Datasets:
Tasks:
Visual Question Answering
Formats:
parquet
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Download annotations/train.jsonl from benchmarkanon/logic_conflict: direct link, hf CLI and curl.
- Browser
- Download file 65.1 MB
-
https://huggingface.co/datasets/benchmarkanon/logic_conflict/resolve/main/annotations/train.jsonl
- Command line
-
hf download hf://datasets/benchmarkanon/logic_conflict/annotations/train.jsonl
-
curl -L -o train.jsonl https://huggingface.co/datasets/benchmarkanon/logic_conflict/resolve/main/annotations/train.jsonl
65.1 MB
- Xet hash:
- 6e9dbfd3b592f106802102cfdb22f9d1d73b5f08ae12ffb5e263b292db3b4993
- Size of remote file:
- 65.1 MB
- SHA256:
- aa8ff91fb5510dd14892ea09a00ee2736490aafd56fc432953a9ce60f0751ed1
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.