Hugging Face's logo

cantillation
/
Teamim-small_WeightDecay-0.05_Augmented_New-Data_nusach-yerushalmi_date-24-07-2024

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Hebrew
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
Teamim-small_WeightDecay-0.05_Augmented_New-Data_nusach-yerushalmi_date-24-07-2024 / README.md
cantillation's picture
cantillation
End of training
53028d4 verified
preview code
|
raw
history blame contribute delete
19.1 kB
metadata
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 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2759
  • Wer: 12.0860
  • Avg Precision Exact: 0.9045
  • Avg Recall Exact: 0.9054
  • Avg F1 Exact: 0.9045
  • Avg Precision Letter Shift: 0.9223
  • Avg Recall Letter Shift: 0.9233
  • Avg F1 Letter Shift: 0.9224
  • Avg Precision Word Level: 0.9250
  • Avg Recall Word Level: 0.9259
  • Avg F1 Word Level: 0.9250
  • Avg Precision Word Shift: 0.9777
  • Avg Recall Word Shift: 0.9785
  • Avg F1 Word Shift: 0.9777
  • 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: 0.0
  • 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-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: 500
  • training_steps: 200000
  • 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 0.0004 1 6.8177 106.5214 0.0004 0.0012 0.0006 0.0038 0.0036 0.0033 0.0030 0.0121 0.0043 0.0322 0.0342 0.0300 0.0 0.0 0.0 0.0909 0.3333 0.1429 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0059 3.7023 10000 0.1748 15.8840 0.8772 0.8813 0.8786 0.9013 0.9056 0.9028 0.9063 0.9103 0.9077 0.9648 0.9693 0.9663 0.9286 0.9375 0.9474 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
0.0045 7.4047 20000 0.2047 15.1038 0.8686 0.8670 0.8673 0.8906 0.8892 0.8894 0.8952 0.8935 0.8938 0.9722 0.9711 0.9710 0.9375 0.9333 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 0.0 0.0 0.0
0.001 11.1070 30000 0.2024 13.8083 0.8862 0.8876 0.8863 0.9076 0.9094 0.9080 0.9109 0.9127 0.9113 0.9743 0.9767 0.9749 1.0 1.0 0.9600 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
0.001 14.8093 40000 0.2188 13.8083 0.8924 0.8918 0.8916 0.9125 0.9118 0.9116 0.9166 0.9156 0.9155 0.9733 0.9730 0.9726 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 0.0 0.0 0.0
0.0005 18.5117 50000 0.2256 13.6464 0.8921 0.8937 0.8924 0.9131 0.9148 0.9135 0.9161 0.9176 0.9164 0.9760 0.9774 0.9762 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 0.0 0.0 0.0
0.0012 22.2140 60000 0.2194 12.8515 0.8896 0.8917 0.8902 0.9089 0.9110 0.9095 0.9116 0.9138 0.9122 0.9748 0.9780 0.9759 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 0.0 0.0 0.0
0.0006 25.9163 70000 0.2265 13.0870 0.8981 0.9013 0.8992 0.9191 0.9224 0.9203 0.9219 0.9249 0.9229 0.9756 0.9776 0.9761 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 0.0 0.0 0.0
0.0001 29.6187 80000 0.2249 13.0870 0.8938 0.8961 0.8945 0.9139 0.9163 0.9146 0.9169 0.9191 0.9175 0.9749 0.9764 0.9752 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 0.0 0.0 0.0
0.0001 33.3210 90000 0.2379 13.2342 0.8960 0.8987 0.8969 0.9160 0.9189 0.9169 0.9197 0.9224 0.9206 0.9759 0.9780 0.9764 1.0 1.0 0.9697 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
0.0 37.0233 100000 0.2302 13.1312 0.8910 0.8958 0.8930 0.9121 0.9171 0.9142 0.9149 0.9195 0.9167 0.9742 0.9786 0.9759 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 0.0 0.0 0.0
0.0004 40.7257 110000 0.2294 12.9987 0.9032 0.9028 0.9025 0.9220 0.9216 0.9213 0.9255 0.9249 0.9247 0.9762 0.9773 0.9763 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 0.0 0.0 0.0
0.0001 44.4280 120000 0.2322 12.6601 0.9038 0.9045 0.9037 0.9234 0.9242 0.9233 0.9262 0.9270 0.9262 0.9766 0.9784 0.9770 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 0.0 0.0 0.0
0.0 48.1303 130000 0.2362 12.5129 0.9054 0.9058 0.9051 0.9241 0.9247 0.9239 0.9277 0.9284 0.9276 0.9763 0.9777 0.9766 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 0.0 0.0 0.0
0.0 51.8327 140000 0.2430 13.1753 0.8973 0.8993 0.8978 0.9184 0.9205 0.9189 0.9216 0.9237 0.9221 0.9766 0.9783 0.9770 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 0.0 0.0 0.0
0.0005 55.5350 150000 0.2325 12.7926 0.9032 0.9032 0.9028 0.9226 0.9228 0.9223 0.9251 0.9252 0.9247 0.9781 0.9785 0.9779 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 0.0 0.0 0.0
0.0 59.2373 160000 0.2428 12.2332 0.9090 0.9104 0.9093 0.9275 0.9289 0.9278 0.9301 0.9315 0.9304 0.9773 0.9791 0.9778 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 0.0 0.0 0.0
0.0 62.9397 170000 0.2499 12.1301 0.9067 0.9081 0.9070 0.9246 0.9261 0.9249 0.9273 0.9286 0.9275 0.9775 0.9794 0.9780 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 0.0 0.0 0.0
0.0003 66.6420 180000 0.2572 12.2185 0.9050 0.9049 0.9045 0.9238 0.9238 0.9234 0.9265 0.9263 0.9260 0.9785 0.9784 0.9780 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 0.0 0.0 0.0
0.0 70.3443 190000 0.2704 12.1449 0.9058 0.9068 0.9059 0.9237 0.9247 0.9238 0.9263 0.9273 0.9264 0.9775 0.9787 0.9777 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 0.0 0.0 0.0
0.0 74.0466 200000 0.2759 12.0860 0.9045 0.9054 0.9045 0.9223 0.9233 0.9224 0.9250 0.9259 0.9250 0.9777 0.9785 0.9777 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 0.0 0.0 0.0

Framework versions

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