cp: LLAMA3 70B: LoRa enabled in all modules instead of only LinearQKV (2181) into r0.3.0

[ko3n1g]

beep boop [🤖]: Hi @rhmukundan 👋,

we've cherry picked #2181 into  for you! 🚀

Please review and approve this cherry pick by your convenience!

Summary by CodeRabbit

• Chores
  • Adjusted LoRA configuration settings for model fine-tuning to target specific linear module types instead of all available modules across multiple GPU architecture configurations.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[ko3n1g]

/ok to test 03dd3c3

[coderabbitai]

📝 Walkthrough

Walkthrough

Three Llama3 70B LoRA configuration functions are updated to explicitly restrict LoRA target modules to linear QKV layers only, by adding a cfg.peft.target_modules = ["linear_qkv"] override after the base config application.

Changes

Cohort / File(s)	Summary
Llama3 LoRA Configuration scripts/performance/configs/llama/llama3_llm_finetune.py	Added cfg.peft.target_modules = ["linear_qkv"] override to three functions (llama3_70b_lora_config_gb300, llama3_70b_lora_config_gb200, llama3_70b_lora_config_h100) to restrict LoRA application scope to QKV-only modules.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• LLAMA3 70B: LoRa enabled in all modules instead of only LinearQKV #2181: Makes identical changes to the same three Llama3 70B LoRA configuration functions, restricting target_modules to ["linear_qkv"].

Suggested labels

r0.3.0

Suggested reviewers

• malay-nagda

🚥 Pre-merge checks | ✅ 3 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Test Results For Major Changes	⚠️ Warning	PR modifies LoRA target modules configuration which affects model convergence, but PR description lacks test results or performance metrics demonstrating no regression.	Add test results to PR description showing convergence metrics and performance on GB300, GB200, and H100 hardware to validate the configuration change.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title clearly describes the main change: enabling LoRA across all modules (instead of only LinearQKV) for LLAMA3 70B, which directly matches the commit message and file changes.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch cherry-pick-2181-r0.3.0

---

No actionable comments were generated in the recent review. 🎉

---

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

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[rhmukundan]

LGTM

[ko3n1g]

Tested and updated golden values