modelscope / modelscope/ms-swift
Support sequential data loading (disable shuffle) for Megatron training
Open
Nobody has claimed this yet.
enhancement
stale
- Dominant language
- Python
- Stars
- 15.7k
- Forks
- 1.7k
- Avg merge
- 1d 16h
- Merged PRs (30d)
- 136
Description
Describe the feature
目前在普通的训练模式下，支持让训练数据按顺序排列（不进行 shuffle，即--train_dataloader_shuffle false）。但是在使用 Megatron 框架进行训练时，该功能似乎无法生效，数据依然会被打乱。
希望能在 Megatron 模式下也支持 --train_dataloader_shuffle false 的功能，以保持和非 Megatron 模式一致的行为。
Paste any useful information
https://github.com/modelscope/ms-swift/issues/3977
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 Megatron training data-loader entry point and the handling of --train_dataloader_shuffle; compare it with ordinary training behavior and the linked issue #3977. Done means false preserves input order in Megatron mode and the relevant tests pass.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100