To use git lfs and huggingface hub: | |
Do once: | |
(1) install git lfs on machine | |
(2) pip install --upgrade huggingface_hub | |
Then: | |
(1) huggingface-cli login | |
Then: | |
(1) git lfs install | |
(2) git clone "https://huggingface.co/datasets/refugee-law-lab/canadian-legal-data" | |
(3) cd into repo | |
For specific new big files tracking using lfs (i.e. new / moved files) | |
(1) git lfs track "d:/xxx/xxx/bigfile.parquet" | |
(2) if over 5gb: huggingface-cli lfs-enable-largefiles . | |
When ready to push: | |
(1) git add . | |
(2) git commit -m "commit message here" | |
(3) git push | |
Notes: | |
- If making a new dataset, easiest to make first on huggingface (make sure to make dataset not model) | |
- If you add files in parquet, json, etc, huggingface will do everything automatically (make sure to include "train" in the file names) | |
- Note that the url includes /datasets/ (forgot that and it was confusing) | |
- | |