Code

gguf : add 64-bit support (GGUF v2)#2821

Code

Merged

ggerganov merged 10 commits intomasterggerganov/llama.cpp:masterfrom gguf-64bitggerganov/llama.cpp:gguf-64bitCopy head branch name to clipboardAug 27, 2023

+164-27

Commits

Commits on Aug 26, 2023

• gguf : bump version to 2

  ggerganov

  committed
  5f1fffd

  Copy full SHA for 5f1fffd

• gguf : add support for 64-bit (no backwards comp yet)

  ggerganov

  committed
  4f0547e

  Copy full SHA for 4f0547e

• gguf : v1 backwards comp

  ggerganov

  committed
  3656b3c

  Copy full SHA for 3656b3c

• gguf.py : bump GGUF version

  ggerganov

  committed
  ba335ff

  Copy full SHA for ba335ff

• gguf.py : uint64_t on all lengths, sizes and counts, enums still uint32_t

  klosax

  authored
  be726c5

  Copy full SHA for be726c5

• gguf.py : string lengths uint32_t

  klosax

  authored
  bc3eaf2

  Copy full SHA for bc3eaf2

• gguf : update all counts to 64-bit

  ggerganov

  committed
  6d369a1

  Copy full SHA for 6d369a1

• gguf.py : string len uint64_t and n_dims uint32_t

  klosax

  authored
  09b6da7

  Copy full SHA for 09b6da7

• gguf : fix typo

  ggerganov

  committed
  b61b170

  Copy full SHA for b61b170

• llama.cpp : print gguf version

  klosax

  authored
  33a5517

  Copy full SHA for 33a5517