[Kernel] Expand MoE weight loading + Add Fused Marlin MoE Kernel #7527
Conversation
|
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge). To run full CI, you can do one of these:
🚀 |
dsikka
changed the title
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py
Outdated
Show resolved
Hide resolved
dsikka
changed the title
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py
Outdated
Show resolved
Hide resolved
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py
Outdated
Show resolved
Hide resolved
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py
Outdated
Show resolved
Hide resolved
|
/ready |
github-actions
bot
added
the
ready
dsikka
force-pushed
the
compressed-tensors-moe-updated
branch
from
ac3c2b9 to
324ecfd
Compare
FILL IN THE PR DESCRIPTION HERE
FIX #xxxx (*link existing issues this PR will resolve*)
**BEFORE SUBMITTING, PLEASE READ THE CHECKLIST BELOW AND FILL IN THE
DESCRIPTION ABOVE**
---
<details>
<!-- inside this <details> section, markdown rendering does not work, so
we use raw html here. -->
<summary><b> PR Checklist (Click to Expand) </b></summary>
<p>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.</p>
<h3>PR Title and Classification</h3>
<p>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:</p>
<ul>
<li><code>[Bugfix]</code> for bug fixes.</li>
<li><code>[CI/Build]</code> for build or continuous integration
improvements.</li>
<li><code>[Doc]</code> for documentation fixes and improvements.</li>
<li><code>[Model]</code> for adding a new model or improving an existing
model. Model name should appear in the title.</li>
<li><code>[Frontend]</code> For changes on the vLLM frontend (e.g.,
OpenAI API server, <code>LLM</code> class, etc.) </li>
<li><code>[Kernel]</code> for changes affecting CUDA kernels or other
compute kernels.</li>
<li><code>[Core]</code> for changes in the core vLLM logic (e.g.,
<code>LLMEngine</code>, <code>AsyncLLMEngine</code>,
<code>Scheduler</code>, etc.)</li>
<li><code>[Hardware][Vendor]</code> for hardware-specific changes.
Vendor name should appear in the prefix (e.g.,
<code>[Hardware][AMD]</code>).</li>
<li><code>[Misc]</code> for PRs that do not fit the above categories.
Please use this sparingly.</li>
</ul>
<p><strong>Note:</strong> If the PR spans more than one category, please
include all relevant prefixes.</p>
<h3>Code Quality</h3>
<p>The PR need to meet the following code quality standards:</p>
<ul>
<li>We adhere to <a
href="https://google.github.io/styleguide/pyguide.html">Google Python
style guide</a> and <a
href="https://google.github.io/styleguide/cppguide.html">Google C++
style guide</a>.</li>
<li>Pass all linter checks. Please use <a
href="https://github.com/vllm-project/vllm/blob/main/format.sh"><code>format.sh</code></a>
to format your code.</li>
<li>The code need to be well-documented to ensure future contributors
can easily understand the code.</li>
<li>Include sufficient tests to ensure the project to stay correct and
robust. This includes both unit tests and integration tests.</li>
<li>Please add documentation to <code>docs/source/</code> if the PR
modifies the user-facing behaviors of vLLM. It helps vLLM user
understand and utilize the new features or changes.</li>
</ul>
<h3>Notes for Large Changes</h3>
<p>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 <code>rfc-required</code>
and might not go through the PR.</p>
<h3>What to Expect for the Reviews</h3>
<p>The goal of the vLLM team is to be a <i>transparent reviewing
machine</i>. 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: </p>
<ul>
<li> 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.</li>
<li> 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.</li>
<li> After the review, the reviewer will put an <code>
action-required</code> label on the PR if there are changes required.
The contributor should address the comments and ping the reviewer to
re-review the PR.</li>
<li> 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.
</li>
</ul>
<h3>Thank You</h3>
<p> 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! </p>
</details>
dsikka
force-pushed
the
compressed-tensors-moe-updated
branch
from
20b5940 to
fe4c9e2
Compare
…nel (vllm-project#7527)" This reverts commit 8678a69.
|
I tried to run deepseek-v2-fp8(https://huggingface.co/neuralmagic/DeepSeek-Coder-V2-Instruct-FP8) with this PR and 8 L40 GPUs, and encountered the following error: |
|
@fengyang95 it looks like your issue is due to incorrectly compiled or linked functions, possibly from older builds. Could you build vllm in a fresh environment? |
|
|
@fengyang95 Do you mind sharing a code snippet of what you're running? |
|
|
@fengyang95 FP8 models will use the triton kernels which should work with the DeepSeek V2 model listed. |
…m-project#7527) Co-authored-by: ElizaWszola <[email protected]>
|
How should I load a w4a16 version of deepseek-v2 by vllm==0.5.5 that was compressed using llm-compressor? File "/usr/local/lib/python3.9/dist-packages/vllm/model_executor/layers/fused_moe/layer.py", line 192, in init |
…m-project#7527) Co-authored-by: ElizaWszola <[email protected]> Signed-off-by: Alvant <[email protected]>
…nel (vllm-project#7527)" (vllm-project#7764) Signed-off-by: Alvant <[email protected]>
…m-project#7527) Co-authored-by: ElizaWszola <[email protected]>
Summary
w4a16by @ElizaWszolaCompressedTensorsMoEMethodto support MoEw4a16models from llm-compressor and compressed-tensorsTODO:
Next Steps:
CompressedTensorsMoEMethodis not leveraging the scheme structure in-place for compressed-tensors in order to keep the scope of this PR focused on the kernel + updated weight loading. Will be updated in a follow-up to use the scheme structureco-authored by @ElizaWszola, from Neural Magic