ianzh0/whisper-small-chinese_base-onnx
By ianzh0 · ↓ 2 · ♥ 1 · Updated 2024-06-27
What's in this repo
This page lists the files in ianzh0/whisper-small-chinese_base-onnx, including decoder_model.onnx, decoder_with_past_model.onnx, encoder_model.onnx.
- 3 ONNX models (.onnx files)
- 30 files, including 9 .json, 5 .1, 5 .0, 3 .onnx, 3 .2, 1 .md
- Last updated 2024-06-27
ONNX files
| File | Size | Download |
|---|---|---|
| onnx/decoder_model.onnx | — | Download |
| onnx/decoder_with_past_model.onnx | — | Download |
| onnx/encoder_model.onnx | — | Download |
How to use
These ONNX files can be loaded with ONNX Runtime. To inspect a model graph, upload the file to the Netron viewer. Download links point to the original files on Hugging Face.
README
Whisper Small Chinese Base
This model is a fine-tuned version of openai/whisper-small on the google/fleurs cmn_hans_cn dataset. It achieves the following results on the evaluation set:
- Loss: 0.3573
- Wer: 16.6439
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: 64
- 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: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0005 | 76.0 | 1000 | 0.3573 | 16.6439 |
| 0.0002 | 153.0 | 2000 | 0.3897 | 16.9749 |
| 0.0001 | 230.0 | 3000 | 0.4125 | 17.2330 |
| 0.0001 | 307.0 | 4000 | 0.4256 | 17.2451 |
| 0.0001 | 384.0 | 5000 | 0.4330 | 17.2300 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2
Model info
Created: 2024-06-27
Updated: 2024-06-27
View on Hugging FaceFiles (30)
Related models
More from ianzh0
No other models by this author.