vllm/vllm at f1cb9b554492bd198ea63cf2233f8599aa850723 - vllm

assets

[Model] Nemotron Parse 1.1 Support (#30864 )

2026-01-05 13:00:14 -08:00

benchmarks

[Misc] support arbitrary MM datasets in spec dec bench (#33486 )

2026-02-02 08:49:48 +00:00

compilation

[fix][torch.compile] Fix cold-start compilation time increase by adding kv cache update to splitting ops (#33441 )

2026-01-31 06:48:34 -08:00

config

[torch.compile] Don't do the fast moe cold start optimization if there is speculative decoding (#33624 )

2026-02-03 03:38:49 +00:00

device_allocator

refactor: find_loaded_library (#31866 )

2026-01-07 06:42:20 +00:00

distributed

[Misc] Replace Optional[X] with X | None syntax (#33332 )

2026-01-30 01:56:59 -08:00

engine

Reduce the kernel overhead when num of active loras is smaller than max loras. Multiple cuda graphs are captured for each num of active-loras. (#32005 )

2026-02-02 12:30:06 -05:00

entrypoints

[Frontend] Add sampling parameters to Responses API (#32609 )

2026-02-03 13:51:10 +08:00

grpc

[grpc] Support gRPC server entrypoint (#30190 )

2026-01-07 23:24:46 -08:00

inputs

[Chore] Remove redundant input parsing methods (#33542 )

2026-02-02 10:50:47 +00:00

kernels

[Kernel] [Helion] [3/N] Helion kernel registry (#33203 )

2026-01-31 15:38:46 +08:00

logging_utils

[Logging] add --disable-access-log-for-endpoints CLI option (#30011 )

2026-01-26 21:49:03 +00:00

lora

Reduce the kernel overhead when num of active loras is smaller than max loras. Multiple cuda graphs are captured for each num of active-loras. (#32005 )

2026-02-02 12:30:06 -05:00

model_executor

Fix quantized Falcon-H1 model loading issues (#32728 )

2026-02-02 22:31:27 -08:00

multimodal

[Bugfix] Fix mm budget setting for Qwen Omni models (#33634 )

2026-02-03 04:56:25 +00:00

platforms

[UX] Format attention backend log line (#33570 )

2026-02-02 18:57:12 +00:00

plugins

[Misc] HF Hub LoRA Resolver (#20320 )

2026-01-26 13:56:32 +00:00

profiler

[Misc] Replace Optional[X] with X | None syntax (#33332 )

2026-01-30 01:56:59 -08:00

ray

[Small] Capture AttributeError when checking ray dependency. (#29024 )

2025-11-20 22:54:10 -08:00

reasoning

[Models] Step-3.5-Flash (#33523 )

2026-02-02 10:21:18 +08:00

renderers

[Refactor] Make Renderer an abstract class (#33479 )

2026-02-01 10:36:30 +08:00

third_party

[FlashMLA] Update FlashMLA to expose new arguments (#32810 )

2026-01-21 22:02:39 -07:00

tokenizers

Improve Mistral format checks. (#33253 )

2026-01-30 06:23:33 -08:00

tool_parsers

[Bugfix] GLM-4 tool parser: incremental string streaming (#33218 )

2026-02-02 11:13:31 +08:00

transformers_utils

[Voxtral Realtime] Introduce global log mel max (#33574 )

2026-02-02 17:01:47 -05:00

triton_utils

[mypy] Enable type checking for more directories (#29674 )

2025-11-28 08:39:27 -08:00

usage

[Deprecation] Remove deprecated environment variables (#32812 )

2026-01-22 02:25:16 +00:00

utils

[Redo] #33110 with threading limit (#33502 )

2026-02-01 09:18:11 +00:00

v1

Reduce the kernel overhead when num of active loras is smaller than max loras. Multiple cuda graphs are captured for each num of active-loras. (#32005 )

2026-02-02 12:30:06 -05:00

vllm_flash_attn

[BugFix] Fix vllm_flash_attn install issues (#17267 )

2025-04-27 17:27:56 -07:00

__init__.py

[V0 Deprecation] Remove V0 executors (#27142 )

2025-10-21 11:09:37 -07:00

_aiter_ops.py

fix[ROCm]: Remove unconditional aiter import (#32902 )

2026-02-02 22:10:02 +08:00

_bc_linter.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633 )

2025-10-12 09:51:31 -07:00

_custom_ops.py

[Bugfix][ROCm] Fixing the skinny gemm dispatch logic from #32831 (#33366 )

2026-01-30 19:05:23 -06:00

_ipex_ops.py

[Misc] Remove unused custom ops copy_blocks and copy_blocks_mla (#30967 )

2025-12-23 18:22:35 -08:00

beam_search.py

[Misc] Replace Optional[X] with X | None syntax (#33332 )

2026-01-30 01:56:59 -08:00

collect_env.py

[Misc] Add Helion version check to collect_env (#32797 )

2026-01-21 21:54:46 +00:00

connections.py

[Misc] Replace urllib's urlparse with urllib3's parse_url (#32746 )

2026-01-22 16:37:15 +08:00

env_override.py

Make mypy opt-out instead of opt-in (#33205 )

2026-01-29 09:12:26 +00:00

envs.py

[Hardware][SM100] Add TRTLLM Kernel for INT4 W4A16 Kernel. (#32437 )

2026-01-30 10:30:46 -08:00

exceptions.py

[OpenAI] Extend VLLMValidationError to additional validation parameters (#31870 )

2026-01-07 14:45:49 +00:00

forward_context.py

[torch.compile] Don't do the fast moe cold start optimization if there is speculative decoding (#33624 )

2026-02-03 03:38:49 +00:00

logger.py

Make mypy opt-out instead of opt-in (#33205 )

2026-01-29 09:12:26 +00:00

logits_process.py

[Misc] Refactor tokenizer interface (#29693 )

2025-11-29 04:02:21 -08:00

logprobs.py

Make mypy opt-out instead of opt-in (#33205 )

2026-01-29 09:12:26 +00:00

model_inspection.py

[UX] Add vLLM model inspection view (#29450 )

2026-01-09 10:12:35 -07:00

outputs.py

[Feature] add session based streaming input support to v1 (#28973 )

2026-01-24 12:06:28 -08:00

pooling_params.py

[Deprecation] Remove deprecated items related to pooling (#33477 )

2026-01-31 08:44:40 -08:00

py.typed

Add py.typed so consumers of vLLM can get type checking (#1509 )

2023-10-30 14:50:47 -07:00

sampling_params.py

[bugfix] Fix online serving crash when text type response_format is received (#26822 )

2026-01-16 12:23:54 +08:00

scalar_type.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633 )

2025-10-12 09:51:31 -07:00

scripts.py

Convert formatting to use ruff instead of yapf + isort (#26247 )

2025-10-05 07:06:22 -07:00

sequence.py

[Chore] Remove more V0 dead code from sequence.py (#31783 )

2026-01-06 10:25:14 +00:00

tasks.py

[Realtime API] Adds minimal realtime API based on websockets (#33187 )

2026-01-30 18:41:29 +08:00

tracing.py

[Misc] Move utils to avoid conflicts with stdlib, and move tests (#27169 )

2025-10-19 05:20:55 -07:00

version.py

Convert formatting to use ruff instead of yapf + isort (#26247 )

2025-10-05 07:06:22 -07:00