ferrox-quant
Quantized weight formats and fused dequant kernels for Ferrox.
Block layouts (Q4_K, Q8_0, …), repacking, and CPU fused dequant+dot used by
ferrox-core and the Metal/CUDA backends.
Quantized weight formats and fused dequant kernels for Ferrox.
Block layouts (Q4_K, Q8_0, …), repacking, and CPU fused dequant+dot used by
ferrox-core and the Metal/CUDA backends.