Datasets:
Tasks:
Text Generation
Modalities:
Text
Sub-tasks:
language-modeling
Languages:
English
Size:
10M - 100M
License:
Download cpp/source_code-train.parquet from shibing624/source_code: direct link, hf CLI and curl.
- Browser
- Download file 43.3 MB
-
https://huggingface.co/datasets/shibing624/source_code/resolve/99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/cpp/source_code-train.parquet
- Command line
-
hf download hf://datasets/shibing624/source_code@99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/cpp/source_code-train.parquet
-
curl -L -o source_code-train.parquet https://huggingface.co/datasets/shibing624/source_code/resolve/99fc6fbf0f6ff8dc482af9c3da67ee8a3e9a07c9/cpp/source_code-train.parquet
43.3 MB
- Xet hash:
- 5b4676b3d462c0b7cce606a354390c2d49818f9bc78050b15e99cde91fd60bb0
- Size of remote file:
- 43.3 MB
- SHA256:
- c3b7ab91048b7f195b25017f3b9b8833ead1b8e4db3efa9800a8c63479c2f538
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.