The Common Pile

The Common Pile: The largest dataset ever built exclusively from public & openly licensed text.

The authors Kandpal et al (2025) also release “Comma v0.1-1T and -2T, two performant 7-billion-parameter LLMs trained on text from the Common Pile”.

“produces models comparable to those trained on an equivalent amount of unlicensed data. Ultimately, we believe that the Common Pile v0.1 represents the first step on the path towards a more ethical language model ecosystem, where performance need not come at the cost of creator rights and legal transparency.”

Abstract
“Large language models (LLMs) are typically trained on enormous quantities of unlicensed text, a practice that has led to scrutiny due to possible intellectual property infringement and ethical concerns. Training LLMs on openly licensed text presents a first step towards addressing these issues, but prior data collection efforts have yielded datasets too small or low-quality to produce performant LLMs. To address this gap, we collect, curate, and release the Common Pile v0.1, an eight terabyte collection of openly licensed text designed for LLM pretraining. The Common Pile comprises content from 30 sources that span diverse domains including research papers, code, books, encyclopedias, educational materials, audio transcripts, and more. Crucially, we validate our efforts by training two 7 billion parameter LLMs on text from the Common Pile: Comma v0.1-1T and Comma v0.1-2T, trained on 1 and 2 trillion tokens respectively. Both models attain competitive performance to LLMs trained on unlicensed text with similar computational budgets, such as Llama 1 and 2 7B. In addition to releasing the Common Pile v0.1 itself, we also release the code used in its creation as well as the training mixture and checkpoints for the Comma v0.1 models.”

https://arxiv.org/pdf/2506.05209