Skip to content

Upstream merge 2025 04 29 #536

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
Merged
merged 226 commits into from
May 2, 2025
Merged

Upstream merge 2025 04 29 #536

merged 226 commits into from
May 2, 2025

Conversation

@gshtras
Copy link
Collaborator

@gshtras gshtras commented May 1, 2025

No description provided.

jpvillam-amd and others added 30 commits April 10, 2025 16:59
@jpvillam-amd
Add gfx950 to the attention archs
f4b308f 
Signed-off-by: jpvillam <[email protected]>
@jpvillam-amd
Linter
e201e58 
Signed-off-by: jpvillam <[email protected]>
@seungrokj
custom all-reduce, gfx950
ae144d6 
Signed-off-by: seungrokjung <[email protected]>
@hmellor
Bump Transformers to 4.51.3 (vllm-project#17116)
0bd7f8f 
Signed-off-by: Harry Mellor <[email protected]>
@hmellor
Use Transformers helper get_text_config() instead of checking for `…
423e9f1 
…text_config` (vllm-project#17105)

Signed-off-by: Harry Mellor <[email protected]>
@reidliu41
[doc] update wrong hf model links (vllm-project#17184)
df5c879 
Signed-off-by: reidliu41 <[email protected]>
Co-authored-by: reidliu41 <[email protected]>
@DarkLight1337
[Misc] Inline Molmo requirements (vllm-project#17190)
9d98ab5 
Signed-off-by: DarkLight1337 <[email protected]>
@russellb @ShangmingCai
[Security] Use safe serialization and fix zmq setup for mooncake pipe (…
a5450f1 
…vllm-project#17192)

Signed-off-by: Shangming Cai <[email protected]>
Co-authored-by: Shangming Cai <[email protected]>
@dyli-google
[V1] Move usage stats to worker and start logging TPU hardware (vllm-…
48cb210 
…project#16211)
@DarkLight1337
[Bugfix] Fix hybrid model tests (vllm-project#17182)
43faa04 
Signed-off-by: DarkLight1337 <[email protected]>
@tiran
Fix Python packaging edge cases (vllm-project#17159)
65e262b 
Signed-off-by: Christian Heimes <[email protected]>
@njhill
[BugFix][Frontend] Fix LLM.chat() tokenization (vllm-project#16081)
7011645 
Signed-off-by: Nick Hill <[email protected]>
@benchislett @luyuzhe111
[V1][Spec Decode] EAGLE-3 Support (vllm-project#16937)
a0e619e 
Signed-off-by: Bryan Lu <[email protected]>
Signed-off-by: Benjamin Chislett <[email protected]>
Co-authored-by: Bryan Lu <[email protected]>
@ruisearch42
[Misc] Refine ray_serve_deepseek example (vllm-project#17204)
c53e073 
Signed-off-by: Rui Qiao <[email protected]>
@heheda12345
[Bugfix] gemma[2,3] interleaved attention when sliding window is disa…
8de2901 
…bled (vllm-project#17180)

Signed-off-by: Chen Zhang <[email protected]>
@rasmith
[AMD][FP8][BugFix] Remove V1 check in arg_utils.py for FP8 since it i…
68af5f6 
…s not necessary (vllm-project#17215)

Signed-off-by: Randall Smith <[email protected]>
@ApostaC
[v1] [P/D] Adding LMCache KV connector for v1 (vllm-project#16625)
5e83a72
@jamesjwu
[Bugfix] [pytorch] Patch AOTAutogradCache._get_shape_env (vllm-projec…
a6e72e1 
…t#17142)

Signed-off-by: James Wu <[email protected]>
@houseroad
[MISC][AMD] Add unused annotation to rocm kernel file (vllm-project#1…
c8e5be3 
…7097)

Signed-off-by: Lu Fang <[email protected]>
@reidliu41
[doc] add Anything LLM integration (vllm-project#17216)
537d5ee 
Signed-off-by: reidliu41 <[email protected]>
Co-authored-by: reidliu41 <[email protected]>
@WoosukKwon
[Minor][Spec Decode] Add use_eagle to SpeculativeConfig (vllm-project…
1cf0719 
…#17213)

Signed-off-by: Woosuk Kwon <[email protected]>
@yarongmu-google
[Doc] Minor fix for the vLLM TPU setup page (vllm-project#17206)
7bd0c77 
Signed-off-by: Yarong Mu <[email protected]>
@WoosukKwon
[Minor][Models] Fix Return Types of Llama & Eagle (vllm-project#17220)
b278911 
Signed-off-by: Woosuk Kwon <[email protected]>
@wenscarl
Allocate kv_cache with stride order (vllm-project#16605)
9e96f56 
Signed-off-by: shuw <[email protected]>
@charlifu
[ROCm][Misc] Follow-ups for Skinny Gemms on ROCm. (vllm-project#17011)
54271bb 
Signed-off-by: charlifu <[email protected]>
@liuzijing2014 @markmc @njhill
[V1][Metrics] Allow V1 AsyncLLM to use custom logger (vllm-project#14661
53e8cf5 
)

Signed-off-by: Zijing Liu <[email protected]>
Signed-off-by: Mark McLoughlin <[email protected]>
Signed-off-by: Nick Hill <[email protected]>
Co-authored-by: Mark McLoughlin <[email protected]>
Co-authored-by: Nick Hill <[email protected]>
@njhill
[BugFix] Avoid race conditions in zero-copy tensor transmission (vllm…
b07bf83 
…-project#17203)

Signed-off-by: Nick Hill <[email protected]>
@WoosukKwon
[CI/test] Fix Eagle Correctness Test (vllm-project#17209)
513f074 
Signed-off-by: Woosuk Kwon <[email protected]>
@njhill
[Core] Remove prompt string from engine core data structures (vllm-pr…
df6f3ce 
…oject#17214)

Signed-off-by: Nick Hill <[email protected]>
@Isotr0py
[Bugfix] Fix missing int type for -n in multi-image example (vllm-p…
8c1c926 
…roject#17223)
LucasWilkinson and others added 18 commits May 2, 2025 03:16
@LucasWilkinson
[Attention] MLA move o_proj q_proj into cuda-graph region (vllm-proje…
afcb3f8 
…ct#17484)

Signed-off-by: Lucas Wilkinson <[email protected]>
@mgoin
[CI] Actually run tests/kv_transfer/test_disagg.py in CI (vllm-projec…
292fc59 
…t#17555)

Signed-off-by: mgoin <[email protected]>
@mgoin
Check if bitblas is installed during support check (vllm-project#17572)
b4003d1 
Signed-off-by: mgoin <[email protected]>
@DarkLight1337
[Misc] Continue refactoring model tests (vllm-project#17573)
f89d0e1 
Signed-off-by: DarkLight1337 <[email protected]>
@mgoin
Fix PixtralHF missing spatial_merge_size (vllm-project#17571)
f192ca9 
Signed-off-by: mgoin <[email protected]>
@jerryzh168
Add pt_load_map_location to allow loading to cuda (vllm-project#16869)
109e15a 
Signed-off-by: Jerry Zhang <[email protected]>
@Isotr0py
[Bugifx] Remove TritonPlaceholder from sys.modules (vllm-project#17317)
9e2de9b 
Signed-off-by: Isotr0py <[email protected]>
@qthequartermasterman @Bryce1010
@Nan2018 @DarkLight1337
[Core] [Bugfix] Add Input Embeddings (vllm-project#15428)
cc2a77d 
Signed-off-by: Andrew Sansom <[email protected]>
Signed-off-by: DarkLight1337 <[email protected]>
Co-authored-by: 临景 <[email protected]>
Co-authored-by: Bryce1010 <[email protected]>
Co-authored-by: Nan2018 <[email protected]>
Co-authored-by: Cyrus Leung <[email protected]>
Co-authored-by: DarkLight1337 <[email protected]>
@robertgshaw2-redhat
[BugFix] Fix Memory Leak (vllm-project#17567)
c777df7 
Signed-off-by: [email protected] <[email protected]>
@DarkLight1337
[Misc] Rename assets for testing (vllm-project#17575)
d754386 
Signed-off-by: DarkLight1337 <[email protected]>
@yangw-dev
add more pytorch related tests for torch nightly (vllm-project#17422)
b8b0859 
Signed-off-by: Yang Wang <[email protected]>
@reidliu41
[doc] add the print result (vllm-project#17584)
6d1479c 
Signed-off-by: reidliu41 <[email protected]>
Co-authored-by: reidliu41 <[email protected]>
@hmellor
Automatically tell users that dict args must be valid JSON in CLI (vl…
785d75a 
…lm-project#17577)

Signed-off-by: Harry Mellor <[email protected]>
@DarkLight1337
[Security] Fix image hash collision (vllm-project#17378)
99404f5 
Signed-off-by: DarkLight1337 <[email protected]>
@mgoin
Support W8A8 INT8 MoE for compressed-tensors (vllm-project#16745)
868c546 
Signed-off-by: mgoin <[email protected]>
@reidliu41
[doc] miss result (vllm-project#17589)
3a500cd 
Signed-off-by: reidliu41 <[email protected]>
Co-authored-by: reidliu41 <[email protected]>
@DarkLight1337
[Misc] Clean up input processing (vllm-project#17582)
cb23495 
Signed-off-by: DarkLight1337 <[email protected]>
@gshtras
Merge branch 'main' into upstream_merge_2025_04_29
2c68ff9
@gshtras gshtras requested a review from shajrawi May 2, 2025 15:36
shajrawi
shajrawi previously approved these changes May 2, 2025
View reviewed changes
@gshtras gshtras requested a review from shajrawi May 2, 2025 19:42
@gshtras gshtras force-pushed the upstream_merge_2025_04_29 branch from db3c89b to f3f620a Compare May 2, 2025 19:49
@gshtras gshtras merged commit 2fea69f into main May 2, 2025
3 of 8 checks passed
@gshtras gshtras deleted the upstream_merge_2025_04_29 branch May 2, 2025 22:01
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@shajrawi shajrawi shajrawi approved these changes

@Alexei-V-Ivanov-AMD Alexei-V-Ivanov-AMD Awaiting requested review from Alexei-V-Ivanov-AMD

@maleksan85 maleksan85 Awaiting requested review from maleksan85

@sunway513 sunway513 Awaiting requested review from sunway513

@hongxiayang hongxiayang Awaiting requested review from hongxiayang

@divakar-amd divakar-amd Awaiting requested review from divakar-amd

@charlifu charlifu Awaiting requested review from charlifu

@mawong-amd mawong-amd Awaiting requested review from mawong-amd

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.