vllm-project/vllm

Pull Requests Commits

[ROCm][CI] Pin TorchCodec to v0.10.0 for ROCm compatibility (#34447)

AndreasKaratzas committed 54 days ago

a976961f

[ci] Use the right tag for CPU arm64 image (#34915)

khluu committed 54 days ago

772f7083

[CI/Build] Fix gRPC version mismatch (#35013)

DarkLight1337 committed 54 days ago

90308959

khluu committed 57 days ago

61c34465

Revert "img build"

khluu committed 57 days ago

31575c1d

khluu committed 57 days ago

5dc8728b

Revert "[Release 2.10] Update to Torch 2.10 - final release (#30525)"

khluu committed 57 days ago

d4db22ab

Revert "[Bugfix][ROCm][GPT-OSS] Use old triton_kernels implementation on ROCm if the new API is not available (#34153)"

khluu committed 57 days ago

ad2a6bb3

khluu committed 65 days ago

2d5be1dd

[Bugfix] Fix MTP accuracy for GLM-5 (#34385)

mgoin committed 65 days ago

7a06e5b0

[Bugfix] send None sentinel on final commit so server properly sends transcription.done (#33963)

pjs102793 committed 65 days ago

946b2f10

[Misc] Bump `fastsafetensors` version for latest fixes (#34273)

njhill committed 65 days ago

5e8adb0c

[Bugfix] fix default is_neox_style is True for deepseek (#34353)

xyDong0223 committed 65 days ago

9be1ff2d

[Model] GLM adaptation (#34124)

jeejeelee committed 65 days ago

b3ee90f9

Patch protobuf for CVE-2026-0994 (#34253)

eicherseiji committed 66 days ago

c44d0c6d

[XPU][9/N] clean up existing ipex code/doc (#34111)

jikunshang committed 66 days ago

83db96d8

[XPU][7/N] enable xpu fp8 moe (#34202)

zufangzhu committed 66 days ago

dbfb79fe

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

ywang96 committed 66 days ago

b2e1fc35

[Bugfix][ROCm][GPT-OSS] Use old triton_kernels implementation on ROCm if the new API is not available (#34153)

gshtras committed 66 days ago

55a1baeb

[torch.compile][Fusion] Fix attention fusion pass removing kv_udpate op. (#33945)

charlifu committed 66 days ago

e1e98416

[XPU][6/N] add xpu scaled_mm kernel (#34117)

zufangzhu committed 66 days ago

5bd63387

[Frontend][last/5] Make pooling entrypoints request schema consensus. (#31127)

noooop committed 68 days ago

Verified 22b64948

[Tiny] Rename encoder budget file to more specific name (#34103)

reaganjlee committed 68 days ago

Verified 7c233dbb

[bug-fix] supported_tasks is breaking backward compatibility at init_app_state (#34027)

kouroshHakha committed 68 days ago

Verified a75a5b54

[Release 2.10] Update to Torch 2.10 - final release (#30525)

atalman committed 69 days ago

Verified f97ca671

Add support for ModelOpt MXFP8 dense models (#33786)

danisereb committed 69 days ago

Verified 084aa19f

glm 4.6 fused tuned inference config for B200 (#32958)

navmarri14 committed 69 days ago

Verified 1ecfabe5

[torch.compile] Add an option to force-enable the MOE cold start optimization (#33735)

zou3519 committed 69 days ago

Verified 4df841fe

[BugFix] Change support no act and mul for marlin (#34088)

TomerBN-Nvidia committed 69 days ago

Verified a263aa61

[Revert] Fix performance regression for GLM-4.7-GPTQ decode and MTP acceptance rate (#33771)

aabbccddwasd committed 69 days ago

Verified 179ae7da

Older