Datasets:
Tasks:
Translation
Modalities:
Text
Formats:
parquet
Size:
1M - 10M
ArXiv:
Tags:
code-to-text
License:
Download java/train/0000.parquet from google/code_x_glue_ct_code_to_text: direct link, hf CLI and curl.
- Browser
- Download file 141 MB
-
https://huggingface.co/datasets/google/code_x_glue_ct_code_to_text/resolve/refs%2Fconvert%2Fparquet/java/train/0000.parquet
- Command line
-
hf download hf://datasets/google/code_x_glue_ct_code_to_text@refs/convert/parquet/java/train/0000.parquet
-
curl -L -o 0000.parquet https://huggingface.co/datasets/google/code_x_glue_ct_code_to_text/resolve/refs%2Fconvert%2Fparquet/java/train/0000.parquet
141 MB
- Xet hash:
- e015d22a684ba0a241c1e620ef1dc9b085c7685c01facb55daae058186e1ab32
- Size of remote file:
- 141 MB
- SHA256:
- 9c1f85dd441e2ea99fd9c9d04e26784d75a0bffc6f1c91b5eb8499ab86fda20b
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.