[https://nvbugs/6396422][fix] Disable kv-cache reuse for MiniMax-M2

[tcherckez-nvidia]

Summary by CodeRabbit

• Tests
  • Updated MiniMax M2 validation to run with block reuse disabled, improving test reliability and consistency.
  • Re-enabled fine-grained FP8 accuracy coverage by removing its test waiver.

Description

Test Coverage

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: c493ff48-0c67-401f-a1f3-314828540476

📥 Commits

Reviewing files that changed from the base of the PR and between 725d9a1 and b5a2a38.

📒 Files selected for processing (2)

• tests/integration/defs/accuracy/test_llm_api_autodeploy.py
• tests/integration/test_lists/waives.txt

💤 Files with no reviewable changes (1)

• tests/integration/test_lists/waives.txt

---

📝 Walkthrough

Walkthrough

The MiniMaxM2 accuracy test now disables KV-cache block reuse by default, and its fine-grained FP8 test waiver is removed.

Changes

MiniMaxM2 accuracy configuration

Layer / File(s)	Summary
Configure block reuse and remove waiver tests/integration/defs/accuracy/test_llm_api_autodeploy.py, tests/integration/test_lists/waives.txt	Sets kv_cache_config.enable_block_reuse to False for MiniMaxM2 and removes the corresponding test_finegrained_fp8 waiver.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#16367: Removes waivers for other FP8 AutoDeploy accuracy tests and also changes KV-cache configuration.
• NVIDIA/TensorRT-LLM#16271: Modifies related accuracy-test waiver coverage.
• NVIDIA/TensorRT-LLM#16295: Adjusts skipped integration-test entries in the same waiver file.

Suggested reviewers: qijune, chienchunhung, bmarimuthu-nv

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description is mostly template text and lacks a filled-in issue description and test coverage details.	Add a short Description section explaining the problem and fix, plus a Test Coverage section listing relevant tests or verification.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title matches the main change and includes the NVBugs ID and fix type in the expected format.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[tcherckez-nvidia]

/bot run --extra-stage "DGX_H100-4_GPUs-AutoDeploy-1"

[tensorrt-cicd]

PR_Github #59249 [ run ] triggered by Bot. Commit: f43a022 Link to invocation

[tensorrt-cicd]

PR_Github #59249 [ run ] completed with state FAILURE. Commit: f43a022
/LLM/main/L0_MergeRequest_PR pipeline #47738 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[tcherckez-nvidia]

/bot run --stage-list "DGX_H100-4_GPUs-AutoDeploy-1"

[tensorrt-cicd]

PR_Github #59363 [ run ] triggered by Bot. Commit: f43a022 Link to invocation

[tensorrt-cicd]

PR_Github #59363 [ run ] completed with state SUCCESS. Commit: f43a022
/LLM/main/L0_MergeRequest_PR pipeline #47839 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[tcherckez-nvidia]

run

[tcherckez-nvidia]

/bot run

[tensorrt-cicd]

PR_Github #59390 [ run ] triggered by Bot. Commit: f43a022 Link to invocation

[tensorrt-cicd]

PR_Github #59390 [ run ] completed with state SUCCESS. Commit: f43a022
/LLM/main/L0_MergeRequest_PR pipeline #47864 completed with status: 'SUCCESS'

CI Report

Link to invocation