Max Hu 9c3fe9936b Flashinfer cuDNN backend for Qwen3 VL ViT attention (#34580) ...

Signed-off-by: Max Hu <maxhu@nvidia.com>
Signed-off-by: Max Hu <hyoung2991@gmail.com>
Co-authored-by: Max Hu <maxhu@nvidia.com>
Co-authored-by: Shang Wang <shangw@nvidia.com>

2026-02-27 20:20:23 +08:00

..

basic_correctness

[offloader] v2: Hide weight onloading latency via prefetching (#29941)

2026-02-25 17:20:59 -08:00

benchmarks

[Benchmark] Simplify SLA scan (#35306)

2026-02-25 22:35:41 -08:00

compile

[torch.compile] Sequence Parallelism threshold compile ranges (#28672)

2026-02-26 05:00:12 +00:00

config

[Bugfix] Exclude language_model_only key from MM AOT compile hash but include in model one (#34508)

2026-02-13 13:59:00 +00:00

cuda

[Bugfix] Fix CUDA compatibility path setting for both datacenter and consumer NVIDIA GPUs (#33992)

2026-02-25 18:15:51 -08:00

detokenizer

[CI/Build] Remove use of skip_v1 (#34699)

2026-02-18 12:19:11 +08:00

distributed

[Bugfix] Fix MessageQueue connect_ip for cross-node data parallelism (#35429)

2026-02-26 22:08:16 +00:00

engine

Turn @config into a dataclass_transform (#31541)

2026-02-03 17:40:59 +00:00

entrypoints

[Bugfix] Replace assert with ValueError for response_format validation in completions endpoint (#35456)

2026-02-27 08:01:06 +00:00

evals

[UX] Add --moe-backend arg for explicit kernel selection (#33807)

2026-02-25 17:44:44 -08:00

kernels

Flashinfer cuDNN backend for Qwen3 VL ViT attention (#34580)

2026-02-27 20:20:23 +08:00

lora

use 'max_active_experts' for moe lora input size (#33197)

2026-02-27 03:50:43 +00:00

model_executor

[Perf] Add opt-in SM100 Oink RMSNorm custom-op path (#31828)

2026-02-24 23:01:53 -08:00

models

[Model] Add nvidia/llama-nemotron-embed-vl-1b-v2 multimodal embedding model (#35297)

2026-02-26 14:17:17 +00:00

multimodal

[MM] Allow audio chunking for offline LLM (#34628)

2026-02-23 21:04:28 -08:00

plugins

[Misc] Update tests and examples for Prithvi/Terratorch models (#34416)

2026-02-13 23:03:51 -08:00

plugins_tests

(bugfix): Fixed encode in LLM entrypoint for IOProcessr plugin prompts (#34618)

2026-02-16 07:33:55 -08:00

prompts

…

quantization

[UX] Add --moe-backend arg for explicit kernel selection (#33807)

2026-02-25 17:44:44 -08:00

reasoning

fix(reasoning): Qwen3ReasoningParser returns truncated output as reasoning (#35230)

2026-02-26 20:30:45 +00:00

renderers

[Bugfix] Fix edge case in UUID data parsing (#34884)

2026-02-19 02:24:30 -08:00

rocm/aiter

[Bugfix][Hardware][AMD] Fix exception types in AITER MLA FP8 check (#31177)

2026-01-06 14:10:59 +08:00

samplers

[CI/Build] Try to make beam search test less flaky (#34885)

2026-02-19 19:16:58 +08:00

standalone_tests

[CI][BugFix] ShellCheck cleanup to remove baseline and preserve runtime behavior (#34514)

2026-02-17 12:22:56 +00:00

system_messages

[V1] Implement Cascade Attention (#11635)

2025-01-01 21:56:46 +09:00

tokenizers_

[Refactor] Simplify BOS/EOS token handling (#34435)

2026-02-12 18:18:24 -08:00

tool_parsers

Fix GLM4 parser tests (#34905)

2026-02-24 22:27:42 +08:00

tool_use

[Refactor] [6/N] to simplify the vLLM openai chat_completion serving architecture (#32240)

2026-01-13 13:01:39 +00:00

tools

Turn @config into a dataclass_transform (#31541)

2026-02-03 17:40:59 +00:00

tracing

[CI][Tracing] Fix race condition by adding server readiness check (#34364)

2026-02-16 12:57:39 +00:00

transformers_utils

[UX] Use gguf repo_id:quant_type syntax for examples and docs (#33371)

2026-01-31 12:14:54 +08:00

utils_

[Refactor] Clean up pooling serial utils (#33665)

2026-02-03 10:29:18 +00:00

v1

[Misc] Move GPUModelRunner.prepare_kernel_block_sizes to utils (#35400)

2026-02-27 11:42:51 +08:00

vllm_test_utils

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

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

weight_loading

[Nightly CI] Remove CT Model (#33530)

2026-02-01 19:09:09 -08:00

__init__.py

…

ci_envs.py

[Model][0/N] Improve all pooling task | clean up (#25817)

2025-10-13 16:44:50 +08:00

conftest.py

[CI] Heavy refactoring of Voxtral multimodal audio model tests (#34294)

2026-02-13 20:04:29 -08:00

test_access_log_filter.py

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

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

test_attention_backend_registry.py

[1/N][Attention] Restructure attention: move files (#31916)

2026-01-09 13:10:24 -08:00

test_config.py

[Bugfix][CPU] Fix basic unit tests failing in CPU platforms (#34677)

2026-02-25 08:36:15 +00:00

test_embedded_commit.py

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

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

test_envs.py

[Perf] Enable environment cache in EngineCore to enable the feature for UniProcExecutor as well (#29289)

2025-12-10 14:13:01 -05:00

test_inputs.py

[Refactor] Pass full VllmConfig to Renderer (#34485)

2026-02-12 22:48:38 -08:00

test_logger.py

[BugFix] respect VLLM_LOGGING_LEVEL in logger (#29761)

2025-12-02 07:49:16 +00:00

test_logprobs.py

[Core] Switch Flat logprob control from environment variable to SamplingParams (#28914)

2025-11-19 02:10:02 +00:00

test_outputs.py

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

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

test_pooling_params.py

[Refactor] Move task outside of PoolingParams.verify (#33796)

2026-02-05 09:33:11 +00:00

test_ray_env.py

[Ray] Propagate third-party env vars to Ray workers via prefix matching (#34383)

2026-02-17 01:08:42 -08:00

test_regression.py

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

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

test_scalartype.py

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

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

test_seed_behavior.py

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

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

test_sequence.py

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

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

test_triton_utils.py

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

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

test_version.py

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

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

test_vllm_port.py

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

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

utils.py

[ROCm][CI] Extending attention backend coverage for Eagle spec decode tests (#35265)

2026-02-25 14:16:18 -08:00