050b8567a0
Drops the hardcoded f32 GLU kernels in favor of a single template. We now load/store in the native tensor type (half or float) to save memory bandwidth, but keep the actual ALU compute in float to avoid exploding math in geglu/swiglu. Also opened up the dispatch gate to allow f16 inputs. |
||
|---|---|---|
| .. | ||
| cmake | ||
| include | ||
| src | ||
| .gitignore | ||
| CMakeLists.txt | ||