Datasets:
Download nodes/swe_polybench_python/0000.parquet from AmazonScience/SpIDER-Bench: direct link, hf CLI and curl.
- Browser
- Download file 111 MB
-
https://huggingface.co/datasets/AmazonScience/SpIDER-Bench/resolve/refs%2Fconvert%2Fparquet/nodes/swe_polybench_python/0000.parquet
- Command line
-
hf download hf://datasets/AmazonScience/SpIDER-Bench@refs/convert/parquet/nodes/swe_polybench_python/0000.parquet
-
curl -L -o 0000.parquet https://huggingface.co/datasets/AmazonScience/SpIDER-Bench/resolve/refs%2Fconvert%2Fparquet/nodes/swe_polybench_python/0000.parquet
111 MB
- Xet hash:
- 9db806ab3b4084a4e7cc886d0c147ce1402d575631ada192c5e19b9be52e33c2
- Size of remote file:
- 111 MB
- SHA256:
- 06cab1b9ccce2d8612f6bc84101a72c764daf1e7028accedc107202aa6bcafcd
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.