What's in the RedPajama-Data-1T LLM training set

RedPajama is “a project to create leading open-source models, starts by reproducing LLaMA training dataset of over 1.2 trillion tokens”. It’s a collaboration between Together, Ontocord.ai, ETH DS3Lab, Stanford CRFM, …

Sheared LLaMA: Accelerating Language Model Pre-training via

Training data used to train LLM models

LaMDA to Red Pajama: How AI's Future Just Got More Exciting!

From ChatGPT to LLaMA to RedPajama: I'm Switching My Interest to

The Latest Open Source LLMs and Datasets

Fine-Tuning Insights: Lessons from Experimenting with RedPajama

RedPajama-Data-v2: An open dataset with 30 trillion tokens for

togethercomputer/RedPajama-Data-V2 · Datasets at Hugging Face

A Survey on Data Selection for Language Models

Finetuning an LLM: RLHF and alternatives (Part I)

Skill it! A Data-Driven Skills Framework for Understanding and

Web LLM runs the vicuna-7b Large Language Model entirely in your

Exploring 10m scraped Shutterstock videos used to train Meta's