modelscope / modelscope/ms-swift
如何自定义Trainer，是否有Trainer的相关文档
Open
Nobody has claimed this yet.
stale
- Dominant language
- Python
- Stars
- 15.7k
- Forks
- 1.7k
- Avg merge
- 1d 16h
- Merged PRs (30d)
- 136
Description
想自定义trainer支持grad_cache, 但是理不顺trainer的逻辑。
比如如何使用model前向推理得到结果，如何复用代码中的embedding推理逻辑
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating the Trainer implementation and the existing embedding inference path in the repository. Read how model forward inference is invoked and how that logic is reused; done means documenting the relevant flow and the supported extension points for a custom Trainer with grad_cache.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Documentation
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100