4. Transformer Fine Tuning
2024. 1. 4. 15:12ㆍNAVER AI Tech/Deep Knowledge Tracing
Hyperopt, Ray 등 라이브러리를 활용해서 hyperparameter tuning을 진행해보자.
Transformer의 hyperparameter
1. layer 수
2. head 수
3. embedding dim
4. sequence length
Training 시 hyperparameter
1. learning rate
2. batch size
3. epochs
'NAVER AI Tech > Deep Knowledge Tracing' 카테고리의 다른 글
| 6. Matrix Factorization(SVD) for Recommend System. (0) | 2024.01.05 |
|---|---|
| 5. Graph Neural Network for Recommend System (2) | 2024.01.05 |
| 3. (Kaggle) Riiid Competition Winner's Solution (0) | 2024.01.04 |
| 2. How to make sequence data (1) | 2024.01.03 |
| 1. Deep Knowledge Tracing (0) | 2024.01.03 |