Michael Goin 0c76bb2de1 [Bugfix] Disable cascade attention with FlashInfer (#26130) ...

Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Signed-off-by: yewentao256 <zhyanwentao@126.com>

2025-10-03 13:35:58 -07:00

..

assets

[Bugfix] Fix GLM4.1V multimodal processor with compatability for Transformers v4.56 (#24822)

2025-09-15 20:45:06 +08:00

attention

[BugFix] ChunkedLocalAttention is currently not CG compatible (#26034)

2025-10-03 13:35:57 -07:00

benchmarks

EAGLE 3: Fix preamble so that measured speedup over Eagle 1 becomes 32% instead of 5% on MTBench (#25916)

2025-10-03 13:35:58 -07:00

compilation

[torch.compile]: Add VLLM_DEBUG_DUMP_PATH environment variable (#25651)

2025-10-03 13:35:56 -07:00

config

[Perf] Fix and reapply move apply w8a8 block fp8 linear to class (#25696)

2025-10-03 13:35:58 -07:00

device_allocator

[sleep mode] save memory for on-the-fly quantization (#24731)

2025-09-12 11:25:19 +00:00

distributed

[Misc]allow disable pynccl (#25421)

2025-10-03 13:35:57 -07:00

engine

[Fix] Improve CPU backend compatibility for RISC-V (#25816)

2025-10-03 13:35:57 -07:00

entrypoints

[gpt-oss] use vLLM instead of openai types for streaming (#25186)

2025-10-03 13:35:57 -07:00

executor

[V0 Deprecation] Remove vllm.worker and update according imports (#25901)

2025-10-03 13:35:56 -07:00

inputs

fix: revert cast to cpu in MsgpackEncoder._encode_tensor to avoid hidden performance regressions (#25738)

2025-10-03 13:35:55 -07:00

logging_utils

Generate _ModelInfo properties file when loading to improve loading speed (#23558)

2025-10-03 13:35:53 -07:00

lora

[V0 deprecation] Clean up LoRA (#25686)

2025-10-03 13:35:55 -07:00

model_executor

Fix MTP with deepep_low_latency (#25904)

2025-10-03 13:35:58 -07:00

multimodal

[Small] Prevent bypassing media domain restriction via HTTP redirects (#26035)

2025-10-03 13:35:57 -07:00

platforms

[New Model] DeepSeek-V3.2 (Rebased to Main) (#25896)

2025-10-03 13:35:56 -07:00

plugins

Change log level from info to debug for IOProcessor (#24999)

2025-09-17 10:21:28 -07:00

profiler

[Docs] Fix warnings in vllm/profiler and vllm/transformers_utils (#25220)

2025-10-03 13:35:53 -07:00

ray

[Misc] Getting and passing ray runtime_env to workers (#22040)

2025-08-01 23:54:40 -07:00

reasoning

[Bugfix] Improve GLM4 MoE Reasoning Parser's is_reasoning_end Condition (#25355)

2025-10-03 13:35:56 -07:00

third_party

fix some typos (#24071)

2025-09-02 20:44:50 -07:00

transformers_utils

[New Model] DeepSeek-V3.2 (Rebased to Main) (#25896)

2025-10-03 13:35:56 -07:00

triton_utils

[Bugfix] Add triton.language.tensor placeholder (#25649)

2025-10-03 13:35:55 -07:00

usage

[Misc] Add logger.exception for TPU information collection failures (#20510)

2025-07-05 07:24:32 -07:00

utils

[Perf] Fix and reapply move apply w8a8 block fp8 linear to class (#25696)

2025-10-03 13:35:58 -07:00

v1

[Bugfix] Disable cascade attention with FlashInfer (#26130)

2025-10-03 13:35:58 -07:00

vllm_flash_attn

[BugFix] Fix vllm_flash_attn install issues (#17267)

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

__init__.py

[CI/Build] Add bc-linter to vLLM CI (#21234)

2025-09-11 15:34:36 -07:00

_bc_linter.py

[CI/Build] Add bc-linter to vLLM CI (#21234)

2025-09-11 15:34:36 -07:00

_custom_ops.py

[New Model] DeepSeek-V3.2 (Rebased to Main) (#25896)

2025-10-03 13:35:56 -07:00

_ipex_ops.py

[Chore] Remove ipex_ops warning (#24835)

2025-09-14 19:41:53 +00:00

beam_search.py

[Chore][V0 Deprecation] Move LogProb to a separate file (#24055)

2025-09-01 12:07:53 -07:00

collect_env.py

[BugFix] python collect_env.py and vllm collect-env compatibility with uv venv (#24066)

2025-09-10 21:25:33 +08:00

connections.py

[Small] Prevent bypassing media domain restriction via HTTP redirects (#26035)

2025-10-03 13:35:57 -07:00

env_override.py

[distributed][rl] remove nccl cumem env var override (#24141)

2025-09-03 06:45:25 +00:00

envs.py

[Small] Prevent bypassing media domain restriction via HTTP redirects (#26035)

2025-10-03 13:35:57 -07:00

forward_context.py

[Bugfix][WideEP] Apply TP Attn + EP MoE fix to other models (#24982)

2025-10-03 13:35:56 -07:00

logger.py

[Logging] allow config logging stream (#24336)

2025-09-10 15:02:01 +00:00

logits_process.py

[mypy] Fix wrong type annotations related to tuple (#25660)

2025-10-03 13:35:55 -07:00

logprobs.py

[Chore][V0 Deprecation] Move LogProb to a separate file (#24055)

2025-09-01 12:07:53 -07:00

outputs.py

[V0 Deprecation] Remove from_seq_group methods (#25330)

2025-10-03 13:35:53 -07:00

pooling_params.py

[Docs] Fix pooling-params doc references in openai_compatible_server.md (#24939)

2025-09-18 04:36:47 -07:00

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

Add backward compatibility for GuidedDecodingParams (#25422)

2025-10-03 13:35:54 -07:00

scalar_type.py

[Doc]: fix typos in Python comments (#24042)

2025-09-01 19:07:45 -07:00

scripts.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

sequence.py

[Misc] Simplify PoolerOutput and move to v1/outputs (#25629)

2025-10-03 13:35:55 -07:00

tasks.py

[V1] Get supported tasks from model runner instead of model config (#21585)

2025-07-25 05:36:45 -07:00

test_utils.py

Remove V0 Encoder-Decoder Support (#24907)

2025-09-15 21:17:14 -07:00

tracing.py

[V1] feat:add engine v1 tracing (#20372)

2025-09-11 17:10:39 -07:00

version.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00