【Hackathon 9th No.29】自定义算子 cutlass_fp8_fp8_half_block_gemm_fused 单测补充

[cloudforge1]

Motivation

Add unit tests for custom operator cutlass_fp8_fp8_half_block_gemm_fused to improve test coverage and prevent regressions.

Modifications

• Added operator unit test file: tests/operators/test_cutlass_fp8_fp8_half_block_gemm_fused.py
• Covered basic correctness, edge cases, and determinism
• Applied pre-commit formatting (black, isort, flake8, ruff)

Usage or Command

python -m pytest tests/operators/test_cutlass_fp8_fp8_half_block_gemm_fused.py -v

Accuracy Tests

Local verification (no GPU):

• py_compile syntax check: passes
• pre-commit (black/isort/flake8/ruff): passes

Tests call CUDA custom ops directly (SM80+ required). Full execution validated by CI run_tests_with_coverage job. Will request AI Studio access for on-device verification if needed.

Checklist

• Add at least a tag in the PR title.
• Format your code, run pre-commit before commit.
• Add unit tests.
• Provide accuracy results. N/A — unit test only.
• If the current PR is submitting to the release branch, cherry-pick from develop. N/A — targeting develop.

[paddle-bot]

Thanks for your contribution!

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (develop@30f9f33). Learn more about missing BASE report.

Additional details and impacted files

@@            Coverage Diff             @@
##             develop    #6693   +/-   ##
==========================================
  Coverage           ?   72.66%           
==========================================
  Files              ?      392           
  Lines              ?    53835           
  Branches           ?     8459           
==========================================
  Hits               ?    39117           
  Misses             ?    11932           
  Partials           ?     2786           

Flag	Coverage Δ
GPU	72.66% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[EmmonsCurse]

@cloudforge1 Please verify the changes locally before submitting to avoid unnecessary CI resource consumption.

@luotao1 cc

[cloudforge1]

@luotao1 Progress update on Hackathon 9th contributions:

9 PRs delivered across custom operator unit tests and log refactoring
(#6682, #6687, #6688, #6690, #6691, #6693, #6694, #6695, #6730):

• 7 custom op test suites: cutlass_fp8 block gemm, gptq_marlin_repack,
  moe_wna16_marlin_gemm, moe_expert_ffn_wint2, winx_unzip,
  multi_head_latent_attention, speculate_set_value
• 1 log refactoring PR (task 88)
• 1 Hackathon 10th config.py test suite

CI triage completed — all current failures are infrastructure-side
(Iluvatar Docker container crashes, XPU OpenAI 500 errors, Approval
gates pending RD assignment). Zero test code failures across all PRs.

We've ramped up quickly on the FastDeploy CI/CD pipeline and codebase
conventions. Going forward we're tightening our pre-push validation
including local tests to match the project's contribution cadence.

Also identified CI/CD optimization opportunities while working across
these PRs — details in PaddlePaddle/Paddle#78233.

Looking forward to RD review assignments on the ready PRs.