Downloads · 30 days
4
7% of all-time downloads
whisper-tiny-zh_CN_lr4_b16
VingeNie/whisper-tiny-zh_CN_lr4_b16
whisper-tiny-zh_CN_lr4_b16 is a automatic speech recognition model from VingeNie. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
58
Public
Parameters
37.8M
6.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors151 MB · 99%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- whisper
- Access
- Public
- Created
- May 2, 2024
- Updated
- May 3, 2024
- SHA
- 39273e58
Base models
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. -->
Whisper Tiny chinese - VingeNie
This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 16.1 dataset. It achieves the following results on the evaluation set:
- Loss: 0.6244
- Cer Ortho: 28.9724
- Cer: 24.2287
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: 0.0001
- train_batch_size: 32
- 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: 25
- training_steps: 6000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer Ortho | Cer |
|---|---|---|---|---|---|
| 0.783 | 0.4796 | 600 | 0.7962 | 54.2372 | 31.6835 |
| 0.6578 | 0.9592 | 1200 | 0.6968 | 32.3398 | 27.9872 |
| 0.5028 | 1.4388 | 1800 | 0.6624 | 40.8123 | 27.6832 |
| 0.4887 | 1.9185 | 2400 | 0.6265 | 31.8488 | 25.8712 |
| 0.3368 | 2.3981 | 3000 | 0.6200 | 31.5221 | 25.3234 |
| 0.3395 | 2.8777 | 3600 | 0.6106 | 38.9218 | 25.3534 |
| 0.2078 | 3.3573 | 4200 | 0.6184 | 28.7037 | 24.7440 |
| 0.1943 | 3.8369 | 4800 | 0.6139 | 32.2586 | 24.3287 |
| 0.1206 | 4.3165 | 5400 | 0.6272 | 30.4763 | 24.2656 |
| 0.102 | 4.7962 | 6000 | 0.6244 | 28.9724 | 24.2287 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.0.1+cu118
- Datasets 2.19.0
- Tokenizers 0.19.1
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- License
- apache-2.0
- Languages
- zh
- Created
- May 2, 2024
- Updated
- May 3, 2024