Datasets:
Tasks:
Video Classification
Formats:
json
Languages:
English
Size:
100K - 1M
ArXiv:
Tags:
reward-model
video-generation
evaluation
preference-learning
temporal-reasoning
synthetic-data
License:
Download data/train.jsonl from stellalisy/IsoVRBench: direct link, hf CLI and curl.
- Browser
- Download file 132 MB
-
https://huggingface.co/datasets/stellalisy/IsoVRBench/resolve/main/data/train.jsonl
- Command line
-
hf download hf://datasets/stellalisy/IsoVRBench/data/train.jsonl
-
curl -L -o train.jsonl https://huggingface.co/datasets/stellalisy/IsoVRBench/resolve/main/data/train.jsonl
132 MB
- Xet hash:
- 8067899a335ad68f783b759fb210ee44e56b52b95ab818546e637b6121964f6e
- Size of remote file:
- 132 MB
- SHA256:
- 813a6b6d10bd569445d73591997c0ba56bc0523db6a553b3ac081e7308df846f
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.