Datasets:

Malfaro43
/

LargeConvo2048

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

A large dataset to train Churro. Every element is under 2048 tokens, when tokenized using the LLaMA Tokenizer.