L2T-NeurIPS-2023 / docs /instructions.md
Dakhoo's picture
still the very first commit
ba48b51
|
raw
history blame
748 Bytes

Instructions

1. Subset the Valid Files

Run the following script to subset the valid files:

./scripts/create_filtered_images_dataset.sh data/all_images_big data/all/ data/all_d/all.list

Use the following command to compress the files:

tar -cvzf all.tar.gz all/

2. Push Large Files

*.tar.* filter=lfs diff=lfs merge=lfs -text
*.jsonl filter=lfs diff=lfs merge=lfs -text

Execute the following command in the root directory of the git repository:

git lfs install
huggingface-cli lfs-enable-largefiles .
git add .
git commit -m "<commit message>"
git push

You can copy the above Markdown-formatted instructions and view them in any Markdown viewer to see the enhanced readability.