Datasets:
Download default/train/0007.parquet from AlgorithmicResearchGroup/arxiv_python_research_code: direct link, hf CLI and curl.
- Browser
- Download file 139 MB
-
https://huggingface.co/datasets/AlgorithmicResearchGroup/arxiv_python_research_code/resolve/refs%2Fconvert%2Fparquet/default/train/0007.parquet
- Command line
-
hf download hf://datasets/AlgorithmicResearchGroup/arxiv_python_research_code@refs/convert/parquet/default/train/0007.parquet
-
curl -L -o 0007.parquet https://huggingface.co/datasets/AlgorithmicResearchGroup/arxiv_python_research_code/resolve/refs%2Fconvert%2Fparquet/default/train/0007.parquet
139 MB
- Xet hash:
- 19b168bc83bde89ac4d03305de9c80bfe55637e08603fee9fb7c57e122caf13f
- Size of remote file:
- 139 MB
- SHA256:
- 444d4f3b2074e7994b856cf5d9e2cec2160c29c484f83437bbcb041281945cd3
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.