Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Refactor TPU requirements file and pin build dependencies #10008

Closed
wants to merge 56 commits into from
Closed

Refactor TPU requirements file and pin build dependencies #10008

wants to merge 56 commits into from

Conversation

richardsliu
Copy link
Contributor

Replaces #9898

@github-actions GitHub Actions
Copy link

github-actions bot commented Nov 4, 2024

👋 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.

🚀

@mergify mergify bot added documentation Improvements or additions to documentation ci/build labels Nov 4, 2024
@mergify Mergify
Copy link

mergify bot commented Nov 4, 2024

This pull request has merge conflicts that must be resolved before it can be
merged. @richardsliu please rebase it. https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

@mergify mergify bot added the needs-rebase label Nov 4, 2024
@bvrockwell
Copy link
Contributor

@robertgshaw2-neuralmagic this fixes the issue in the docs and the requirements file.

Thank you @richardsliu !!

@mergify mergify bot removed the needs-rebase label Nov 4, 2024
richardsliu and others added 22 commits November 4, 2024 23:46
@richardsliu
Pin dockerfile
59776c5 
Signed-off-by: Richard Liu <[email protected]>
@richardsliu
fix manual installation
1fc4621 
Signed-off-by: Richard Liu <[email protected]>
@richardsliu
refactor requirements tpu
0c971e4 
Signed-off-by: Richard Liu <[email protected]>
@richardsliu
update torch-xla
d00f062 
Signed-off-by: Richard Liu <[email protected]>
@richardsliu
fix documentation
23dfa46 
Signed-off-by: Richard Liu <[email protected]>
@joerunde @richardsliu
[Bugfix][Frontend] Reject guided decoding in multistep mode (vllm-pro…
9efb5fe 
…ject#9892)

Signed-off-by: Joe Runde <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@youkaichao @richardsliu
[torch.compile] directly register custom op (vllm-project#9896)
0458bc6 
Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@mgoin @richardsliu
[Bugfix] Fix layer skip logic with bitsandbytes (vllm-project#9887)
fd71c52 
Signed-off-by: mgoin <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@youkaichao @richardsliu
[torch.compile] rework test plans (vllm-project#9866)
2f80b13 
Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@mgoin @richardsliu
[Model] Support bitsandbytes for MiniCPMV (vllm-project#9891)
7c07ae3 
Signed-off-by: mgoin <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@CRZbulabula @youkaichao @richardsliu
[torch.compile] Adding torch compile annotations to some models (vllm…
92b845d 
…-project#9876)

Signed-off-by: youkaichao <[email protected]>
Co-authored-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@DarkLight1337 @richardsliu
[Doc] Update multi-input support (vllm-project#9906)
76936dc 
Signed-off-by: Richard Liu <[email protected]>
@DarkLight1337 @richardsliu
[Frontend] Chat-based Embeddings API (vllm-project#9759)
6432426 
Signed-off-by: Richard Liu <[email protected]>
@mgoin @richardsliu
[CI/Build] Add Model Tests for PixtralHF (vllm-project#9813)
cc856c0 
Signed-off-by: Richard Liu <[email protected]>
@DarkLight1337 @richardsliu
[Frontend] Use a proper chat template for VLM2Vec (vllm-project#9912)
3678bf1 
Signed-off-by: Richard Liu <[email protected]>
@tjohnson31415 @prashantgupta24
@patrickvonplaten @richardsliu
[Bugfix] Fix edge cases for MistralTokenizer (vllm-project#9625)
f2f6195 
Signed-off-by: Travis Johnson <[email protected]>
Signed-off-by: Prashant Gupta <[email protected]>
Co-authored-by: Prashant Gupta <[email protected]>
Co-authored-by: Patrick von Platen <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@andrejonasson @richardsliu
[Core] Refactor: Clean up unused argument in Scheduler._preempt (vllm…
7a720dc 
…-project#9696)

Signed-off-by: André Jonasson <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@youkaichao @richardsliu
[torch.compile] use interpreter with stable api from pytorch (vllm-pr…
a04aeb0 
…oject#9889)

Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@pavanimajety @richardsliu
[Bugfix/Core] Flashinfer k_scale and v_scale (vllm-project#9861)
e66b29b 
Signed-off-by: Richard Liu <[email protected]>
@youkaichao @richardsliu
[1/N] pass the complete config from engine to executor (vllm-project#…
522cc4a 
…9933)

Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@GeneDer @richardsliu
[Bugfix] PicklingError on RayTaskError (vllm-project#9934)
771f4fd 
Signed-off-by: Gene Su <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@dependabot @khluu @richardsliu
[ci/build] Bump the patch-update group with 10 updates (vllm-project#…
7655274 
…9897)

Signed-off-by: dependabot[bot] <[email protected]>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Kevin H. Luu <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
youkaichao and others added 12 commits November 4, 2024 23:47
@youkaichao @richardsliu
[4/N] make quant config first-class citizen (vllm-project#9978)
6eb2ed0 
Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@jeejeelee @richardsliu
[Misc]Reduce BNB static variable (vllm-project#9987)
a2a024e 
Signed-off-by: Jee Jee Li <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@mzusman @richardsliu
[Model] factoring out MambaMixer out of Jamba (vllm-project#8993)
db45c39 
Signed-off-by: mzusman <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@robertgshaw2-neuralmagic @richardsliu
[CI] Basic Integration Test For TPU (vllm-project#9968)
af148c8 
Signed-off-by: Robert Shaw <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@hissu-hyvarinen @richardsliu
[Bugfix][CI/Build][Hardware][AMD] Shard ID parameters in AMD tests ru…
6d353ed 
…nning parallel jobs (vllm-project#9279)

Signed-off-by: Hissu Hyvarinen <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@ywang96 @richardsliu
[Doc] Update VLM doc about loading from local files (vllm-project#9999)
db1e670 
Signed-off-by: Roger Wang <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@robertgshaw2-neuralmagic @richardsliu
[Bugfix] Fix MQLLMEngine hanging (vllm-project#9973)
d667c05 
Signed-off-by: [email protected] <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@lk-chen @richardsliu
[Misc] Refactor benchmark_throughput.py (vllm-project#9779)
af37b9b 
Signed-off-by: Linkun Chen <[email protected]>
Co-authored-by: Linkun Chen <[email protected]>
Co-authored-by: Linkun Chen <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@tomeras91 @richardsliu
[Frontend] Add max_tokens prometheus metric (vllm-project#9881)
20f6878 
Signed-off-by: Tomer Asida <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@bnellnm @richardsliu
[Bugfix] Upgrade to pytorch 2.5.1 (vllm-project#10001)
767e8ad 
Signed-off-by: Bill Nell <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@youkaichao @richardsliu
[4.5/N] bugfix for quant config in speculative decode (vllm-project#1…
fc7d716 
…0007)

Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Richard Liu <[email protected]>
@richardsliu
Update tpu-installation.rst
b98cf25 
Signed-off-by: Richard Liu <[email protected]>
@mergify Mergify
Copy link

mergify bot commented Nov 4, 2024

This pull request has merge conflicts that must be resolved before it can be
merged. @richardsliu please rebase it. https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

@richardsliu richardsliu closed this Nov 4, 2024
@richardsliu richardsliu deleted the requirements branch November 4, 2024 23:54
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@youkaichao youkaichao Awaiting requested review from youkaichao youkaichao is a code owner

@tlrmchlsmth tlrmchlsmth Awaiting requested review from tlrmchlsmth tlrmchlsmth is a code owner

@WoosukKwon WoosukKwon Awaiting requested review from WoosukKwon WoosukKwon is a code owner

@njhill njhill Awaiting requested review from njhill njhill is a code owner

@LiuXiaoxuanPKU LiuXiaoxuanPKU Awaiting requested review from LiuXiaoxuanPKU LiuXiaoxuanPKU is a code owner

@DarkLight1337 DarkLight1337 Awaiting requested review from DarkLight1337 DarkLight1337 is a code owner

@ywang96 ywang96 Awaiting requested review from ywang96 ywang96 is a code owner

@robertgshaw2-neuralmagic robertgshaw2-neuralmagic Awaiting requested review from robertgshaw2-neuralmagic robertgshaw2-neuralmagic is a code owner

@simon-mo simon-mo Awaiting requested review from simon-mo simon-mo is a code owner

@zhuohan123 zhuohan123 Awaiting requested review from zhuohan123 zhuohan123 is a code owner

@alexm-neuralmagic alexm-neuralmagic Awaiting requested review from alexm-neuralmagic alexm-neuralmagic is a code owner

@comaniac comaniac Awaiting requested review from comaniac comaniac is a code owner

Assignees
No one assigned
Labels
ci/build documentation Improvements or additions to documentation frontend needs-rebase
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.