metadata
base_model: openai/whisper-base
datasets:
- Oyounghyun/whisper_study_data
language:
- ko
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
model-index:
- name: study0702
results: []
study0702
This model is a fine-tuned version of openai/whisper-base on the train dataset. It achieves the following results on the evaluation set:
- Loss: 0.1707
- Cer: 6.3423
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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- training_steps: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.2677 | 0.9615 | 100 | 0.2343 | 8.5421 |
| 0.1167 | 1.9231 | 200 | 0.1853 | 7.1773 |
| 0.0532 | 2.8846 | 300 | 0.1747 | 6.6635 |
| 0.0284 | 3.8462 | 400 | 0.1718 | 6.7919 |
| 0.0174 | 4.8077 | 500 | 0.1705 | 6.3423 |
| 0.0081 | 5.7692 | 600 | 0.1689 | 5.9891 |
| 0.0069 | 6.7308 | 700 | 0.1702 | 6.2781 |
| 0.0053 | 7.6923 | 800 | 0.1702 | 6.5511 |
| 0.0045 | 8.6538 | 900 | 0.1703 | 6.3263 |
| 0.0048 | 9.6154 | 1000 | 0.1707 | 6.3423 |
Framework versions
- Transformers 4.43.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1