t4a-private-gemm-x86
x86-64 matmul impl (tensor4all fork of private-gemm-x86; adds the masked-store offset clamp, a triangular-destination regression test and the diagonal reversal for negative depth strides)
Activity
- Latest release
- 15h ago
- Total releases
- 5
- Cadence
- ~daily
- Last 12 months
- 5
Details
- License
- MIT
- First release
- Oct 19, 2026
Releases