whisper-large-v2-zh-hk-2gpu
This model is a fine-tuned version of openai/whisper-medium on the mozilla-foundation/common_voice_11_0 zh-HK dataset. It achieves the following results on the evaluation set:
- Loss: 0.2237
- Wer: 0.4573
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: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 16
- total_eval_batch_size: 16
- 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 |
|---|---|---|---|---|
| 0.1544 | 1.14 | 1000 | 0.2260 | 0.5485 |
| 0.0745 | 2.28 | 2000 | 0.2132 | 0.4967 |
| 0.0213 | 3.42 | 3000 | 0.2114 | 0.4718 |
| 0.0117 | 4.57 | 4000 | 0.2196 | 0.4643 |
| 0.0014 | 5.71 | 5000 | 0.2237 | 0.4573 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1
- Datasets 2.8.0
- Tokenizers 0.13.2