ember.cpp 0.1.0
C++ inference engine for Llama-family and paperkiln models (CPU, CUDA, WebAssembly)
Loading...
Searching...
No Matches
block_q3_K Struct Reference

3-bit K-quantized block structure More...

#include <quantization.h>

Collaboration diagram for block_q3_K:
Collaboration graph

Public Attributes

uint8_t hmask [GGML_QK_K/8]
 
uint8_t qs [GGML_QK_K/4]
 
uint8_t scales [12]
 
uint16_t d
 
uint16_t dmin
 

Detailed Description

3-bit K-quantized block structure

Stores weights quantized to 3 bits with block-wise scaling. Balances compression and precision between Q2_K and Q4_K.

Definition at line 116 of file quantization.h.

Member Data Documentation

◆ d

uint16_t block_q3_K::d

Block scale

Definition at line 120 of file quantization.h.

◆ dmin

uint16_t block_q3_K::dmin

Block minimum value

Definition at line 121 of file quantization.h.

◆ hmask

uint8_t block_q3_K::hmask[GGML_QK_K/8]

High bit masks

Definition at line 117 of file quantization.h.

◆ qs

uint8_t block_q3_K::qs[GGML_QK_K/4]

Quantized values

Definition at line 118 of file quantization.h.

◆ scales

uint8_t block_q3_K::scales[12]

Sub-block scales

Definition at line 119 of file quantization.h.


The documentation for this struct was generated from the following file: