Chenggang Zhao a51629ddf9 Code polishing

2025-04-18 16:43:00 +08:00

..

include/deep_gemm

Code polishing

2025-04-18 16:43:00 +08:00

jit

Use swizzling instead of padding (#86)

2025-04-14 15:20:58 +08:00

jit_kernels

Support TMA multicast on B with m_grouped_gemm_contiguous.

2025-04-18 15:38:04 +08:00

__init__.py

fix typo

2025-02-26 18:37:22 +08:00

utils.py

Correctly flush L2, as reconstructing the tensors on every iteration effectively put them in the L2, and gave the GPU enough idle time to avoid thermal throttling in a potentially unrealistic way.

2025-03-15 20:46:24 +00:00