fix: return proper attention for llama4 lora kernel and fsdp2 llama4 example fix

[NanoCode012]

Description

• fix llama4 attention lora kernels
• fix llama4 fsdp2 example cpu ram efficient qlora

Motivation and Context

How has this been tested?

Screenshots (if appropriate)

Types of changes

Social Handles (Optional)

Summary by CodeRabbit

• Bug Fixes

  • Disabled an incompatible configuration option for CPU RAM efficient loading when using 8-bit or 4-bit loading modes.

• New Features

  • Added support for a new attention mechanism specific to "llama4" model types.

[coderabbitai]

Walkthrough

A new conditional branch for the "llama4" model type was added to the get_attention_cls_from_config function to directly import and return the Llama4TextAttention class. Additionally, a configuration option for FSDP CPU RAM efficient loading was commented out and annotated as incompatible with 8-bit/4-bit loading.

Changes

File(s)	Change Summary
src/axolotl/monkeypatch/lora_kernels.py	Added a conditional branch to handle "llama4" model type in get_attention_cls_from_config, importing Llama4TextAttention.
examples/llama-4/scout-qlora-flexattn-fsdp2.yaml	Commented out fsdp_cpu_ram_efficient_loading: true and annotated as incompatible with 8bit/4bit loading.

Possibly related PRs

• Fix: adding magistral fsdp config, fixing not eval with test_datasets, handle mllama attention #2789: Adds support for "mllama" attention in the same function, modifying similar logic for different model types.
• misc fixes 202507 #2937: Also adds support for the "llama4" model type in get_attention_cls_from_config with similar import logic.

Suggested labels

ready to merge

Suggested reviewers

• winglian
• SalmanMohammadi

Poem

In the land of code, a llama appears,
With new attention, it now steers.
RAM settings tweaked, a note in the air,
Eight-bit and four-bit—handle with care!
Hopping ahead, the rabbit is keen,
On fresh model magic, swift and clean.
🐇✨

---

📜 Recent review details

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 170322a and e78519c.

📒 Files selected for processing (2)

• examples/llama-4/scout-qlora-flexattn-fsdp2.yaml (1 hunks)
• src/axolotl/monkeypatch/lora_kernels.py (1 hunks)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (6)

• GitHub Check: PyTest from Source Dist (3.11, 2.7.0)
• GitHub Check: PyTest (3.11, 2.7.0)
• GitHub Check: PyTest from Source Dist (3.11, 2.6.0)
• GitHub Check: PyTest (3.11, 2.6.0)
• GitHub Check: PyTest from Source Dist (3.11, 2.7.1)
• GitHub Check: PyTest (3.11, 2.7.1)

🔇 Additional comments (2)

examples/llama-4/scout-qlora-flexattn-fsdp2.yaml (1)

77-77: LGTM: Proper handling of incompatible FSDP settings.

The change correctly disables fsdp_cpu_ram_efficient_loading when using quantized loading (load_in_4bit: true). The comment clearly explains the incompatibility, preventing potential runtime errors.

src/axolotl/monkeypatch/lora_kernels.py (1)

154-157: Verify llama4 support in your Transformers dependency
The sandbox couldn’t import transformers, so please confirm that your project’s pinned transformers version actually includes the llama4 module and the Llama4TextAttention class. In your local environment, run:

python -c "from transformers.models.llama4.modeling_llama4 import Llama4TextAttention"

and ensure your requirements.txt or pyproject.toml specifies a release of transformers that provides transformers.models.llama4.modeling_llama4.

✨ Finishing Touches

• 📝 Generate Docstrings

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

• Review comments: Directly reply to a review comment made by CodeRabbit. Example:
  • I pushed a fix in commit <commit_id>, please review it.
  • Explain this complex logic.
  • Open a follow-up GitHub issue for this discussion.
• Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
  • @coderabbitai explain this code block.
  • @coderabbitai modularize this function.
• PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
  • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
  • @coderabbitai read src/utils.ts and explain its main purpose.
  • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
  • @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

• @coderabbitai pause to pause the reviews on a PR.
• @coderabbitai resume to resume the paused reviews.
• @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
• @coderabbitai full review to do a full review from scratch and review all the files again.
• @coderabbitai summary to regenerate the summary of the PR.
• @coderabbitai generate docstrings to generate docstrings for this PR.
• @coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
• @coderabbitai resolve resolve all the CodeRabbit review comments.
• @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
• @coderabbitai help to get help.

Other keywords and placeholders

• Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
• Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
• Add @coderabbitai anywhere in the PR title to generate the title automatically.

CodeRabbit Configuration File (.coderabbit.yaml)

• You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
• Please see the configuration documentation for more information.
• If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

• Visit our Documentation for detailed information on how to use CodeRabbit.
• Join our Discord Community to get help, request features, and share feedback.
• Follow us on X/Twitter for updates and announcements.

[codecov]

Codecov Report

Attention: Patch coverage is 33.33333% with 2 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/axolotl/monkeypatch/lora_kernels.py	33.33%	2 Missing ⚠️

📢 Thoughts on this report? Let us know!