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

[MFM] Merge changes from vllm-project/vllm main #28

Merged
merged 30 commits into from
Jan 27, 2025
Merged

[MFM] Merge changes from vllm-project/vllm main #28

merged 30 commits into from
Jan 27, 2025

Conversation

tjtanaa
Copy link

@tjtanaa tjtanaa commented Jan 27, 2025 •
edited by github-actions bot
Loading

No description provided.

ywang96 and others added 30 commits January 24, 2025 00:22
@ywang96
[Misc] Use VisionArena Dataset for VLM Benchmarking (#12389)
3c818bd 
Signed-off-by: Roger Wang <[email protected]>
@youkaichao
[ci/build] fix wheel size check (#12396)
c7c9851 
Signed-off-by: youkaichao <[email protected]>
@MohitIntel
[Hardware][Gaudi][Doc] Add missing step in setup instructions (#12382)
9a0f3bd
@youkaichao
[ci/build] sync default value for wheel size (#12398)
e784c6b 
Signed-off-by: youkaichao <[email protected]>
@jsato8094
[Misc] Enable proxy support in benchmark script (#12356)
3bb8e2c 
Signed-off-by: Junichi Sato <[email protected]>
@LucasWilkinson
[Bugfix][Kernel] Fix CUDA 11.8 being broken by FA3 build (#12375)
ab5bbf5 
Signed-off-by: Lucas Wilkinson <[email protected]>
@DarkLight1337
[Misc] Remove deprecated code (#12383)
df5dafa 
Signed-off-by: DarkLight1337 <[email protected]>
@LucasWilkinson
[Bugfix][Kernel] FA3 Fix - RuntimeError: This flash attention build o…
3132a93 
…nly supports pack_gqa (for build size reasons). (#12405)

Signed-off-by: Lucas Wilkinson <[email protected]>
@ElizaWszola
[Bugfix][Kernel] Fix moe align block issue for mixtral (#12413)
221d388
@DarkLight1337
[Bugfix] Fix BLIP-2 processing (#12412)
fb30ee9 
Signed-off-by: DarkLight1337 <[email protected]>
@divakar-amd
[ROCm][MoE] MI300 tuned configs Mixtral-8x(7B,22B) | fp16, fp8 (#12408)
bf21481 
Signed-off-by: Divakar Verma <[email protected]>
@Isotr0py
[Misc] Add FA2 support to ViT MHA layer (#12355)
f1fc051 
Signed-off-by: Isotr0py <[email protected]>
@lsy323
[TPU][CI] Update torchxla version in requirement-tpu.txt (#12422)
324960a 
Signed-off-by: Siyuan Liu <[email protected]>
@ywang96 @Isotr0py
[Misc][Bugfix] FA3 support to ViT MHA layer (#12435)
2a0309a 
Signed-off-by: Roger Wang <[email protected]>
Signed-off-by: Isotr0py <[email protected]>
Co-authored-by: Isotr0py <[email protected]>
@youngkent
[V1][Perf] Reduce scheduling overhead in model runner after cuda sync (…
fa63e71 
…#12094)

Signed-off-by: Keyun Tong <[email protected]>
@ywang96
[V1][Bugfix] Fix assertion when mm hashing is turned off (#12439)
0ee349b 
Signed-off-by: Roger Wang <[email protected]>
@ywang96
[Misc] Revert FA on ViT #12355 and #12435 (#12445)
a525527
@mhendrey @ShangmingCai
@youkaichao @hmellor @terrytangyuan @Isotr0py @DarkLight1337 @heheda12345 @wangxiyuan
[Frontend] generation_config.json for maximum tokens(#12242)
9ddc352 
Signed-off-by: Matthew Hendrey <[email protected]>
Signed-off-by: Shangming Cai <[email protected]>
Signed-off-by: youkaichao <[email protected]>
Signed-off-by: Harry Mellor <[email protected]>
Signed-off-by: Yuan Tang <[email protected]>
Signed-off-by: Isotr0py <[email protected]>
Signed-off-by: DarkLight1337 <[email protected]>
Signed-off-by: Chen Zhang <[email protected]>
Signed-off-by: wangxiyuan <[email protected]>
Co-authored-by: shangmingc <[email protected]>
Co-authored-by: youkaichao <[email protected]>
Co-authored-by: Harry Mellor <[email protected]>
Co-authored-by: Yuan Tang <[email protected]>
Co-authored-by: Isotr0py <[email protected]>
Co-authored-by: Cyrus Leung <[email protected]>
Co-authored-by: Chen Zhang <[email protected]>
Co-authored-by: wangxiyuan <[email protected]>
@tlrmchlsmth @mgoin
[Bugfix] Disable w16a16 2of4 sparse CompressedTensors24 (#12417)
aa2cd2c 
Signed-off-by: Tyler Michael Smith <[email protected]>
Co-authored-by: mgoin <[email protected]>
@tlrmchlsmth
[Bugfix/CI] Fix broken kernels/test_mha.py (#12450)
72f4880
@LucasWilkinson
[Bugfix][Kernel] Fix perf regression caused by PR #12405 (#12434)
68f1114 
Signed-off-by: Lucas Wilkinson <[email protected]>
@tlrmchlsmth
[Build/CI] Fix libcuda.so linkage (#12424)
72bac73 
Signed-off-by: Tyler Michael Smith <[email protected]>
@K-Mistele
[Frontend] Rerank API (Jina- and Cohere-compatible API) (#12376)
0034b09 
Signed-off-by: Kyle Mistele <[email protected]>
@terrytangyuan
[DOC] Add link to vLLM blog (#12460)
582cf78 
Signed-off-by: Yuan Tang <[email protected]>
@WoosukKwon
[V1] Avoid list creation in input preparation (#12457)
28e0750 
Signed-off-by: Woosuk Kwon <[email protected]>
@pooyadavoodi
[Frontend] Support scores endpoint in run_batch (#12430)
0cc6b38 
Signed-off-by: Pooya Davoodi <[email protected]>
@DarkLight1337
[Bugfix] Fix Granite 3.0 MoE model loading (#12446)
5204ff5 
Signed-off-by: DarkLight1337 <[email protected]>
@Isotr0py
[Bugfix] Fix missing seq_start_loc in xformers prefill metadata (#12464)
372bf08 
Signed-off-by: Isotr0py <[email protected]>
@WoosukKwon
[V1][Minor] Minor optimizations for update_from_output (#12454)
624a1e4 
Signed-off-by: Woosuk Kwon <[email protected]>
@Isotr0py
[Bugfix] Fix gpt2 GGUF inference (#12467)
ce69f7f 
Signed-off-by: Isotr0py <[email protected]>
@tjtanaa tjtanaa merged commit e8e548c into EmbeddedLLM:port-rocm-cpa-credit Jan 27, 2025
3 of 4 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

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

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.