metadata
library_name: transformers
language:
- en
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- Chinese_english_4000_106
metrics:
- wer
model-index:
- name: Chinese_english_4000_106
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Chinese English
type: Chinese_english_4000_106
args: 'config: default, split: test'
metrics:
- name: Wer
type: wer
value: 24.70214455917395
Chinese_english_4000_106
This model is a fine-tuned version of openai/whisper-tiny on the Chinese English dataset. It achieves the following results on the evaluation set:
- Loss: 0.8355
- Wer: 24.7021
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-06
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0002 | 1.2837 | 1000 | 0.8259 | 24.7021 |
| 0.0002 | 2.5674 | 2000 | 0.8297 | 24.5433 |
| 0.0002 | 3.8511 | 3000 | 0.8339 | 24.6545 |
| 0.0002 | 5.1348 | 4000 | 0.8355 | 24.7021 |
Framework versions
- Transformers 4.49.0
- Pytorch 2.6.0+cu124
- Datasets 3.4.1
- Tokenizers 0.21.1