Split the eagle test into two files #7170
Conversation
There was a problem hiding this comment.
Summary of Changes
Hello @merrymercy, 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 refactors the EAGLE inference tests by splitting a single large test file into two smaller, more focused files. This change primarily affects the test suite structure and the organization of the test classes, aiming to improve maintainability without altering the test logic itself.
Highlights
- Test File Splitting: The large
test_eagle_infer.pyfile has been split into two smaller files:test_eagle_infer_a.pyandtest_eagle_infer_b.py. This improves test organization and potentially makes the files easier to manage. - Test Suite Update: The main test suite runner (
run_suite.py) has been updated to include the two new test files (test_eagle_infer_a.pyandtest_eagle_infer_b.py) and remove the reference to the originaltest_eagle_infer.py. - Test Class Distribution: The
TestEAGLEEngine,TestEAGLEEngineTokenMap,TestEAGLE3Engine, andTestEAGLEDraftExtendtest classes (along with their variations for different attention backends) have been moved to the newtest_eagle_infer_a.pyfile. - Remaining Tests: The
TestEAGLEServertest class and its numerous variations (for streaming, batching, logprobs, correctness, and combinations thereof) remain in thetest_eagle_infer_b.pyfile.
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 is currently in preview and 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 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 configureGemini 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 splits the test_eagle_infer.py test file into two separate files, test_eagle_infer_a.py and test_eagle_infer_b.py, presumably for better organization and management of tests. The changes in test/srt/run_suite.py correctly reflect this split.
My review focuses on the newly created test_eagle_infer_a.py file. I've identified a couple of minor issues related to Python coding conventions (PEP 8) for constants and an unused variable. These are aimed at improving code clarity and maintainability.
Style Guide:
As no specific style guide was provided, I've referred to PEP 8 for Python, particularly for constant naming conventions.
| torch_dtype = torch.float16 | ||
| prefill_tolerance = 5e-2 | ||
| decode_tolerance: float = 5e-2 |
There was a problem hiding this comment.
Module-level constants should be named in UPPER_CASE_WITH_UNDERSCORES as per PEP 8 conventions for better readability and to clearly distinguish them from variables.1
Style Guide References
| torch_dtype = torch.float16 | |
| prefill_tolerance = 5e-2 | |
| decode_tolerance: float = 5e-2 | |
| TORCH_DTYPE = torch.float16 | |
| PREFILL_TOLERANCE = 5e-2 | |
| DECODE_TOLERANCE: float = 5e-2 |
Footnotes
-
PEP 8 recommends that constants are usually defined on a module level and written in all capital letters with underscores separating words. Examples include
MAX_OVERFLOWandTOTAL. ↩
| speed = ( | ||
| output["meta_info"]["completion_tokens"] | ||
| / output["meta_info"]["e2e_latency"] | ||
| ) |
There was a problem hiding this comment.
The speed variable is calculated but not used anywhere in this method or class. Consider removing it if it's not intended for future use or debugging, to avoid confusion and dead code.
No description provided.