Commits vllm-project/vllm

[Frontend] Support multimodal inputs for late-interaction scoring (ColQwen3) + NewModel: nvidia/nemotron-colembed (#34574)

craftsangjae committed 30 days ago

Verified 5719a4e4

[Realtime] Add Qwen3-ASR realtime streaming support (#34613)

pougetat committed 30 days ago

Verified 11be2c74

[Kernel] Optimize sample_recovered_tokens_kernel (#34974)

xyang16 committed 30 days ago

Verified 7a5adad4

Support prompt_embeds for pooling requests in output processor (#34904)

laviier committed 30 days ago

Verified 59c62332

[Misc] Fix mypy errors in vllm/profiler and remove from exclude list (#34959)

taneem-ibrahim committed 30 days ago

Verified d38cd3dd

[ROCm][Bugfix]: Only save unpadded sizes for shared_experts in MoERunner to fix rmsnorm pad fusion (#34636)

Rohan138 committed 30 days ago

Verified ded333fb

[Kernel] [Helion] [9/N] Canonicalize GPU variant names to base model names (#34928)

gmagogsfm committed 30 days ago

Verified 9d7577b2

[CI/Build] Add opentelemetry libs in default vllm build (requirements/common.txt) (#34466)

vladmihailescu committed 30 days ago

Verified e739c29e

[LoRA] Support Quantized Adapters (#30286)

yugong333 committed 30 days ago

Verified a55caf6a

Revert "[Llama4,Quantization] Simplify and generalize logic for Q/K permutations in quantized self-attn layers " (#34997)

LucasWilkinson committed 30 days ago

Verified 0e22cd61

Bump Flashinfer Version and Re-enable DeepSeek NVFP4 AR+Norm Fusion (#34899)

wzhao18 committed 30 days ago

Verified ea5f903f

[AMD][CI] Fix test_custom_allreduce for A100 testgroup (#34735)

rjrock committed 30 days ago

Verified 0632ed87

[CI] Revert PRs 34818 and 33600 (#34979)

LucasWilkinson committed 30 days ago

Verified aaefc58e

[Test] Add FP8 KV Cache Testing for MLA Backends (#34473)

wzhao18 committed 30 days ago

Verified f24b2de3

[CI] Remove failing prime-rl integration test (#34843)

mgoin committed 30 days ago

Verified fac1507f

[compile] Fix torch.compile time discrepancy in logging. (#34912)

zhxchen17 committed 30 days ago

Verified f8639940

[compile] Move torch_aot_compile directory under torch_compile_cache (#34831)

zhxchen17 committed 30 days ago

Verified e4a5d8c6

[Kernel] [Helion] [6/N] Add num_tokens dimension to silu_mul autotuning and dispatching (#34185)

gmagogsfm committed 30 days ago

Verified a6d0299c

Ensure that MkDocs v2 does not get installed (#34958)

hmellor committed 30 days ago

Verified 6ce80f70

[perf] Avoid dtype promotion sync in mamba_get_block_table_tensor (#34870)

hl475 committed 30 days ago

Verified 1fe46216

[Refactor] Extract Harmony streaming SSE event builders into streaming_events.py (#34909)

sfeng33 committed 30 days ago

Verified ed31a020

[V0 Deprecation] Remove unused MM placeholders in request output (#34944)

DarkLight1337 committed 30 days ago

Verified f9ac1920

[BUGFIX] Fix `_dummy_run` missing `prepare_inputs_event` synchronization (#34866)

vadiklyutiy committed 30 days ago

Verified 59965aff

[Kernel] Optimize grouped topk kernel (#34206)

xyang16 committed 30 days ago

Verified b1c4f0b2

[Bugfix][Hardware][AMD] Fix ROCM_AITER_FA speculative decoding support (#32877)

c0de128 committed 31 days ago

Verified 8de7c636

[Minor] Add logging when using MXFP4 MXFP8 TRTLLM backend (#34916)

frankwang28 committed 31 days ago

Verified 05977923

[Models] LFM2: Support LoRA (#34921)

tianshu-Michael-yu committed 31 days ago

Verified ea37530b

[ROCm][CI] Loosen RemoteOpenAIServer Startup Timeout (#34922)

micah-wil committed 31 days ago

Verified f5432e35

[Misc] Add deprecated environment variable utilities (#33677)

carlory committed 31 days ago

Verified 07cab212

[CI][AMD][BugFix][P/D] Add default_vllm_config to test_moriio_connector.py so tests pass (#33739)

rasmith committed 31 days ago

Verified 0c1dc427