Firehawk52-backup/llama
28
0
Fork 0
mirror of https://github.com/meta-llama/llama.git synced 2026-01-15 16:32:54 -03:00
Code Issues Actions Packages Projects Releases Wiki Activity

merge

Browse Source
This commit is contained in:
Ubuntu
2023-09-07 01:25:45 +00:00
parent 7706271705 2a61ead637
commit ad4ff84cca
2 changed files with 18 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@@ -34,6 +34,8 @@ In a conda env with PyTorch / CUDA available, clone the repo and run in the top-
pip install -e .
```
Then to run the script: `./download.sh`. There are optional ENV variables that can be used for automation.
## Inference
Different models require different model-parallel (MP) values:

21
download.sh
View File

@@ -5,10 +5,21 @@
set -e
read -p "Enter the URL from email: " PRESIGNED_URL
echo ""
read -p "Enter the list of models to download without spaces (7B,13B,70B,7B-chat,13B-chat,70B-chat), or press Enter for all: " MODEL_SIZE
TARGET_FOLDER="." # where all files should end up
# Check if PRESIGNED_URL environment variable is set
if [ -z "$PRESIGNED_URL" ]; then
read -p "Enter the URL from email: " PRESIGNED_URL
fi
# Check if MODEL_SIZE environment variable is set
if [ -z "$MODEL_SIZE" ]; then
read -p "Enter the list of models to download without spaces (7B,13B,70B,7B-chat,13B-chat,70B-chat), or press Enter for all: " MODEL_SIZE
fi
# Check if TARGET_FOLDER environment variable is set
if [ -z "$TARGET_FOLDER" ]; then
TARGET_FOLDER="." # Default target folder
fi
mkdir -p ${TARGET_FOLDER}
if [[ $MODEL_SIZE == "" ]]; then
@@ -67,4 +78,4 @@ do
else
(cd ${TARGET_FOLDER}"/${MODEL_PATH}" && md5sum -c checklist.chk)
fi
done
done