Wei ZhaoandWei Zhao 2427094152 [Feature] Support EPLB for DeepSeek v4 Mega Moe (#43339) ...

Signed-off-by: wzhao18 <wzhao18.sz@gmail.com>
Co-authored-by: Wei Zhao (Engrg-Hardware 1) <weizha@login-lyris01.lyris.clusters.nvidia.com>

2026-06-02 10:56:44 -07:00

..

__init__.py

[MoE Refactor] Remove SharedFusedMoE class (#35782)

2026-04-21 18:12:12 -04:00

argparse_utils.py

[Misc] Support Human-readable (k/K/m/M..) json cli arg (#40473)

2026-04-23 09:42:23 +02:00

async_utils.py

[Bugfix] Offload blocking tokenizer ops to shared thread pool to unblock event loop (#34789)

2026-03-26 22:17:00 -07:00

cache.py

[Docs] Mock all imports for docs (#27873)

2025-11-01 10:02:23 +00:00

collection_utils.py

[Plugin] Simplify IO Processor Plugin interface (#34236)

2026-02-10 19:47:39 -08:00

counter.py

[Misc] Separate out utils.counter and move utils.Device to engine (#27588)

2025-10-28 12:20:46 +00:00

cpu_resource_utils.py

[BugFix][MyPy]: Module has no attribute "sched_getaffinity" [attr-defined] (#41465)

2026-05-05 13:20:37 +00:00

cpu_triton_utils.py

[CPU] Fix spec decode kernel signatures for synthetic mode compatibility (#41932)

2026-05-10 12:07:15 +00:00

deep_gemm.py

[Feature] Support EPLB for DeepSeek v4 Mega Moe (#43339)

2026-06-02 10:56:44 -07:00

flashinfer.py

Integrate flashinfer b12x MoE and FP4 GEMM kernels for SM120/121 (#40082)

2026-05-20 17:21:11 +00:00

func_utils.py

[ROCm][Engine] Fix GPU memory leaks in engine shutdown and test workaround for async KV prefix cache reset (#38503)

2026-04-25 05:25:20 +00:00

gc_utils.py

[Doc]: fixing typos in multiple files. (#29685)

2025-11-28 08:41:41 -08:00

hashing.py

[Core] Add xxHash as a high-performance hash option for accelerating prefix caching (#29163)

2025-12-03 16:06:57 +00:00

import_utils.py

[MM][CG] Profile encoder CUDA graph pool memory (#41714)

2026-06-02 12:27:34 +08:00

jsontree.py

[mypy] Pass type checking for vllm/utils and vllm/v1/pool (#29666)

2025-11-28 20:40:47 +08:00

math_utils.py

[Refactor] Remove unused utils (#38153)

2026-03-26 17:08:19 +08:00

mem_constants.py

[Tests] Add Qwen3-VL multimodal memory leak check (#39268)

2026-04-09 04:54:46 -07:00

mem_utils.py

[Bugfix] Use is_integrated to detect UMA GPUs for memory reporting (#35356)

2026-04-13 11:07:40 -07:00

mistral.py

[Feature] Avoid eager import of the "mistral_common" package. (#40043)

2026-04-24 02:49:16 +00:00

multi_stream_utils.py

[DSV4] Add knob to enable pre-attn gemm (#41443)

2026-05-01 12:23:17 -07:00

nccl.py

[mypy] Pass type checking for vllm/utils and vllm/v1/pool (#29666)

2025-11-28 20:40:47 +08:00

network_utils.py

[Fix] Add missing space in IP fallback warning (#40359)

2026-04-20 20:26:06 -07:00

numa_utils.py

[Misc][NUMA] Auto-bind to PCT priority cores on DGX B300 + widen EngineCore across shard NUMA nodes (#43270)

2026-05-29 10:07:44 +08:00

numa_wrapper.sh

[Feature] NUMA binding support for GPU workers (#38635)

2026-04-08 09:55:24 -07:00

nvtx_pytorch_hooks.py

[Feature] Add Layer-wise NVTX Support (#29990)

2025-12-05 11:20:07 +00:00

ompmultiprocessing.py

[CPU Backend] Improve cpu thread utilization (#42666)

2026-05-18 03:04:41 -07:00

platform_utils.py

[CPU] Experimentally enable Triton and MRV2 (#43225)

2026-05-22 01:48:17 -07:00

print_utils.py

[Examples][1/n] Resettle basic examples. (#35579)

2026-03-08 20:22:53 -07:00

registry.py

[mypy] Pass type checking for vllm/utils and vllm/v1/pool (#29666)

2025-11-28 20:40:47 +08:00

serial_utils.py

[Frontend] Add multimodal support to /inference/v1/generate endpoint (#38405)

2026-04-17 20:31:56 -07:00

system_utils.py

[Frontend] DP Supervisor (#40841)

2026-05-21 20:33:16 -07:00

tensor_schema.py

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

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

torch_utils.py

[Perf][3/n] Eliminate GPU<->CPU syncs in attention impls (#41434)

2026-05-07 19:44:24 -07:00

tqdm_utils.py

[Renderer] Move InputPreprocessor into Renderer (2/2) (#34560)

2026-02-17 05:29:01 -08:00