vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Introduce GGML migration tool for new file format

Browse source 
If you deleted your old Meta LLaMA .pth files, then the
migrate-ggml-2023-03-30-pr613.py script will allow you to convert your
old ggml files into the new mmap()'able format.

See #613
This commit is contained in:
Justine Tunney 2023-03-30 05:42:56 -07:00
parent 6f23ba5ee2
commit ee0c40dd6d
3 changed files with 326 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

8
convert-pth-to-ggml.py
View file

@ -1,4 +1,4 @@
# Convert a LLaMA model checkpoint to a ggml compatible file
# Convert a LLaMA model checkpoint to a ggjt compatible file
#
# Load the model using Torch
# Iterate over all variables and write them to a binary file.
@ -52,8 +52,8 @@ GGML_BLCK_SIZE = {
}
GGML_TYPE_SIZE = {
GGML_TYPE_Q4_0: 4 + QK/2,
GGML_TYPE_Q4_1: 4*2 + QK/2,
GGML_TYPE_Q4_0: 4 + QK//2,
GGML_TYPE_Q4_1: 4*2 + QK//2,
GGML_TYPE_I8: 1,
GGML_TYPE_I16: 2,
GGML_TYPE_I32: 4,
@ -245,11 +245,9 @@ def main():
fname_model = f"{dir_model}/consolidated.00.pth"
fname_out = f"{dir_model}/ggml-vocab.bin"
print(f"Extracting only the vocab from '{fname_model}'\n")
model = torch.load(fname_model, map_location="cpu")
with open(fname_out, "wb") as fout:
write_header(fout, hparams, ftype)
write_tokens(fout, tokenizer)
del model
print(f"Done. Output file: {fname_out}\n")
return