OmniCorpus-CC: 988 Million Image-Text Interleaved Documents

OmniCorpus-CC is a unified multimodal corpus of 10 billion-level images interleaved with text. It contains 988 million image-text interleaved documents collected from Common Crawl. The dataset was created by OpenGVLab and was last updated on the platform in March 2025.

Use Cases

• Train multimodal large language models based on the interleaved image-text structure.
• Develop visual question answering systems based on the described image-text pairs.
• Conduct research on cross-modal representation learning based on the web-crawled image-text documents.
• Fine-tune image captioning models based on the large-scale collection of paired data.

Strengths

• Contains 988 million image-text interleaved documents.
• Part of a larger corpus targeting 10 billion-level images.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for this specific subset is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

OpenGVLab

Collection Method

Collected from Common Crawl.

Freshness

Last updated 2025-03-20 12:32:06; freshness should be verified.