Datasets:
Download default/train/0000.parquet from Orion-zhen/github-python-code-fim: direct link, hf CLI and curl.
- Browser
- Download file 155 MB
-
https://huggingface.co/datasets/Orion-zhen/github-python-code-fim/resolve/refs%2Fconvert%2Fparquet/default/train/0000.parquet
- Command line
-
hf download hf://datasets/Orion-zhen/github-python-code-fim@refs/convert/parquet/default/train/0000.parquet
-
curl -L -o 0000.parquet https://huggingface.co/datasets/Orion-zhen/github-python-code-fim/resolve/refs%2Fconvert%2Fparquet/default/train/0000.parquet
155 MB
- Xet hash:
- cd69bf5c183cf3157a4b6928e2404bc56dda216cb0f2a5d60023f70dddf5752d
- Size of remote file:
- 155 MB
- SHA256:
- acf3975035fa26ee3c27899f60fbda12a6c73d0fc5e3c3ae1d1ecc6b43b41123
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.