Xingyu Liu 97e1752922

Merge branch 'main' into model_arch_cfg

2025-12-12 15:30:09 -08:00

..

assets

…

attention

[Kernel] Support CUDA Graphs in 3D Triton Attention Kernel (#28306)

2025-12-12 16:55:40 +01:00

benchmarks

[Misc] Consistent case for vllm bench serve results (#30403)

2025-12-10 09:44:02 -08:00

compilation

[compile] Parse compile range cache keys as Range during cache loading. (#30516)

2025-12-12 04:30:51 +00:00

config

Merge branch 'main' into model_arch_cfg

2025-12-12 15:30:09 -08:00

device_allocator

…

distributed

[LMCache] Relax lmcache version requirement (#30425)

2025-12-11 18:18:47 -09:00

engine

[Bug] Fix attention_backend arg string parsing (#30534)

2025-12-12 08:40:50 -07:00

entrypoints

[Frontend] Fixes anthropic streaming message_start usage nesting (#30266)

2025-12-12 16:28:54 +00:00

inputs

…

logging_utils

…

lora

…

model_executor

[MoE-FP8-modelopt] Add FlashInfer alignment padding for intermediate dimensions (#29748)

2025-12-12 20:42:32 +00:00

multimodal

[Fix] Update lazing loading of video loader backend (#30444)

2025-12-11 10:14:57 +00:00

platforms

[CPU][FIX] Fix build failures on Arm CPUs with torch nightly (#30481)

2025-12-12 02:09:25 +00:00

plugins

…

profiler

…

ray

…

reasoning

[IMPROVEMENT] Change MistralReasoningParser behavior (#30391)

2025-12-11 17:53:26 +01:00

third_party

…

tokenizers

[DeepSeek V3.2] Proper drop_thinking logic (#30490)

2025-12-12 15:01:06 +00:00

transformers_utils

Merge branch 'main' into model_arch_cfg

2025-12-12 15:30:09 -08:00

triton_utils

…

usage

…

utils

[Fix]Load kv-cache dtype from hf_quant_config.json automatically (#29980)

2025-12-12 11:27:47 -08:00

v1

[CI] Fix mypy for vllm/v1/executor (#30517)

2025-12-12 18:05:34 +00:00

vllm_flash_attn

…

__init__.py

…

_aiter_ops.py

…

_bc_linter.py

…

_custom_ops.py

[Attention] Use sparse prefill kernel for fp8 kv-cache in DeepSeek-v3.2 (#27532)

2025-12-12 05:57:47 -08:00

_ipex_ops.py

…

beam_search.py

…

collect_env.py

…

connections.py

…

env_override.py

…

envs.py

[Attention] Use sparse prefill kernel for fp8 kv-cache in DeepSeek-v3.2 (#27532)

2025-12-12 05:57:47 -08:00

forward_context.py

…

logger.py

Make the httpx logger less annoying when Transformers v5 is installed (#30480)

2025-12-11 15:44:56 +00:00

logits_process.py

…

logprobs.py

…

outputs.py

…

pooling_params.py

…

py.typed

…

sampling_params.py

…

scalar_type.py

…

scripts.py

…

sequence.py

…

tasks.py

…

tracing.py

…

version.py

…