Morph MORPH
SIGN IN SIGN UP
tejas / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

0 0 122 Python

COMMITS

main
October 19, 2025
W
[Minor] Remove unused env variable (#27161)
Woosuk Kwon committed
fb86067
October 18, 2025
T
[V1][Metrics][Plugin] Add plugin support for custom `StatLoggerBase` implementations (#22456)
Tova Movshovitz committed
83e760c
L
[BugFix] Disable fp8 kv-cache by default for DeepSeek V3.2 (#27121)
Lucas Wilkinson committed
c2bba69
B
[BugFix] fix graph partition signature (#27139)
Boyuan Feng committed
e133d6d
D
[Chore] Separate out profiling utilities from vllm.utils (#27150)
dongbo910220 committed
a1946c9
L
[BugFix] Fix failing gemma-3-1b-it test: `test_lm_eval_accuracy_v1_engine[google/gemma-3-1b-it]` (#27111)
Lucas Wilkinson committed
9f020f4
N
[Minor] Add some clarifying comments to recent changes (#27130)
Nick Hill committed
3b45075
Y
Fix incorrect string formatting in barrier timeout exceptions (#27149)
Yongtao Huang committed
168e578
I
[Chore] Clean up pytorch helper functions in `vllm.utils` (#26908)
Isotr0py committed
6ac5e06
L
[Models][QwenVL] Remove unnecessary `.contiguous()` calls (#27106)
Lukas Geiger committed
5c2acb2
N
[Misc] Refactor `get_kv_cache_spec` into `AttentionLayerBase` (#26587)
Nicolò Lucchesi committed
b26b70b
F
[fix][cpu] fix prefill attention in CPU attention backend (#27035)
Fadi Arafeh committed
ab4be40
W
[Feature] Batch Invariant: Support DeepGEMM and Blackwell (#27127)
Wentao Ye committed
245e4f2
I
[Chore] Separate out `vllm.utils.mem_utils` (#27143)
iAmir97 committed
1d165d6
D
[Test] Add test for /health endpoint on engine failure (#26074)
dongbo910220 committed
8300402
C
[DOC][FEATURES][CPU]update cpu feature for v1 (#27135)
Chendi.Xue committed
12e2170
V
[Misc] Rev DeepEP (#27122)
Varun Sundar Rabindranath committed
30a33b9
H
[GPT-OSS] Structure_Tag support for gpt-oss tool-call in cot (#25515)
Hanchenli committed
7c57254
H
[CI/Build] tests(v1): feed Triton attention the (num_blocks, 2, …) KV cache layout in backend-correctness tests (#26663)
Huamin Li committed
c312320
Z
[Perf] Add H100 fused MoE config (#25398)
ZiTian Zhao committed
c981f0e
October 17, 2025
L
[BugFix][Core] Fix error when enable async-scheduling in multi-node env (#25887)
Lehua Ding committed
6367bde
W
[Test] Make `test_failure` more stable for batch invariance (#27054)
Wentao Ye committed
f50cc22
P
[V1][Spec Decode] Fix greedy temperature detection after sampler refactor (#27077)
Pradyun92 committed
acedc74
Z
[Minor] Remove unnecessary error message (#27115)
Zhuohan Li committed
d29483b
M
[Bugfix] Use PIECEWISE cudagraphs on Blackwell if max_model_len > 131072 (#27114)
Michael Goin committed
950cf9e
I
[Chore] Remove unused `PolyNorm` layer (#27110)
Isotr0py committed
3125d79
V
[Bugfix] [AITER] [ROCm] Fix Quark MoE Quant Config and AITER Fused MoE quant type logic (#27029)
vllmellm committed
e33ee23
R
[ROCm][Bugfix][Model] Fix illegal memory access when running qwen3_moe models with rms_norm (Qwen3-235B-A22B, Qwen3-30B-A3B, etc.) (#26192)
rasmith committed
b10c64c
A
[ROCM] MoE fp4 CK kernel (#26545)
Aleksandr Malyshev committed
0925b28
N
[CI] Remove forbidden slash (#27112)
Nicolò Lucchesi committed
99722d5