tzhouam 0dd5f3e17d Use blockwise Flux offload within single-L4 memory capacity ...

Signed-off-by: tzhouam <tzhouam@connect.ust.hk>

2026-10-03 22:34:10 +08:00

..

accuracy

Fit weekly serving fixtures to assigned GPUs and initialize accuracy groups

2026-10-03 22:29:17 +08:00

features

[Model] YuE2: async single-stage synthesis and CUDA graph acceleration (#8407)

2026-10-02 05:17:06 -07:00

offline_inference

[Perf][AuK] Graph the encoder prefill, cache reference latents and fuse the codec activation (#8305)

2026-10-02 15:43:55 +00:00

online_serving

Use blockwise Flux offload within single-L4 memory capacity

2026-10-03 22:34:10 +08:00

__init__.py

[ROCm] [CI] Add ROCm Dockerfile and AMD CI (vllm v0.12.0) (#280)

2025-12-21 11:51:37 +08:00