[None][fix] impl fused triton kernel for e8m0 resmooth (target release/1.2.0rc6.post1, cherry-pick from https://github.com/NVIDIA/TensorRT-LLM/pull/10327 and https://github.com/NVIDIA/TensorRT-LLM/pull/10770)#10771
Conversation
…ry footprint (NVIDIA#10327) Signed-off-by: Nekofish-L <liuxiangyang@mail.ustc.edu.cn> Co-authored-by: Kanghwan <861393+karljang@users.noreply.github.com> Signed-off-by: Yuxian Qiu <142763828+yuxianq@users.noreply.github.com>
Signed-off-by: Yuxian Qiu <142763828+yuxianq@users.noreply.github.com>
|
/bot run --disable-fail-fast |
📝 WalkthroughWalkthroughReplaces Python-based FP8 E4M3 per-block resmoothing logic with a Triton GPU kernel implementation. Updates Changes
Estimated code review effort🎯 4 (Complex) | ⏱️ ~40 minutes 🚥 Pre-merge checks | ✅ 1 | ❌ 2❌ Failed checks (2 warnings)
✅ Passed checks (1 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Fix all issues with AI agents
In `@tensorrt_llm/quantization/utils/fp8_utils.py`:
- Around line 109-133: Before launching _resmooth_kernel, validate that
weight_scale's last two dims match the expected grid tiles: compute M,K from
orig_shape, BLOCK_M,BLOCK_K from block_size and expected_tiles = (triton.cdiv(M,
BLOCK_M), triton.cdiv(K, BLOCK_K)); if weight_scale.shape[-2:] does not equal
expected_tiles, either raise a clear ValueError (including expected vs actual
and the symbols weight_scale, M, K, BLOCK_M, BLOCK_K) or reshape/pad
weight_scale to match before creating s_view so _resmooth_kernel's loads/stores
cannot go out-of-bounds.
- Around line 98-102: The function signature for resmooth_to_fp8_e8m0 currently
uses PEP 585 syntax "tuple[int, int]" which breaks on Python 3.8; update the
type annotation to use typing.Tuple (i.e., Tuple[int, int]) in the function
definition (resmooth_to_fp8_e8m0) so it is compatible with Python 3.8,
leveraging the existing Tuple import from typing.
🧹 Nitpick comments (1)
tensorrt_llm/quantization/utils/fp8_utils.py (1)
106-113: Avoid implicit device migration and ensure contiguity beforeview.
.cuda()moves tensors to the current device (may differ fromweight.device) and the subsequent.view(...)will fail on non-contiguous inputs. Consider preserving the original device and usingreshapeorcontiguous.♻️ Proposed adjustment
- weight = weight.cuda() - weight_scale = weight_scale.cuda() + device = weight.device if weight.is_cuda else torch.device("cuda") + weight = weight.to(device).contiguous() + weight_scale = weight_scale.to(device).contiguous() ... - w_view = weight.view(-1, M, K) - s_view = weight_scale.view(-1, weight_scale.shape[-2], - weight_scale.shape[-1]) + w_view = weight.reshape(-1, M, K) + s_view = weight_scale.reshape(-1, weight_scale.shape[-2], + weight_scale.shape[-1])
|
PR_Github #32394 [ run ] triggered by Bot. Commit: |
|
/bot run --disable-fail-fast --add-multi-gpu-test |
|
PR_Github #32400 [ run ] triggered by Bot. Commit: |
|
PR_Github #32400 [ run ] completed with state
|
|
/bot run --disable-fail-fast --add-multi-gpu-test |
|
PR_Github #32414 [ run ] triggered by Bot. Commit: |
|
PR_Github #32414 [ run ] completed with state |
|
/bot run --disable-fail-fast --add-multi-gpu-test |
|
PR_Github #32423 [ run ] triggered by Bot. Commit: |
|
PR_Github #32423 [ run ] completed with state
|
|
/bot run --disable-fail-fast --add-multi-gpu-test |
|
PR_Github #32441 [ run ] triggered by Bot. Commit: |
|
PR_Github #32441 [ run ] completed with state
|
|
/bot run --disable-fail-fast --add-multi-gpu-test |
|
PR_Github #32492 [ run ] triggered by Bot. Commit: |
|
/bot run --disable-fail-fast --only-multi-gpu-test |
|
PR_Github #32493 [ run ] triggered by Bot. Commit: |
|
/bot skip --comment "skip CI since this emergent fix has been validated locally" |
|
PR_Github #32505 [ skip ] triggered by Bot. Commit: |
|
PR_Github #32505 [ skip ] completed with state |
4 participants
Summary by CodeRabbit
✏️ Tip: You can customize this high-level summary in your review settings.
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)
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
/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...Provide a user friendly way for developers to interact with a Jenkins server.
Run
/bot [-h|--help]to print this help message.See details below for each supported subcommand.
Details
run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental)]Launch build/test pipelines. All previously running jobs will be killed.
--reuse-test (optional)pipeline-id(OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.--disable-reuse-test(OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.--disable-fail-fast(OPTIONAL) : Disable fail fast on build/tests/infra failures.--skip-test(OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.--stage-list "A10-PyTorch-1, xxx"(OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.--gpu-type "A30, H100_PCIe"(OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.--test-backend "pytorch, cpp"(OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.--only-multi-gpu-test(OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.--disable-multi-gpu-test(OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.--add-multi-gpu-test(OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.--post-merge(OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx"(OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".--detailed-log(OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.--debug(OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip testing for latest commit on pull request.
--comment "Reason for skipping build/test"is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.reuse-pipeline
reuse-pipelineReuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.