Llava Onevision 1.5: Mid-Training Webdataset for Vision-Language Models

A webdataset likely containing 3 million examples for training multimodal AI models, as indicated by the title. It was published by author mvp-lab on the Hugging Face platform and last updated on September 20, 2025. The dataset appears to be associated with the LLaVA (Large Language and Vision Assistant) project, suggesting it contains paired image-text data.

Use Cases

• Fine-tuning a vision-language model for instruction following (inferred from domain, verify after download)
• Pre-training a multimodal model's alignment module (inferred from domain, verify after download)
• Benchmarking model performance on visual question-answering tasks (inferred from domain, verify after download)

Strengths

• Published on the Hugging Face platform, a major repository for AI datasets and models.
• Last updated on 2025-09-20 04:46:39, indicating recent maintenance.
• Title suggests a scale of 3 million data points.

Limitations

• Metadata is minimal; actual content requires verification after download.
• Column-level documentation, sample data, file formats, license, and exact row count are unknown.
• Data may reflect the biases inherent to its source web crawl.

Provenance

Source

mvp-lab on Hugging Face.

Collection Method

Likely scraped or assembled from web sources, given the 'webdataset' label.

Time Range

null

Freshness

Last updated 2025-09-20 04:46:39.

Geography

null