]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: d8d4e86)
Remove torch GPU dependencies from the Docker.full image (#665)
authorbsilvereagle <redacted>
Sun, 2 Apr 2023 22:13:03 +0000 (15:13 -0700)
committerGitHub <redacted>
Sun, 2 Apr 2023 22:13:03 +0000 (00:13 +0200)
commita0c05164168297c04737936ad0cad849a512547a
tree30f6e731817e123d4402f95091aa3e6d320345c8tree
parentd8d4e865cd481b18f10508ffee35db903767ef5ccommit | diff
Remove torch GPU dependencies from the Docker.full image (#665)

By using `pip install torch --index-url https://download.pytorch.org/whl/cpu`
instead of `pip install torch` we can specify we want to install a CPU-only version
of PyTorch without any GPU dependencies. This reduces the size of the Docker image
from 7.32 GB to 1.62 GB
.devops/full.Dockerfile diff | blob | history
Packaging of ggml-org/llama.cpp