MFTCoder: Boosting Code LLMs with Multitask Fine-Tuning.
Bingchang LiuChaoyu ChenZi GongCong LiaoHuan WangZhichao LeiMing LiangDajun ChenMin ShenHailian ZhouWei JiangHang YuJianguo LiPublished in: KDD (2024)
Keyphrases
- fine tuning
- multi task
- multi task learning
- multitask learning
- feature selection
- binary classification problems
- viable alternative
- fine tune
- learning tasks
- multi class
- base learners
- multiple tasks
- transfer learning
- learning problems
- fine tuned
- gaussian processes
- maximum margin
- multiclass support vector machines
- ensemble methods
- cost sensitive
- text classification
- information gain
- machine learning
- text categorization
- labeled data
- active learning
- training data
- learning algorithm