Downloads · 30 days
5
17% of all-time downloads
whisper-small-ne
saurastha/whisper-small-ne
whisper-small-ne is a automatic speech recognition model from saurastha. Use it when you need speech turned into text. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
29
Public
Repo size
1.9 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin967 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- Model type
- whisper
- Access
- Public
- Created
- Oct 9, 2023
- Updated
- Oct 9, 2023
- SHA
- 6a4c7f7f
Model card
From the Hugging Face model README
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
model
This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6003
- Wer Ortho: 72.9876
- Wer: 49.5585
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 5
- training_steps: 100
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Wer Ortho |
|---|---|---|---|---|---|
| 1.5813 | 0.04 | 5 | 1.2523 | 73.4580 | 94.3667 |
| 0.9924 | 0.08 | 10 | 1.0420 | 76.0991 | 94.0559 |
| 0.8522 | 0.12 | 15 | 0.9581 | 71.8012 | 91.8415 |
| 0.7602 | 0.17 | 20 | 0.9154 | 64.7802 | 88.3450 |
| 0.6813 | 0.21 | 25 | 0.8780 | 65.9285 | 87.9953 |
| 0.7084 | 0.25 | 30 | 0.8520 | 63.5171 | 86.4413 |
| 0.5974 | 0.29 | 35 | 0.8120 | 62.9593 | 85.5866 |
| 0.5631 | 0.33 | 40 | 0.7904 | 59.9409 | 84.1492 |
| 0.6962 | 0.38 | 45 | 0.7631 | 58.0709 | 82.6340 |
| 0.4946 | 0.42 | 50 | 0.7486 | 60.7448 | 83.1779 |
| 0.6118 | 0.46 | 55 | 0.7216 | 59.8097 | 84.4600 |
| 0.5004 | 0.5 | 60 | 0.7018 | 58.8255 | 82.4786 |
| 0.4357 | 0.54 | 65 | 0.6932 | 57.4311 | 81.2743 |
| 0.4478 | 0.58 | 70 | 0.6839 | 55.8399 | 80.8469 |
| 0.4561 | 0.62 | 75 | 0.6856 | 59.3012 | 80.8081 |
| 0.429 | 0.67 | 80 | 0.6646 | 55.0853 | 79.9145 |
| 0.4072 | 0.71 | 85 | 0.6601 | 52.2638 | 76.0684 |
| 0.4096 | 0.75 | 90 | 0.6477 | 52.3950 | 76.8842 |
| 0.4471 | 0.79 | 95 | 0.6393 | 51.2795 | 76.1072 |
| 0.3282 | 0.83 | 100 | 0.6003 | 72.9876 | 49.5585 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- Created
- Oct 9, 2023
- Updated
- Oct 9, 2023