|
TinyLlama.cpp 1.0
A lightweight C++ implementation of the TinyLlama language model
|
Data structures for GGUF (GPT-Generated Unified Format) file format. More...
#include <cstdint>#include <map>#include <string>#include <variant>#include <vector>#include <sys/mman.h>#include <sys/stat.h>#include <fcntl.h>#include <unistd.h>#include "ggml_types.h"

Go to the source code of this file.
Classes | |
| struct | GGUFArray |
| Represents an array in GGUF metadata. More... | |
| struct | GGUFHeader |
| Header structure for GGUF files. More... | |
| struct | GGUFTensorInfo |
| Information about a tensor stored in a GGUF file. More... | |
| struct | GGUFData |
| Complete representation of a GGUF file's contents. More... | |
Typedefs | |
| using | GGUFMetadataValue = std::variant< uint8_t, int8_t, uint16_t, int16_t, uint32_t, int32_t, float, bool, std::string, uint64_t, int64_t, double, GGUFArray > |
| Type for storing metadata values of various types. | |
Data structures for GGUF (GPT-Generated Unified Format) file format.
This file defines the structures used to represent and manipulate data stored in GGUF format files. GGUF is a format designed for storing large language models and their associated metadata efficiently.
Definition in file gguf_structs.h.
| using GGUFMetadataValue = std::variant<uint8_t, int8_t, uint16_t, int16_t, uint32_t, int32_t, float, bool, std::string, uint64_t, int64_t, double, GGUFArray> |
Type for storing metadata values of various types.
Uses std::variant to support multiple value types that can be stored in GGUF metadata, including numeric types, strings, and arrays.
Definition at line 56 of file gguf_structs.h.