root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 18:34:36 +00:00
Code Issues Actions 7 Packages Projects Releases Wiki Activity

python/pytorch compat notes (#44)

Browse Source
This commit is contained in:
Oleksandr Nikitin 2023-03-12 14:16:33 +02:00 committed by GitHub
parent 129c7d1ea8
commit b9bd1d0141
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
README.md
View File

@ -136,6 +136,7 @@ ls ./models
65B 30B 13B 7B tokenizer_checklist.chk tokenizer.model
# install Python dependencies
# preferred versions: python 3.10 (not 3.11), torch 1.13.1+
python3 -m pip install torch numpy sentencepiece
# convert the 7B model to ggml FP16 format

1
convert-pth-to-ggml.py
View File

@ -86,6 +86,7 @@ for p in range(n_parts):
if (p > 0):
fname_out = sys.argv[1] + "/ggml-model-" + ftype_str[ftype] + ".bin" + "." + str(p)
# weights_only requires torch 1.13.1, remove this param or update if you get an "invalid keyword argument" error
model = torch.load(fname_model, map_location="cpu", weights_only=True)
fout = open(fname_out, "wb")