[Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend #9823

yma11 · 2024-10-30T01:43:31Z

Upgrade to Pytorch 2.5 for xpu backend

BEFORE SUBMITTING, PLEASE READ THE CHECKLIST BELOW AND FILL IN THE DESCRIPTION ABOVE

PR Checklist (Click to Expand)

Thank you for your contribution to vLLM! Before submitting the pull request, please ensure the PR meets the following criteria. This helps vLLM maintain the code quality and improve the efficiency of the review process.

PR Title and Classification

Only specific types of PRs will be reviewed. The PR title is prefixed appropriately to indicate the type of change. Please use one of the following:

[Bugfix] for bug fixes.
[CI/Build] for build or continuous integration improvements.
[Doc] for documentation fixes and improvements.
[Model] for adding a new model or improving an existing model. Model name should appear in the title.
[Frontend] For changes on the vLLM frontend (e.g., OpenAI API server, LLM class, etc.)
[Kernel] for changes affecting CUDA kernels or other compute kernels.
[Core] for changes in the core vLLM logic (e.g., LLMEngine, AsyncLLMEngine, Scheduler, etc.)
[Hardware][Vendor] for hardware-specific changes. Vendor name should appear in the prefix (e.g., [Hardware][AMD]).
[Misc] for PRs that do not fit the above categories. Please use this sparingly.

Note: If the PR spans more than one category, please include all relevant prefixes.

Code Quality

The PR need to meet the following code quality standards:

We adhere to Google Python style guide and Google C++ style guide.
Pass all linter checks. Please use format.sh to format your code.
The code need to be well-documented to ensure future contributors can easily understand the code.
Include sufficient tests to ensure the project to stay correct and robust. This includes both unit tests and integration tests.
Please add documentation to docs/source/ if the PR modifies the user-facing behaviors of vLLM. It helps vLLM user understand and utilize the new features or changes.

Adding or changing kernels

Each custom kernel needs a schema and one or more implementations to be registered with PyTorch.

Make sure custom ops are registered following PyTorch guidelines: Custom C++ and CUDA Operators and The Custom Operators Manual
Custom operations that return Tensors require meta-functions. Meta-functions should be implemented and registered in python so that dynamic dims can be handled automatically. See above documents for a description of meta-functions.
Use torch.libary.opcheck() to test the function registration and meta-function for any registered ops. See tests/kernels for examples.
When changing the C++ signature of an existing op, the schema must be updated to reflect the changes.
If a new custom type is needed, see the following document: Custom Class Support in PT2.

Notes for Large Changes

Please keep the changes as concise as possible. For major architectural changes (>500 LOC excluding kernel/data/config/test), we would expect a GitHub issue (RFC) discussing the technical design and justification. Otherwise, we will tag it with rfc-required and might not go through the PR.

What to Expect for the Reviews

The goal of the vLLM team is to be a transparent reviewing machine. We would like to make the review process transparent and efficient and make sure no contributor feel confused or frustrated. However, the vLLM team is small, so we need to prioritize some PRs over others. Here is what you can expect from the review process:

After the PR is submitted, the PR will be assigned to a reviewer. Every reviewer will pick up the PRs based on their expertise and availability.
After the PR is assigned, the reviewer will provide status update every 2-3 days. If the PR is not reviewed within 7 days, please feel free to ping the reviewer or the vLLM team.
After the review, the reviewer will put an action-required label on the PR if there are changes required. The contributor should address the comments and ping the reviewer to re-review the PR.
Please respond to all comments within a reasonable time frame. If a comment isn't clear or you disagree with a suggestion, feel free to ask for clarification or discuss the suggestion.

Thank You

Finally, thank you for taking the time to read these guidelines and for your interest in contributing to vLLM. Your contributions make vLLM a great tool for everyone!

github-actions · 2024-10-30T01:43:44Z

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can do one of these:

Add ready label to the PR
Enable auto-merge.

🚀

yma11 · 2024-11-04T02:53:36Z

@zhouyuan @jikunshang please help take a review first. Thanks.

jikunshang · 2024-11-04T02:55:25Z

requirements-xpu.txt

+
+torch @ https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/torch-2.5.0a0%2Bgite84e33f-cp310-cp310-linux_x86_64.whl
+intel-extension-for-pytorch @ https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/intel_extension_for_pytorch-2.5.10%2Bgit9d489a8-cp310-cp310-linux_x86_64.whl
+oneccl_bind_pt == 2.3.100+xpu @ https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/oneccl_bind_pt-2.5.0%2Bxpu-cp310-cp310-linux_x86_64.whl


don't need version here?

zhouyuan · 2024-11-04T03:01:32Z

related
#9911
#8932

jikunshang

LGTM

yma11 · 2024-11-05T01:34:57Z

@youkaichao can you help review this PR? Thanks!

youkaichao

thanks for the great job! do you want to upgrade to pytorch 2.5.1 directly?

youkaichao · 2024-11-05T01:49:51Z

Dockerfile.xpu

    -r requirements-xpu.txt

+RUN git clone https://github.com/intel/pti-gpu && \


why do we need to build it here?

do you have released version for this?

This library doesn't have a released binary version. One alternative way is to download corresponding installation file and install it during docker build. But actually the build doesn't take much time. which way do you prefer?

wget https://registrationcenter-download.intel.com/akdlm/IRC_NAS/884eaa22-d56f-45dc-9a65-901f1c625f9e/l_intel-pti-dev_p_0.9.0.38_offline.sh sh ./l_intel-pti-dev_p_0.9.0.38_offline.sh source <path-of-installation>/pti/latest/env/vars.sh

@youkaichao by the way, for installing through installation file, we need to force using silent mode and it's not so reliable. Better to use current compile way. Any more concerns?

zhouyuan · 2024-11-05T02:22:58Z

thanks for the great job! do you want to upgrade to pytorch 2.5.1 directly?

Hi @youkaichao, thanks for the check! For this patch we want close PyTorch 2.5 support gap in #8932 as fast as we can
so the IPEX binary is targeting for 2.5.0 only. There will be a new IPEX release around early Dec to support newer PyTorch. BTW the 2.5.1 release have some fixes for CUDA backend however it should not bring big impact on XPU.

thanks, -yuan

Signed-off-by: Kunshang Ji <[email protected]>

Signed-off-by: yan ma <[email protected]>

youkaichao

thanks for the upgrade!

Signed-off-by: Kunshang Ji <[email protected]> Signed-off-by: yan ma <[email protected]> Co-authored-by: Kunshang Ji <[email protected]> Signed-off-by: Mozhou <[email protected]>

Signed-off-by: Kunshang Ji <[email protected]> Signed-off-by: yan ma <[email protected]> Co-authored-by: Kunshang Ji <[email protected]> Signed-off-by: Loc Huynh <[email protected]>

Signed-off-by: Kunshang Ji <[email protected]> Signed-off-by: yan ma <[email protected]> Co-authored-by: Kunshang Ji <[email protected]> Signed-off-by: Sumit Dubey <[email protected]>

yma11 requested review from mgoin and robertgshaw2-neuralmagic as code owners October 30, 2024 01:43

mergify bot added the ci/build label Oct 30, 2024

yma11 marked this pull request as draft October 30, 2024 02:34

DarkLight1337 added the Intel GPU label Oct 30, 2024

yma11 force-pushed the ipex-25 branch from f36c6f0 to c6a2373 Compare November 4, 2024 02:45

yma11 changed the title ~~[WIP]Upgrade to Pytorch 2.5 and AWQ/GPTQ support for xpu backend~~ Upgrade to Pytorch 2.5 for xpu backend Nov 4, 2024

yma11 changed the title ~~Upgrade to Pytorch 2.5 for xpu backend~~ [Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend Nov 4, 2024

yma11 force-pushed the ipex-25 branch from c6a2373 to d266fd2 Compare November 4, 2024 02:48

yma11 marked this pull request as ready for review November 4, 2024 02:53

jikunshang reviewed Nov 4, 2024

View reviewed changes

jikunshang approved these changes Nov 4, 2024

View reviewed changes

yma11 force-pushed the ipex-25 branch from 01e589c to f7a9023 Compare November 5, 2024 01:31

youkaichao reviewed Nov 5, 2024

View reviewed changes

jikunshang and others added 7 commits November 5, 2024 17:07

add softcap for gemma2 model

e641e87

Signed-off-by: Kunshang Ji <[email protected]>

fix libpti_view.so not found issue.

75f40e5

Signed-off-by: Kunshang Ji <[email protected]>

ipex api change

a40eeca

Signed-off-by: Kunshang Ji <[email protected]>

fix paged_attention, decide use v1/v2 on ipex level

29fc5cb

Signed-off-by: Kunshang Ji <[email protected]>

update dependencies and revert API change

4c38158

Signed-off-by: yan ma <[email protected]>

fix format

9ef1354

Signed-off-by: yan ma <[email protected]>

address comments

f7a9023

Signed-off-by: yan ma <[email protected]>

youkaichao approved these changes Nov 7, 2024

View reviewed changes

youkaichao merged commit d3859f1 into vllm-project:main Nov 7, 2024
31 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend #9823

[Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend #9823

yma11 commented Oct 30, 2024 •

edited

Loading

github-actions bot commented Oct 30, 2024

yma11 commented Nov 4, 2024

jikunshang Nov 4, 2024

zhouyuan commented Nov 4, 2024 •

edited

Loading

jikunshang left a comment

yma11 commented Nov 5, 2024

youkaichao left a comment

youkaichao Nov 5, 2024

yma11 Nov 5, 2024

yma11 Nov 6, 2024

zhouyuan commented Nov 5, 2024 •

edited

Loading

youkaichao left a comment

		-r requirements-xpu.txt

		RUN git clone https://github.com/intel/pti-gpu && \

[Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend #9823

[Misc][XPU] Upgrade to Pytorch 2.5 for xpu backend #9823

Conversation

yma11 commented Oct 30, 2024 • edited Loading

PR Title and Classification

Code Quality

Adding or changing kernels

Notes for Large Changes

What to Expect for the Reviews

Thank You

github-actions bot commented Oct 30, 2024

yma11 commented Nov 4, 2024

jikunshang Nov 4, 2024

Choose a reason for hiding this comment

zhouyuan commented Nov 4, 2024 • edited Loading

jikunshang left a comment

Choose a reason for hiding this comment

yma11 commented Nov 5, 2024

youkaichao left a comment

Choose a reason for hiding this comment

youkaichao Nov 5, 2024

Choose a reason for hiding this comment

yma11 Nov 5, 2024

Choose a reason for hiding this comment

yma11 Nov 6, 2024

Choose a reason for hiding this comment

zhouyuan commented Nov 5, 2024 • edited Loading

youkaichao left a comment

Choose a reason for hiding this comment

yma11 commented Oct 30, 2024 •

edited

Loading

zhouyuan commented Nov 4, 2024 •

edited

Loading

zhouyuan commented Nov 5, 2024 •

edited

Loading