Commit History

expose ggml's get_rows, and rename the package to match the repo
e9319be

marcsun13 HF Staff commited on

Fail when a shipped artifact is not a loadable library
9b43109

marcsun13 HF Staff commited on

Load the kernel in tests the way a consumer does
1df3b99

marcsun13 HF Staff commited on

Metal: cover every quant type upstream's metallib has a gemv for
0becb0a

Marc Sun commited on

GGUF kernels: dequantize + fused gemv over packed blocks, 9 CUDA variants
2c20074
verified

marcsun13 HF Staff commited on