hsliuustc0106 7db3a0cd5e [Performance] Add per-component DLO offload policy ...

Signed-off-by: hsliuustc0106 <liuhongsheng4@huawei.com>

2026-08-26 07:23:36 +08:00

..

attention

[perf] Add mask-free TRTLLM packed-padding path (#6542)

2026-08-25 07:45:21 +00:00

cache

[Feature][MiniMax-H3] Support diffusion continuous batching (#5810)

2026-08-24 21:08:53 +08:00

diffusion_kv

[Diffusion] Add native paged KV backend for diffusion workers (#6102)

2026-08-22 21:45:45 +08:00

distributed

[Bugfix] Use dedicated WORLD group for distributed VAE communication (#6401)

2026-08-20 16:13:34 +08:00

executor

[Core] Add UniProc diffusion executor for single-GPU (#6308)

2026-08-22 21:58:21 +08:00

hooks

[Refactor][Misc] Remove unused code in diffusion hooks/model_loader (#5270)

2026-07-22 09:45:17 +08:00

layers

[MUSA] Fix SwiGLU to use fused op (#6364)

2026-08-20 14:19:14 +08:00

lora

[Feature] Support MiniMax-H3 Turbo LoRA with the legacy manager (#6476)

2026-08-24 06:21:42 +08:00

model_loader

[Diffusion] Register HWR mmap for direct DLO H2D (#6591)

2026-08-25 23:30:11 +08:00

models

[Performance] Add per-component DLO offload policy

2026-08-26 07:23:36 +08:00

offloader

[Performance] Add per-component DLO offload policy

2026-08-26 07:23:36 +08:00

postprocess

[Bugfix] Fix RIFE device selection for CPU-transported videos (#2876)

2026-04-17 19:19:34 +08:00

profiler

[Bugfix] Report stepwise Qwen-Image profiler metrics (#4517)

2026-07-03 04:32:22 +00:00

quantization

[Bugfix] Fix HSDP compatibility with the new online FP8 linear method (#5677)

2026-08-14 15:45:55 +08:00

sched

[Diffusion] Add native KV cache initialization and Scheduler-managed block allocation (#6094)

2026-08-18 07:00:24 +00:00

utils

[Model][Frontend] MiniMax-H3: Add opt-in planar video response encoding (#6288)

2026-08-22 04:04:14 +00:00

worker

[Diffusion] Register HWR mmap for direct DLO H2D (#6591)

2026-08-25 23:30:11 +08:00

__init__.py

[Diffusion] Qwen image support (#82)

2025-11-28 14:22:11 +08:00

compile.py

fix: keep DLO hooks outside regional compilation (#6073)

2026-08-22 21:44:03 +08:00

config.py

[Diffusion][Attention] Support per-role attention backend via CLI (#2681)

2026-05-10 11:59:05 +08:00

data.py

[Performance] Add per-component DLO offload policy

2026-08-26 07:23:36 +08:00

diffusion_engine.py

[Bugfix] Make the async-output wait bound configurable and default it higher (#6255)

2026-08-22 22:10:30 +08:00

envs.py

[Refactor] Extract has_flash_attn_pkg in OmniPlatform (#3068)

2026-04-27 04:07:34 +00:00

forward_context.py

[Feature][MiniMax-H3] Support diffusion continuous batching (#5810)

2026-08-24 21:08:53 +08:00

inline_stage_diffusion_client.py

[Diffusion] Fix DLO DP concurrent request execution (#5864)

2026-08-08 08:24:42 +08:00

io_support.py

[Model] Add native SANA-Video 2B T2V and I2V support (#5508)

2026-08-25 09:46:02 +08:00

ipc.py

[Bugfix] Fix lost async diffusion outputs in request-level batching (#6023)

2026-08-15 18:27:51 +08:00

model_metadata.py

[Model] Add SANA-WM support(Stage-1 only) (#4061)

2026-08-18 23:57:55 -06:00

output_formatter.py

[Metrics] Add Diffusion Metrics (#4755)

2026-08-13 06:59:13 +00:00

prompt_update.py

[Frontend] Add midway prompt update for streaming video generation (#4652)

2026-07-27 14:07:43 +08:00

registry.py

[Model] Add native SANA-Video 2B T2V and I2V support (#5508)

2026-08-25 09:46:02 +08:00

request.py

[Model] Add native SANA-Video 2B T2V and I2V support (#5508)

2026-08-25 09:46:02 +08:00

stage_diffusion_client.py

[Frontend] Add midway prompt update for streaming video generation (#4652)

2026-07-27 14:07:43 +08:00

stage_diffusion_proc.py

[Refact] Refactor diffusion output streaming (#5166)

2026-07-23 10:06:08 +08:00

vllm_config.py

[Diffusion] Add native paged KV backend for diffusion workers (#6102)

2026-08-22 21:45:45 +08:00