Morph MORPH
SIGN IN SIGN UP
dhruvhead / vllm UNCLAIMED
0 0 1 Python

COMMITS

main
August 26, 2026
T
[Agents] Add CUDA IMA debugging skill (#53702)
Thien Tran committed
d125b54
W
[Kimi Bug] Fix k3 torch.AcceleratorError: CUDA error: an illegal memory access was encountered (#53773)
Wentao Ye committed
a447955
August 25, 2026
C
[Kimi K3][Kernel] Enable low-latency decode GEMM dispatch on SM100 (#53534)
Canlin Guo committed
bc2d63e
A
[Bugfix][KV Offload] Defer request-level cascade of in-flight primary keys (#53329)
Almog Tavor committed
bc39ded
D
[Core] Disable fuse_allreduce_rms under VLLM_BATCH_INVARIANT (non-deterministic under TP) (#51292)
Donald Tolley committed
80771bb
W
[CI Bug] Fix kimi test `AssertionError: Aligned Mamba state indices must be precomputed` (#53766)
Wentao Ye committed
0e30bd6
W
[Perf] Autotune batch invariance triton kernel in blackwell, 33.6% E2E latency reduction (#53649)
Wentao Ye committed
384d079
N
[Bugfix][MRV2] Run cudagraph memory profiling in a throwaway graph pool (#53682)
Nick Hill committed
e6cc089
M
[CI] Preserve Rust Docker cache across commits (#53290)
Misha Goin committed
c9331d8
M
[Bugfix][Multimodal] Reject malformed base64 audio with 400 instead of 500 (#53744)
Moe Huzaifa committed
299ebd0
M
[Spec Decode] Enable adaptive DSpark on SM100 sparse MLA (#52783)
Misha Goin committed
7de9605
W
[Mypy Fix] Mypy fix for "vllm/model_executor/models/[gG]" (#53616)
Wentao Ye committed
70021a0
M
[Bugfix][Frontend] Apply the stop string limit to Cohere requests (#53750)
Moe Huzaifa committed
217a0b5
M
[Bugfix][Frontend] Keep credentials out of the Rust frontend launch log (#53738)
Moe Huzaifa committed
3e83393
M
[Bugfix][Tokenizer] Replace bare asserts in the DeepSeek V4 encoder (#53747)
Moe Huzaifa committed
19406fa
L
[Bugfix] Resolve B12X modules before Dynamo tracing (#53326)
Luke Alonso committed
3454335
C
[CI] Add GSM8K accuracy test for amd/DeepSeek-V4-Flash-MXFP4 (#50632)
Colin Z committed
bc11eca
Y
[XPU] update key supported models (#53494)
Yan Ma committed
10ade93
W
[K3 Perf] Optimize k3 mamba metadata preparation, 6.6~7.6x kernel performance improvement (#52388)
Wentao Ye committed
41729fc
N
[Feature][DSpark]: Logprobs adaptive verification (#52242)
Naveenraj Kamalakannan committed
d5cadce
E
[Quantization][Humming] Support MXFP4 weight + block-FP8 activation for MoE (#51332)
Elvir Crnčević committed
ca29cd4
T
[Bugfix] Handle parenthesized Gemma4 tool calls (#53657)
Taneem Ibrahim committed
9c8e90e
T
[Pooling UX] Improve serve --task error guidance (#53467)
Taneem Ibrahim committed
2014e47
E
[Profiler] Fix start_profile permanently no-op after max_iterations auto-stop (#51839)
Elvir Crnčević committed
afc91fa
D
[ROCm][CI] Skip ModernBERT FP8 MTEB test when no FP8 ScaledMM kernel exists (#53589)
djramic committed
d9fbe52
T
[Config][EC] Normalize producer-only encoder config (#53656)
Tianyu Guo committed
12c84b9
J
[Bugfix][DeepSeek V4] Handle trailing system messages in prompt rendering (#51262)
Jiahao Liang committed
59d7fc9
N
[sleep functionality] code refactor about sleep/wake_up (#50431)
Ning Xie committed
23310bc
T
[Agents] Add kernel microbenchmark skill (#53688)
Thien Tran committed
5e379a3
J
[Bugfix][MM] Fix JinaVL processing cache order (#53553)
Jiatai Wang committed
06ecec7