← Back to KHAO

Tech ·

Adding a column or creating a subset for a given filter means that one has to scan and rewrite

2 min read

Compiled by KHAO Editorial — aggregated from 1 source. See llms.txt for citation guidance.

★ Tier-1 Source

Data pipeline: source datasets, prepare images, re-caption with a VLM, write Mosaic Data Shards.

These two formats play off each other throughout this post and the PRX data pipeline: Lance to build, MDS to stream.

Key facts

Summary

In one sentence: they assemble training data from a mix of public and internal datasets, re-caption the images with a VLM, and turn the result into the streamable corpus they trained PRX on. At a high level, the data pipeline looks like this:. The goal was to assemble a large, diverse dataset for pre-training. That is a problem of coverage and diversity, not of per-image perfection. The team assemble their pre-training data from a mix of public and internal datasets. Where a source already comes quality-filtered, deduplicated, and filtered for NSFW content and personal information, they build on that work instead of repeating it at scale. In their experience, what matters most for pre-training is to use long captions that accurately describe everything in the image. The team saw this directly in Part 2, where switching from short captions to long ones substantially improved sample quality.

Read full article at Hugging Face →