lyj-jjjandCursor 821286794f [HY-Imgae3.0] support hunyuan image3 dit fa-fp8 on npu (#3540) ...

Signed-off-by: lyj-jjj <liuyingjun5@huawei.com>
Co-authored-by: Cursor <cursoragent@cursor.com>

2026-05-18 09:35:30 -07:00

..

attention

[BugFix] fix(omni): isolate diffusion KV-cache dtype from vLLM --kv-cache-dtype #3585 (#3596)

2026-05-15 11:22:59 +08:00

cache

[Feat] Add helios support cache dit (#3470)

2026-05-16 21:11:34 +08:00

distributed

[Quantization] Redo Z-Image text encoder FP8 online quantization (#3279)

2026-05-06 16:37:36 +00:00

executor

[Bugfix] Support diffusion worker dead detect when use inline engine (#3214)

2026-05-16 20:50:29 +08:00

hooks

[Bugfix][HunyuanImage3.0] Fix KV reuse compatibility in SP scenarios (#3546)

2026-05-13 22:40:18 +08:00

layers

[ROCm] Bugfix wan22 (#3463)

2026-05-13 20:50:57 +08:00

lora

[Feature] Enable LoRA adapter injection for BAGEL (#2490)

2026-04-06 19:30:52 +08:00

model_loader

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

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

models

[HY-Imgae3.0] support hunyuan image3 dit fa-fp8 on npu (#3540)

2026-05-18 09:35:30 -07:00

offloader

[New Model]: Add sensenova u1 support (#3319)

2026-05-11 22:19:00 +08:00

postprocess

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

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

profiler

[Feature] support sp for hunyuan (#2163)

2026-03-27 17:28:13 +08:00

sched

Fix diffusion engine cleanup lifecycle (#3494)

2026-05-18 08:54:57 +08:00

utils

[Bugfix] enforce max_sequence_length for Qwen-Image and Wan2.2 series before encoding (#2847)

2026-04-17 10:40:36 +08:00

worker

[Bugfix]update process name for dit stage (#3602)

2026-05-16 20:51:31 +08:00

__init__.py

[Diffusion] Qwen image support (#82)

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

compile.py

[core] add torch compile for diffusion (#684)

2026-01-09 04:05:44 +13:00

config.py

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

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

data.py

[bugfix] Fix diffusers backend input bug after #2913 (#3644)

2026-05-15 09:55:09 +00:00

diffusion_engine.py

Fix diffusion engine cleanup lifecycle (#3494)

2026-05-18 08:54:57 +08:00

envs.py

[Refactor] Extract has_flash_attn_pkg in OmniPlatform (#3068)

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

forward_context.py

support online FP8 quantization for FA on NPU #2236 (#2640)

2026-05-13 11:46:00 +08:00

inline_stage_diffusion_client.py

[Bugfix] Support diffusion worker dead detect when use inline engine (#3214)

2026-05-16 20:50:29 +08:00

ipc.py

[Feature] Add trajectory recording to BAGEL denoising loop (#2483)

2026-04-08 10:43:01 +08:00

model_metadata.py

[Feature] HunyuanImage-3.0 IT2I: multi-image input + prompt API cleanup (#3444)

2026-05-14 16:58:40 +08:00

registry.py

[New Model]: Add sensenova u1 support (#3319)

2026-05-11 22:19:00 +08:00

request.py

[Feature] Port Bagel RDMA flow to latest main (#2000)

2026-04-08 14:55:35 +08:00

stage_diffusion_client.py

[Rebase] Rebase to vllm v0.21.0 (#3530)

2026-05-16 18:18:43 +08:00

stage_diffusion_proc.py

[Feat] DiffusionEngine Support async batch infer (#2729)

2026-05-06 09:47:13 +12:00