MaciejBalaNVandRoger Wang 13c2de684c Cosmos3 logging and prompt improvements (#6325) ...

Signed-off-by: Maciej Bala <mbala@nvidia.com>
Co-authored-by: Roger Wang <hey@rogerw.io>

2026-08-20 12:04:49 -07:00

..

attention

[CI/Build] Strengthen pre-commit with markdownlint, SPDX, and policy hooks (#6273)

2026-08-20 21:07:04 +08:00

cache

[Perf] Fuse MiniMax H3 modulation with FP32 accumulation (#6281)

2026-08-18 20:13:54 -07:00

diffusion_kv

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

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

distributed

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

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

executor

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

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

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

[ROCm] [CI] Fix circular import for rocm due to patch (#6287)

2026-08-18 09:45:05 +00:00

model_loader

[Core] Support online FP8 with DLO AllGather (#6279)

2026-08-18 23:48:07 +08:00

models

Cosmos3 logging and prompt improvements (#6325)

2026-08-20 12:04:49 -07:00

offloader

[REFACT]Refactor diffusion parallel state (#5531)

2026-08-19 11:23:10 +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

[Bugfix] Prevent DiffusionResultPump crash on cancelled futures (#5793) (#5983)

2026-08-15 16:19:45 +00:00

worker

[REFACT]Refactor diffusion parallel state (#5531)

2026-08-19 11:23:10 +08:00

__init__.py

[Diffusion] Qwen image support (#82)

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

compile.py

[Rebase] Rebase to vllm 0.27.0 (#5976)

2026-08-11 15:23:46 +00:00

config.py

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

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

data.py

[REFACT]Refactor diffusion parallel state (#5531)

2026-08-19 11:23:10 +08:00

diffusion_engine.py

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

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

envs.py

[Refactor] Extract has_flash_attn_pkg in OmniPlatform (#3068)

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

forward_context.py

[Attention] Refine TRTLLM attention support for MiniMax H3 (#5779)

2026-08-06 09:38:17 +00:00

inline_stage_diffusion_client.py

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

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

io_support.py

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

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

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 SANA-WM support(Stage-1 only) (#4061)

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

request.py

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

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

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

[Bugfix | Model] Fix SenseNova & Use Well-defined Model Configs (#5877)

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