minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[diffusion] feat: dispatch fp8 companions in mixed NVFP4 checkpoints (#36066)

Browse Source
This commit is contained in:
Mick
2026-08-25 11:26:37 +08:00
committed by GitHub
parent 833be86c15
commit 67853c5804
5 changed files with 32 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/cookbook/diffusion/MiniMax/MiniMax-H3.mdx
+3 -2
View File
@@ -319,8 +319,9 @@ compute capability 10.0 or newer. Pass a pruned FL2VA / Ref2VA file such as
`Abiray/Minimax-H3-nvfp4-INT4-INT8-Convrot/MiniMax_H3_FL2VA_pruned_nvfp4.safetensors`
to `--transformer-weights-path` and omit `--quantization`. SGLang infers the
packed group size and Comfy scale layout from the checkpoint; FSDP is rejected.
Mixed files may mark selected linears as `int8_tensorwise`; SGLang dispatches
those layers to the serialized Kitchen INT8 ConvRot path automatically.
Mixed files may mark selected linears as `int8_tensorwise` or dynamic/static
FP8; SGLang dispatches those layers to their serialized Kitchen INT8 or native
FP8 path automatically.
### Advanced: precomputed AdaLN cache