metadata
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: whisper-medium-zh-CN
results: []
whisper-medium-zh-CN
This model is a fine-tuned version of openai/whisper-medium on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2311
- Wer: 100.0
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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.4153 | 0.16 | 200 | 0.3310 | 100.0 |
| 0.3728 | 0.32 | 400 | 0.2904 | 100.0 |
| 0.3358 | 0.48 | 600 | 0.2695 | 100.0636 |
| 0.304 | 0.65 | 800 | 0.2435 | 100.0 |
| 0.2817 | 0.81 | 1000 | 0.2311 | 100.0 |
Framework versions
- Transformers 4.32.0.dev0
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3