update models shell #5299
update models shell #5299
Conversation
There was a problem hiding this comment.
Summary of Changes
Hello @Jintao-Huang, 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 primarily focuses on expanding the training capabilities for various models by introducing new shell scripts and updating existing ones. I've also improved the documentation to guide users on finding model-specific parameter definitions and performed a minor code refactoring to optimize imports.
Highlights
- New Model Training Scripts: I've added new training shell scripts for several models, including
deepseek_vl2,internvl3,ovis2, andphi4_multimodal. These scripts provide the necessary configurations and commands to train these models. - Documentation Updates: I've updated both the Chinese and English documentation files (
命令行参数.mdandCommand-line-parameters.md) to include a note clarifying where users can find the definitions for specific model parameters. - Training Script Adjustments: I've made minor adjustments to existing training scripts, such as removing the
MAX_PIXELSsetting fromminicpmv/train.shand increasing themax_lengthparameter inocr.shfrom 2048 to 4096. - Code Refactoring: I've refactored the
swift/trainers/mixin.pyfile by moving specific imports (DeepSpeedSchedulerWrapperandSaveStrategy) to be local within the_save_flash_checkpointfunction, which helps in better managing dependencies.
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 or fill out our survey to provide feedback.
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 adds several new training shell scripts, updates documentation, and refactors imports in swift/trainers/mixin.py. While the new scripts and documentation changes are good, the import refactoring in mixin.py introduces a critical issue. The local imports for DeepSpeedSchedulerWrapper and SaveStrategy are not compatible with newer versions of the transformers library, which are used in the new shell scripts. This will lead to an ImportError at runtime. I have provided a detailed comment with a suggested fix to ensure version compatibility.
| from transformers.trainer import DeepSpeedSchedulerWrapper | ||
| from transformers.trainer_utils import SaveStrategy |
There was a problem hiding this comment.
The local imports for DeepSpeedSchedulerWrapper and SaveStrategy will fail with newer versions of transformers (e.g., 4.51.* as used in the new shell scripts), as they have been removed. This will cause a runtime ImportError when use_flash_ckpt is enabled.
To ensure compatibility across transformers versions, these imports should be handled with try...except blocks. Additionally, the usage of DeepSpeedSchedulerWrapper needs to be guarded.
Here is the suggested change for the imports:
And at line 498, the check should be updated to:
is_deepspeed_custom_scheduler = (
self.is_deepspeed_enabled
and DeepSpeedSchedulerWrapper is not None
and not isinstance(self.lr_scheduler, DeepSpeedSchedulerWrapper)
)| from transformers.trainer import DeepSpeedSchedulerWrapper | |
| from transformers.trainer_utils import SaveStrategy | |
| try: | |
| from transformers.trainer import DeepSpeedSchedulerWrapper | |
| except ImportError: | |
| DeepSpeedSchedulerWrapper = None # For transformers>=4.43 | |
| try: | |
| from transformers.trainer_utils import SaveStrategy | |
| except ImportError: | |
| from transformers.trainer_utils import IntervalStrategy as SaveStrategy # For transformers>=4.42 |
No description provided.