For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo function

matmul_Q6_K_pack_b

def matmul_Q6_K_pack_b(b_tt: TileTensor[DType.uint8, Storage=b_tt.Storage, linear_idx_type=b_tt.linear_idx_type], b_packed_tt: TileTensor[DType.uint8, Storage=b_packed_tt.Storage, linear_idx_type=b_packed_tt.linear_idx_type])

Packs Q6_K quantized weights into the blocked layout consumed by the compute kernel.

Args:

b_tt (TileTensor[DType.uint8, Storage=b_tt.Storage, linear_idx_type=b_tt.linear_idx_type]): Source tensor holding the unpacked Q6_K quantized weights.
b_packed_tt (TileTensor[DType.uint8, Storage=b_packed_tt.Storage, linear_idx_type=b_packed_tt.linear_idx_type]): Destination tensor for the packed weights.