feat: add support for Modelopt fp8_pb_wo quantization scheme #6106

achartier · 2025-07-16T22:32:13Z

PR title

feat: add support for Modelopt fp8_pb_wo quantization scheme

Description

Map fp8_pb_wo to the existing FP8_BLOCK_SCALES implementation so it can re-use the existing DS implementation
Squeeze weight tensors so they have the expected shape

Test Coverage

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.

run [--disable-fail-fast --skip-test --stage-list "A10-1, xxx" --gpu-type "A30, H100_PCIe" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-[Post-Merge]-1, xxx"]

Launch build/test pipelines. All previously running jobs will be killed.

--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 "A10-1, xxx" (OPTIONAL) : Only run the specified test stages. Examples: "A10-1, 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.

--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. Will also run 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-[Post-Merge]-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-[Post-Merge]-1, xxx".

For guidance on mapping tests to stage names, see docs/source/reference/ci-overview.md.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip 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-pipeline

Reuse 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.

Summary by CodeRabbit

Bug Fixes
- Improved handling of quantization algorithm identifiers from configuration files, ensuring consistent recognition of specific aliases.
- Adjusted tensor shape processing for weight scale parameters to prevent issues with singleton dimensions during model loading.

Signed-off-by: Aurelien Chartier <[email protected]>

coderabbitai · 2025-07-16T22:32:20Z

"""

Walkthrough

The changes update the handling of the quantization algorithm identifier in configuration files. Specifically, they introduce a mapping so that when the string "fp8_pb_wo" is encountered, it is replaced with "FP8_BLOCK_SCALES" in both model and utility code. Additionally, tensor shape handling is improved by squeezing singleton dimensions before copying weight scale tensors in the linear module.

Changes

File(s)	Change Summary
tensorrt_llm/_torch/model_config.py	Added conditional logic to map "fp8_pb_wo" to "FP8_BLOCK_SCALES" in quantization config loading.
tensorrt_llm/llmapi/llm_utils.py	Updated quant_algo parsing to map "fp8_pb_wo" to QuantAlgo.FP8_BLOCK_SCALES when reading HF quant config.
tensorrt_llm/_torch/modules/linear.py	Added `.squeeze()` to loaded weight scale tensors before copying into module parameters in FP8BlockScalesLinearMethod.

Poem

In configs and tensors, a subtle tune,
"fp8_pb_wo" now sings a new monsoon.
Scales are squeezed, shapes align,
Quantization flows just fine.
A rabbit hops with cheerful glee,
For tidy code and clarity! 🐇✨
"""

📜 Recent review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 327d42d and 58c5594.

📒 Files selected for processing (1)

tensorrt_llm/_torch/model_config.py (1 hunks)

🚧 Files skipped from review as they are similar to previous changes (1)

tensorrt_llm/_torch/model_config.py

✨ Finishing Touches

📝 Generate Docstrings

🪧 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.

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.

achartier · 2025-07-16T22:38:05Z

/bot run

tensorrt-cicd · 2025-07-16T22:44:36Z

PR_Github #12121 [ run ] triggered by Bot

tensorrt-cicd · 2025-07-17T03:26:47Z

PR_Github #12121 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #9006 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

tensorrt_llm/_torch/model_config.py

Signed-off-by: Aurelien Chartier <[email protected]>

achartier · 2025-07-18T02:15:21Z

/bot reuse-pipeline

tensorrt-cicd · 2025-07-18T02:20:17Z

PR_Github #12256 [ reuse-pipeline ] triggered by Bot

tensorrt-cicd · 2025-07-18T02:34:59Z

PR_Github #12256 [ reuse-pipeline ] completed with state SUCCESS
Reusing PR_Github #12121 for commit aa41757

…6106) Signed-off-by: Aurelien Chartier <[email protected]> Co-authored-by: Haohang Huang <[email protected]>

…6106) Signed-off-by: Aurelien Chartier <[email protected]> Co-authored-by: Haohang Huang <[email protected]> Signed-off-by: Shreyas Misra <[email protected]>

feat: add support for Modelopt fp8_pb_wo quantization scheme

327d42d

Signed-off-by: Aurelien Chartier <[email protected]>

achartier requested a review from symphonylyh July 16, 2025 22:32

achartier requested review from a team as code owners July 16, 2025 22:32

achartier requested a review from juney-nvidia July 16, 2025 22:32

symphonylyh approved these changes Jul 16, 2025

View reviewed changes

Merge branch 'main' into fp8_pb_wo

cef7781

QiJune reviewed Jul 18, 2025

View reviewed changes

tensorrt_llm/_torch/model_config.py Show resolved Hide resolved

QiJune approved these changes Jul 18, 2025

View reviewed changes

achartier added 2 commits July 18, 2025 02:10

address review

58c5594

Signed-off-by: Aurelien Chartier <[email protected]>

Merge branch 'main' into fp8_pb_wo

aa41757

achartier enabled auto-merge (squash) July 18, 2025 02:15

achartier merged commit 812243b into NVIDIA:main Jul 18, 2025
3 checks passed

achartier deleted the fp8_pb_wo branch July 18, 2025 02:36

reasonsolo pushed a commit to reasonsolo/TensorRT-LLM that referenced this pull request Jul 21, 2025

feat: add support for Modelopt fp8_pb_wo quantization scheme (NVIDIA#…

24eddd7

…6106) Signed-off-by: Aurelien Chartier <[email protected]> Co-authored-by: Haohang Huang <[email protected]>

timlee0212 pushed a commit to timlee0212/TensorRT-LLM that referenced this pull request Jul 21, 2025

feat: add support for Modelopt fp8_pb_wo quantization scheme (NVIDIA#…

4cc54d5

…6106) Signed-off-by: Aurelien Chartier <[email protected]> Co-authored-by: Haohang Huang <[email protected]>

This was referenced Jul 30, 2025

Deepseek R1 FP8 Support on Blackwell #6486

Merged

[TRTLLM-6992][feat] Add runtime swap AB for SM100 FP8 blockwise GEMM #6586

Open

achartier mentioned this pull request Aug 14, 2025

[https://nvbugs/5449155][fix] Fix DeepSeek R1 weight loading for TP16 #6913

Merged

This was referenced Aug 25, 2025

Add blockscale fp8 gemm to qwen2.5vl vlm part #7204

Open

[OMNIML-2336][feat] Add NVFP4 x FP8 #6809

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: add support for Modelopt fp8_pb_wo quantization scheme #6106

feat: add support for Modelopt fp8_pb_wo quantization scheme #6106

Uh oh!

achartier commented Jul 16, 2025 •

edited by coderabbitai bot

Loading

Uh oh!

coderabbitai bot commented Jul 16, 2025 •

edited

Loading

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

achartier commented Jul 16, 2025

Uh oh!

tensorrt-cicd commented Jul 16, 2025

Uh oh!

tensorrt-cicd commented Jul 17, 2025

Uh oh!

Uh oh!

achartier commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

Uh oh!

Uh oh!

feat: add support for Modelopt fp8_pb_wo quantization scheme #6106

feat: add support for Modelopt fp8_pb_wo quantization scheme #6106

Uh oh!

Conversation

achartier commented Jul 16, 2025 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR title

Description

Test Coverage

GitHub Bot Help

kill

skip

reuse-pipeline

Summary by CodeRabbit

Uh oh!

coderabbitai bot commented Jul 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Poem

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

achartier commented Jul 16, 2025

Uh oh!

tensorrt-cicd commented Jul 16, 2025

Uh oh!

tensorrt-cicd commented Jul 17, 2025

Uh oh!

Uh oh!

achartier commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

Uh oh!

Uh oh!

achartier commented Jul 16, 2025 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Jul 16, 2025 •

edited

Loading