Datasets:
Tasks:
Text Generation
Modalities:
Text
Sub-tasks:
language-modeling
Languages:
English
Size:
10M - 100M
License:
Download java/source_code-train.parquet from shibing624/source_code: direct link, hf CLI and curl.
- Browser
- Download file 39 MB
-
https://huggingface.co/datasets/shibing624/source_code/resolve/99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/java/source_code-train.parquet
- Command line
-
hf download hf://datasets/shibing624/source_code@99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/java/source_code-train.parquet
-
curl -L -o source_code-train.parquet https://huggingface.co/datasets/shibing624/source_code/resolve/99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/java/source_code-train.parquet
39 MB
- Xet hash:
- 78343f62b172ee757ae8402b7d9fc2a84c33e116d8e7c427b8c9372678a6f3d4
- Size of remote file:
- 39 MB
- SHA256:
- 15956e83d8c526c915e6ba3025300d6eb76a2d776199283d6d3c7df30f943d16
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.