Gahow Wang d29c39d74e fix: GEMV NaN bug — skip custom kernel for small N (<256) ...

The custom launch_gemv_bf16 kernel produces NaN when output dimension N
is small (e.g. N=32 for the MoE router). Fall back to cuBLAS GemmEx for
N < 256. Also removes the padding workaround in gpt_oss MoE forward.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-05-30 15:20:04 +08:00

..

src

fix: GEMV NaN bug — skip custom kernel for small N (<256)

2026-05-30 15:20:04 +08:00

Cargo.toml

fix(xserv-chat): UTF-8/CJK-aware line input

2026-05-29 11:36:54 +08:00