vllm-project/vllm

Pull Requests Commits

deprecate head_first

yewentao256 committed 3 days ago

6010cedb

[Model Runner V2] Use pinned memory for write_contents (#34222)

WoosukKwon committed 3 days ago

Verified a2443de5

[Docs] Speed up build environment set-up (#34240)

hmellor committed 3 days ago

Verified f84a2a8f

[BUGFIX] Fix accuracy bugs in Qwen3-Next MTP (#34077)

vadiklyutiy committed 3 days ago

Verified 000214c4

[Core][BugFix] Fix PP KV cache sharding memory validation (#33698)

junuxyz committed 3 days ago

Verified c5a66d16

[Perf][Kernel] Add faster topKperRow decode kernel for DeepSeek-V3.2 sparse attention (#33680)

LopezCastroRoberto committed 3 days ago

Verified afdce12c

[compile] Enable AOT compile with 2.10 in trunk. (#34155)

zhxchen17 committed 3 days ago

Verified 82e11973

[ROCm][Quantization] GPT_OSS in amd-quark format model loading and emulations (#29008)

xuebwang-amd committed 3 days ago

Verified b129136c

Support benchmarking of Geospatial models (#33922)

mgazz committed 3 days ago

Verified 599e4335

add --insecure arg to the vllm bench to skip TLS (#34026)

fanyang-real committed 3 days ago

Verified a1946570

Bump `mamba-ssm` version in CI for Transformers v5 compatibility (#34233)

hmellor committed 3 days ago

Verified d0bc5205

[V1][BugFix] Fix EAGLE3 encoder cache miss with disable_chunked_mm_input (#34220)

KrxGu committed 3 days ago

Verified 748625cd

Stop testing for slow tokenizers as they will not exist soon (#34235)

hmellor committed 3 days ago

Verified 61413973

[Misc] allow specify is_mm_prefix_lm in hf_config (#34215)

lkhphuc committed 3 days ago

Verified 94de8715

Add flagos in MiniCPM-o (#34126)

tc-mb committed 3 days ago

Verified e042d7e6

[Bugfix] Fix FI kernel`chunk_gated_delta_rule` output shape for Qwen3.5 (#34219)

ywang96 committed 3 days ago

Verified ae4e2806

[Docs] Fix format error in KV load failure recovery doc (#34137)

zzaebok committed 3 days ago

Verified cbea11c9

[Bugfix] Fix `--trust-remote-code` conflict (#34218)

DarkLight1337 committed 3 days ago

Verified 2c32558a

[Bugfix] Fix memory inconsistency in cross-process shared memory (#32022)

slippersss committed 3 days ago

Verified 5f970120

Revert #34208 (#34216)

DarkLight1337 committed 3 days ago

Verified 998e2d91

[Perf] Optimize detokenizer python logic (#32975)

yewentao256 committed 3 days ago

Verified e1060a71

[BugFix] Avoid prefix cache hit in the same schedule step for mamba layers (#29387)

heheda12345 committed 3 days ago

Verified 97fa8f65

[Frontend][CI] Consolidate instrumentator entrypoints (#34123)

noooop committed 3 days ago

Verified dab1de9f

[Bugfix] Sort hf_weights_files in fastsafetensors_weights_iterator to match #33491 (#34190)

jaim12005 committed 3 days ago

Verified 8d48d0a9

[responsesAPI] fix simpleContext streaming output_messages (#34188)

qandrew committed 3 days ago

Verified 9608844f

[Bugfix] Add `--trust-remote-code` to dataset bench args (#34208)

DarkLight1337 committed 3 days ago

Verified f69b903b

[Bugfix] Fix DP Attention Padding in Dummy Run (#34187)

LucasWilkinson committed 4 days ago

Verified 81e217fe

[CI/Build] Relax `test_mcp_tool_call` (#34204)

DarkLight1337 committed 4 days ago

Verified ab97bcf6

[Doc] Update usage of `--limit-mm-per-prompt` (#34148)

DarkLight1337 committed 4 days ago

Verified 25e48a3a

[Bugfix][Core] Fix CPU memory leak from Request reference cycle in prefix caching (#34183)

ywang96 committed 4 days ago

Verified 8a5e0e2b

Older