Jack HuandDouglas Lehr 2cab53ddee [Model][Hardware][AMD]: Part 1/2 -> Enable e2e QK Norm + RoPE + KV Cache runtime fusion for Qwen3-30B-A3B on ROCM_AITER_FA, and ROCM_AITER_UNIFIED_ATTN (#42749) ...

Signed-off-by: Jack Hu <Jack.Hu@amd.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>

2026-07-16 17:39:04 -05:00

..

__init__.py

[Model] Add DiffusionGemma Support (#45163)

2026-06-11 22:17:35 -07:00

attention.py

[1/N] Add dense MHA path for sparse MLA short sequences (#47327)

2026-07-14 00:29:56 +00:00

cache.py

[2/N][Core] support partial prefix cache hit for hybrid model (#46384)

2026-07-12 05:37:51 +00:00

compilation.py

[Model][Hardware][AMD]: Part 1/2 -> Enable e2e QK Norm + RoPE + KV Cache runtime fusion for Qwen3-30B-A3B on ROCM_AITER_FA, and ROCM_AITER_UNIFIED_ATTN (#42749)

2026-07-16 17:39:04 -05:00

device.py

platforms: add uses_cpu_device() hook to Platform for DeviceConfig (#42313)

2026-05-12 12:39:17 -07:00

diffusion.py

[Model] Add DiffusionGemma Support (#45163)

2026-06-11 22:17:35 -07:00

ec_transfer.py

[Misc] Introduce ec_both role EC (encoder cache) connector (#34182)

2026-02-10 18:55:35 +00:00

kernel.py

BF16x3 router GEMM (#47973)

2026-07-16 17:04:20 +08:00

kv_events.py

[Mypy] Better fixes for the mypy issues in vllm/config (#37902)

2026-03-25 06:14:43 -07:00

kv_transfer.py

[Connector] Remove P2pNcclConnector (#44854)

2026-06-08 18:58:29 +08:00

load.py

[Bugfix][Config] Keep pydantic validation for fields with a TYPE_CHECKING Literal alias (#46220)

2026-06-24 12:25:50 +00:00

lora.py

[LoRA] Support 2D and 3D MoE LoRA adapter at the same time (#42242)

2026-05-18 15:22:26 +08:00

mamba.py

[Mamba] Flashinfer selective_state_update (#36162)

2026-04-14 15:10:58 -04:00

model_arch.py

[Model] Support Unlimited OCR (#46564)

2026-06-27 23:09:18 -07:00

model.py

[Model] Add Inkling model support [1/N] (#48799)

2026-07-15 23:40:07 -07:00

multimodal.py

Vram semaphore infra (#44465)

2026-06-26 17:32:51 -07:00

observability.py

[Feat] Add runtime monitor for post-warmup CuTeDSL compilation (#46167)

2026-06-23 23:33:17 +00:00

offload.py

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

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

parallel.py

[CI Bug] Fully solve accuracy issue for DSv3.2 + MTP + Sequence Parallel (#48036)

2026-07-14 10:00:24 -04:00

pooler.py

Log fully resolved pooling config at startup (#48030)

2026-07-14 22:00:46 +00:00

profiler.py

Remove need for explicit \n in docstring lists for --help formatting (#38350)

2026-03-27 08:38:00 -07:00

quantization.py

[Quant] Add nvfp4_per_token online MoE quantization (#48538)

2026-07-16 14:25:27 -07:00

reasoning.py

[Feature] Add auto-detection for reasoning_config when only reasoning_parser is set (#38214)

2026-04-10 01:36:26 +00:00

scheduler.py

[Bugfix] Reject negative values for max_logprobs and long_prefill_token_threshold (#44002)

2026-06-30 13:01:03 +01:00

speculative.py

[Spec Decode][DSpark] Add Gemma4-12B DSpark draft model (#47216)

2026-07-16 21:51:47 +00:00

speech_to_text.py

support hotwords for FunASR model (#39674)

2026-04-22 02:25:06 -07:00

structured_outputs.py

Remove unused disable_fallback field (#36546)

2026-03-09 20:56:54 -07:00

utils.py

[EPLB] Mask padding in EPLB load recording (#38128)

2026-06-28 19:43:58 -07:00

vllm.py

[Model][Hardware][AMD]: Part 1/2 -> Enable e2e QK Norm + RoPE + KV Cache runtime fusion for Qwen3-30B-A3B on ROCM_AITER_FA, and ROCM_AITER_UNIFIED_ATTN (#42749)

2026-07-16 17:39:04 -05:00

weight_transfer.py

Weight sync refactor + move sparse nccl engine (#44353)

2026-07-01 01:25:19 -07:00