Commits vllm-project/vllm

m3 engine parser (reasoning-only) stage

BugenZhao committed 125 days ago

Verified 7fc60bb2

Merge branch 'main' into minimax-m3-streaming-reasoning-text-markers-main

Palaiologos1453 committed 126 days ago

Verified 1f2c614c

[Bugfix][Model Runner V2] Fix min_tokens off-by-one in the V2 GPU sampler (#46243)

Sunt-ing committed 126 days ago

Verified 183a430c

[Bugfix] Fix NVFP4/OCP MX MoE emulation (#46254)

mawong-amd committed 126 days ago

Verified a346d589

[Core] Ensure memory is pinned prior to async h2d copy (#45424)

njhill committed 126 days ago

Verified 7df3d7da

[Misc] Fix stale doc URL and docstring module path (#35530)

umut-polat committed 126 days ago

Verified 8dd1b702

[Render] Add reasoning/tool parsing to /derender + fix byte-fallback FFFD (#45919)

aoshen02 committed 126 days ago

Verified f57ac274

[Perf] Skip/shrink all_token_ids copy in scheduler for non-async and V2 runner (#45840)

amanchugh89 committed 126 days ago

Verified 6e919960

[SimpleCPUOffloadConnector] PCP + DCP support (#39831)

jonathanc-n committed 126 days ago

Verified c88d3d47

[Perf][KVConnector][Mooncake] Compact chunk-hash keys and zero-copy lookup wire format (#45969)

ivanium committed 126 days ago

Verified ab7fcbdd

[KV-Offloading] : Expose CPU cache usage metric (#45737)

varun-sundar-rabindranath committed 126 days ago

Verified 3b4a76b6

[KV Offload] Support packed HMA KV cache layout (#46205)

LucasWilkinson committed 126 days ago

Verified cc22621b

[Bugfix] Move extract_layer_index back inside is_v32 guard (#46199)

tlrmchlsmth committed 126 days ago

Verified 77148992

[Frontend] Report cache usage in Anthropic /v1/messages API (#40912)

zhangshuoming990105 committed 126 days ago

Verified 891cc4b9

[ROCm] [Bugfix] Bugfix ROCm Sparse Indexer (#46222)

tjtanaa committed 126 days ago

Verified 1bdf9810

Stop setting CUDA_VISIBLE_DEVICES internally in vLLM, add device_ids arg (#45026)

tlrmchlsmth committed 126 days ago

Verified ebfbcfe4

[Bugfix] Guard model_config access in _log_compilation_config (#46198)

tlrmchlsmth committed 126 days ago

Verified e9de72fe

[Perf] Optimize Qwen3-VL multi-video prompt processing (#46026)

Sirius29 committed 127 days ago

Verified d272418f

Revert "Fix Stale Encoder Cache After Weight Update" (#46125)

SumanthRH committed 127 days ago

Verified 7ff7f5c8

Merge branch 'main' into minimax-m3-streaming-reasoning-text-markers-main

Palaiologos1453 committed 127 days ago

Verified f24d8d5b

[Hardware][AMD][CI] Fix e2e core test group (#46024)

mawong-amd committed 127 days ago

Verified dced2907

[Bugfix] Fix gridDim.y overflow for large row counts (#45255)

JasonLi314 committed 127 days ago

Verified 93bad119

[ROCm] Fix VRAM not freed in test_phi3v (#46046)

djramic committed 127 days ago

Verified 0fbf42af

[ROCm][CI] Skip Qwen3.5-35B-A3B-MXFP4-AITER-TP2 for non gfx950 (#46109)

charlifu committed 127 days ago

Verified e6cd8913

[Bugfix][Parser] Fix U+FFFD leak at reasoning-to-content transition in engine parsers (#46159)

bbrowning committed 127 days ago

Verified 859e4d43

[ROCm][CI] Pin `test_rocm_compressed_tensors_w8a8` to TRITON_ATTN (#46180)

micah-wil committed 127 days ago

Verified 4a083cc8

Move CI failure diagnosis docs into ci-fails-buildkite skill (#45975)

vadiklyutiy committed 127 days ago

Verified ca7e1f2c

[ROCm] Use vLLM's fp8 quant max in AITER hipBLASLt accuracy test (#46176)

djramic committed 127 days ago

Verified dec860fb

Fix dead link in docs (#46181)

hmellor committed 127 days ago

Verified 0a49fb2b

[Test] Migrate test_openai_schema.py to schemathesis 4.x (#46173)

bbrowning committed 127 days ago

Verified 4a8abf37