biondizzle e0814eb54e fix: cast expert_offsets to int32 for CuTeDSL kernel ...

CuTeDSL's grouped GEMM uses int32 for expert offsets internally.
Our cumsum produced int64, causing a type mismatch inside a dynamic
if-branch (prev_off changes from Int32 to Int64).

Also cast tokens_per_expert to int32 before cumsum.

2026-05-16 07:15:57 +00:00

..

patches

fix: missing tqdm import in weight_loader

2026-05-16 06:31:14 +00:00

nvfp4_cutedsl.py

fix: cast expert_offsets to int32 for CuTeDSL kernel

2026-05-16 07:15:57 +00:00