DatasetMan/QGO-8B
Image-Text-to-Text • 9B • Updated
Datasets:
The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 336492670 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.
PM4Bench QGO training data
This is the synthetic multilingual OCR corpus used for QGO-8B GRPO:
- 19,500 training rows
- 500 validation rows
- embedded image bytes
- prompt, rule-based reward target, answer, split, and auxiliary metadata
The release builder preserves row order and learning content. It changes only
historical machine-specific image path strings to portable basenames. Per-shard
source and release SHA-256 values are recorded in
metadata/release_inventory.json.
The dataset is synthetic and released under Apache-2.0.
metadata/release_inventory.json records every shard's row count, byte size,
source digest, release digest, and the only permitted transformation. The
repository-level SHA256SUMS covers the complete public snapshot.
Training code and the exact reward implementation are available at https://github.com/opendatalab/PM4Bench/tree/main/recipes/qgo.
- Downloads last month
- -
