Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @chenyushuo, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request addresses a bug related to FSDP (Fully Sharded Data Parallel) model offloading. It ensures that the FSDP model is correctly loaded to the GPU before operations that require it, such as weight synchronization and state dict uploads. After these operations, the model is then offloaded back to the CPU, preventing potential issues and optimizing memory usage when offloading is active.
Highlights
- Ensured FSDP models are loaded to GPU before critical operations: The
actor_module_fsdpis now explicitly loaded to GPU beforesetup_weight_sync_group,sync_weight, andupload_state_dictroutines if parameter offloading is active. - Implemented consistent model offloading: After these operations complete, the FSDP model is offloaded back to the CPU to free up GPU memory, specifically for actor modules when offloading is enabled.
- Added synchronization and cache clearing:
torch.distributed.barrier()andtorch.cuda.empty_cache()calls are strategically placed after offloading operations to ensure proper distributed synchronization and GPU memory management.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
There was a problem hiding this comment.
Code Review
This pull request addresses a bug in FSDP offloading by ensuring the model is loaded onto the GPU before operations that require it and offloaded back to the CPU afterward to conserve GPU memory. The changes are applied to setup_weight_sync_group, sync_weight, and upload_state_dict methods. The logic appears correct. My main feedback is to refactor the repeated code for loading and offloading into a context manager to improve code maintainability and readability.
|
/unittest-module-trainer |
Summary
Failed Tests
Tests
Github Test Reporter by CTRF 💚 |
|
/unittest-module-trainer |
Summary
Tests
Github Test Reporter by CTRF 💚 |
|
/unittest-module-trainer |
Summary
Tests
Github Test Reporter by CTRF 💚 |
2 participants
Description
As the title says.
Checklist
Please check the following items before code is ready to be reviewed.