 3.9651 100.0 0.0047 0.0506 0.0084 0.0116 0.1258 0.0211 0.0156 0.1669 0.0281 0.0278 0.2904 0.0501 0.0 0.0 0.0 0.1429 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
3.9935 2.3768 46000 3.9619 100.0 0.0051 0.0507 0.0092 0.0122 0.1230 0.0219 0.0165 0.1640 0.0295 0.0298 0.2889 0.0532 0.0 0.0 0.0 0.1667 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
4.0324 2.4801 48000 3.9600 100.0 0.0052 0.0485 0.0092 0.0126 0.1197 0.0225 0.0171 0.1597 0.0304 0.0326 0.2909 0.0577 0.0 0.0 0.0 0.1667 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
4.003 2.5834 50000 3.9592 100.0 0.0053 0.0482 0.0094 0.0129 0.1183 0.0228 0.0173 0.1573 0.0306 0.0334 0.2899 0.0588 0.0 0.0 0.0 0.1667 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

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.2.1
  • Datasets 2.20.0
  • Tokenizers 0.19.1