The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 324830859 bytes, limit is 300000000 bytes
Make sure that
1. the Parquet files contain a page index to enable random access without loading entire row groups2. otherwise use smaller row-group sizes when serializing the Parquet files
Error code: TooBigContentError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Jev-Bench
Public dataset backup: 300 rows; 0 external image files.
Conversations, answers, row order, IDs and RL extra_info are preserved. SFT/RL image paths are relocated; image bytes are archived unchanged. Image files are hashed by original path for naming, not deduplicated by image content.
Download and restore
hf download hansQAQ/Jev-Bench --repo-type dataset --local-dir ./Jev-Bench
python ./Jev-Bench/restore_jev_dataset.py --dataset-dir ./Jev-Bench
SearchWeave English 300-row version. Images are already embedded in test.parquet; no image archives are required.
- Downloads last month
- 56