Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Remove hardcoded DEBUG logging level in gpt_oss.py#15236

Merged
blisc merged 2 commits into
NVIDIA-NeMo:mainfrom
yurekami:fix/debug-logging
Jan 13, 2026
Merged

Remove hardcoded DEBUG logging level in gpt_oss.py#15236
blisc merged 2 commits into
NVIDIA-NeMo:mainfrom
yurekami:fix/debug-logging

Conversation

@yurekami
Copy link
Copy Markdown
Contributor

@yurekami yurekami commented Dec 29, 2025

Summary

Fixes #15116

Removes the hardcoded logging.getLogger("nemo").setLevel(logging.DEBUG) line that was forcing DEBUG level logging regardless of user configuration.

This line overrode any user-configured logging level, making it impossible to reduce log verbosity.

Test plan

  • Verified syntax check passes
  • Logging level now respects user configuration

The HFGPTOSSImporter.apply() method had a leftover debug statement
`logging.setLevel(logging.DEBUG)` that forced verbose logging globally
whenever importing GPT-OSS models from HuggingFace.

This caused:
- Excessive logs in production environments
- Performance degradation from logging overhead
- Loss of control over logging configuration

The fix removes this hardcoded statement, allowing the logging level
to be determined by the application's configuration.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <[email protected]>
Signed-off-by: yurekami <[email protected]>
@blisc
Copy link
Copy Markdown
Collaborator

blisc commented Jan 13, 2026

Thanks for your contribution!

@blisc blisc merged commit 4da048c into NVIDIA-NeMo:main Jan 13, 2026
192 checks passed
AkCodes23 pushed a commit to AkCodes23/NeMo that referenced this pull request Jan 28, 2026
NVIDIA-NeMo#15236)

The HFGPTOSSImporter.apply() method had a leftover debug statement
`logging.setLevel(logging.DEBUG)` that forced verbose logging globally
whenever importing GPT-OSS models from HuggingFace.

This caused:
- Excessive logs in production environments
- Performance degradation from logging overhead
- Loss of control over logging configuration

The fix removes this hardcoded statement, allowing the logging level
to be determined by the application's configuration.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Signed-off-by: yurekami <[email protected]>
Co-authored-by: yurekami <[email protected]>
Co-authored-by: Claude Opus 4.5 <[email protected]>
Co-authored-by: Jason <[email protected]>
Signed-off-by: Akhil Varanasi <[email protected]>
nune-tadevosyan pushed a commit to nune-tadevosyan/NeMo that referenced this pull request Mar 13, 2026
NVIDIA-NeMo#15236)

The HFGPTOSSImporter.apply() method had a leftover debug statement
`logging.setLevel(logging.DEBUG)` that forced verbose logging globally
whenever importing GPT-OSS models from HuggingFace.

This caused:
- Excessive logs in production environments
- Performance degradation from logging overhead
- Loss of control over logging configuration

The fix removes this hardcoded statement, allowing the logging level
to be determined by the application's configuration.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Signed-off-by: yurekami <[email protected]>
Co-authored-by: yurekami <[email protected]>
Co-authored-by: Claude Opus 4.5 <[email protected]>
Co-authored-by: Jason <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Remove hardcoded DEBUG logging level in gpt_oss.py

3 participants