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 module

qmatmul_k

Provides CPU kernels for K-quant block-wise quantized matrix multiplication.

Functions

• ​matmul_Q4_K: Computes a matrix multiplication with Q4_K block-quantized weights.
• ​matmul_Q4_K_pack_b: Packs Q4_K quantized weights into the blocked layout consumed by the compute kernel.
• ​matmul_Q6_K: Computes a matrix multiplication with Q6_K block-quantized weights.
• ​matmul_Q6_K_pack_b: Packs Q6_K quantized weights into the blocked layout consumed by the compute kernel.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit