metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
base_model: openai/whisper-large-v3
model-index:
- name: whisper_final
results: []
๊ตฌ์์ฅ์ ํ์๋ฅผ ์ํ ์์ฑ์ธ์ ๋ชจ๋ธ
ํ๋ก์ ํธ ์ ๋ณด
์ฌ๋จ๋ฒ์ธ ๋ฏธ๋์ ์ํํธ์จ์ด์ ํจ๊ปํ๋ ์ 3ํ์์ด๋์ด ๊ณต๋ชจ์
ํ๋ก์ ํธ ๋ช
"๊ตฌ์์ฅ์ ์์ฑ ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ ๊ณ ๋ น ํ์์ ์์ฌ์ํต ๊ฐ์ ๋ฐฉ์"
๋ชจ๋ธ ์ค๋ช
- openai/whisper-large-v3์ ๋ํ ํ์ธํ๋ ๋ชจ๋ธ
- ๋ณธ ๋ชจ๋ธ์ "๊ตฌ์์ฅ์ ์์ฑ ๋ฐ์ดํฐ๋ฅผ ํ์ฉํ ๊ณ ๋ น ํ์์ ์์ฌ์ํต ๊ฐ์ ๋ฐฉ์" ํ๋ก์ ํธ์ ๊ตฌ์์ฅ์ ํ์๋ค์ ๋ํ ํ๊ตญ์ด ์์ฑ์ธ์ ๋ชจ๋ธ์. OpenAI์ Whisper ๋ชจ๋ธ์ ํ์ธํ๋ ํ์ฌ ๊ตฌ์์ฅ์ ์ ์์ฑ์ ํน์ฑ์ ๋ฐ์ํ ๋ชจ๋ธ์ ๊ตฌ์ถํ์์.
- ์ค๋ฅธ์ชฝ "Inference API"๋ฅผ ํตํด ์์ฑ์ธ์ ๋ชจ๋ธ์ ํ ์คํธ ํด๋ณผ ์ ์์ต๋๋ค.
ํ์ต ๋ชจ๋ธ
- Paper: Radford, A., Kim, J. W., Xu, T., Brockman, G., McLeavey, C., & Sutskever, I. (2023, July). Robust speech recognition via large-scale weak supervision. In International Conference on Machine Learning (pp. 28492-28518). PMLR.
- URL: https://proceedings.mlr.press/v202/radford23a.html
ํ์ต ๋ฐ์ดํฐ
- AIHub "๊ตฌ์์ฅ์ ์์ฑ ๋ฐ์ดํฐ" (KOR)
- URL: https://aihub.or.kr/aihubdata/data/view.do?currMenu=115&topMenu=100&aihubDataSe=data&dataSetSn=608
ํ์ต ํ๋ผ๋ฏธํฐ
- learning_rate: 5e-07
- train_batch_size: 8
- 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
- mixed_precision_training: Native AMP
ํ์ต ๊ฒฐ๊ณผ
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
4.2932 | 0.09 | 10 | 4.6306 | 16.0442 |
4.2744 | 0.18 | 20 | 4.1942 | 16.2348 |
3.7418 | 0.27 | 30 | 3.7625 | 15.5107 |
3.2037 | 0.36 | 40 | 3.5635 | 14.6723 |
3.4714 | 0.45 | 50 | 3.4383 | 14.3674 |
2.8962 | 0.55 | 60 | 3.3494 | 14.1768 |
2.7958 | 0.64 | 70 | 3.2752 | 18.2927 |
2.8691 | 0.73 | 80 | 3.2208 | 19.5884 |
2.8693 | 0.82 | 90 | 3.1857 | 20.6174 |
2.9474 | 0.91 | 100 | 3.1644 | 20.6555 |
3.1712 | 1.0 | 110 | 3.1551 | 20.6174 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1