biondizzle c56f5dda7e fix: use UINT8 TMA for packed FP4 instead of 16U4_ALIGN8B ...

The 16U4_ALIGN8B TMA data type is not supported on this driver
(CUDA_ERROR_INVALID_VALUE). Use UINT8 TMA to load raw bytes and let
the UMMA descriptor interpret SMEM as packed FP4 for mxf4nvf4.
TMA dimensions stay in bytes (like UINT8).

2026-05-12 18:05:11 +00:00

..

apis

fix: add kInt8 dtype support to TMA descriptor + change activation tensors to kInt8

2026-05-11 22:54:47 +00:00

indexing

[Public release 26/04] Introducing Mega MoE, FP4 Indexer and other features/fixes (#304)

2026-04-17 09:45:14 +08:00

jit

fix: gran_k=16 in transform_sf + sm_100a arch for NVFP4 mega_moe

2026-05-11 16:11:11 +00:00

jit_kernels

fix: use UINT8 TMA for packed FP4 instead of 16U4_ALIGN8B

2026-05-12 18:05:11 +00:00

utils

[Public release 26/04] Introducing Mega MoE, FP4 Indexer and other features/fixes (#304)

2026-04-17 09:45:14 +08:00

python_api.cpp

feat: NVFP4 mega MoE kernel (scale_vec::4X, UE4M3 block scales)

2026-05-11 05:41:08 +00:00