jd-opensource / jd-opensource/JoyAI-VL-Interaction
about SFT
Open
- Dominant language
- Python
- Stars
- 1.9k
- Forks
- 190
- PR merge metrics
- No merged PRs in 30d
Description
在llamafactory上进行sft时，请问能提供交叉熵损失加权计算的实现吗
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with the LLaMA-Factory SFT entry point and locate the cross-entropy loss calculation. Clarify the weighting input and expected behavior before implementation; done should provide a working weighted cross-entropy path for SFT, but this issue names no file or test.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100