Datasets:
Tasks:
Text Retrieval
Modalities:
Text
Formats:
parquet
Languages:
code
Size:
10K - 100K
ArXiv:
License:
Download python-corpus/test/0000.parquet from mteb/CodeSearchNetRetrieval: direct link, hf CLI and curl.
- Browser
- Download file 397 kB
-
https://huggingface.co/datasets/mteb/CodeSearchNetRetrieval/resolve/refs%2Fconvert%2Fparquet/python-corpus/test/0000.parquet
- Command line
-
hf download hf://datasets/mteb/CodeSearchNetRetrieval@refs/convert/parquet/python-corpus/test/0000.parquet
-
curl -L -o 0000.parquet https://huggingface.co/datasets/mteb/CodeSearchNetRetrieval/resolve/refs%2Fconvert%2Fparquet/python-corpus/test/0000.parquet
397 kB
- Xet hash:
- e281c720d0f3a38b1b7793b4ac475beab859c0bf93b3dc86955aef7fc3b6c84e
- Size of remote file:
- 397 kB
- SHA256:
- c3ffed7d172b9ef900d0c41a5081a42ff364f0e93ae156154bf314609709c962
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.