Databricks’ Dolly 2 is an open-source large language model available on GitHub:

https://github.com/databrickslabs/dolly

This repository contains the code, model weights, and the fine-tuning dataset used to train Dolly 2. You can use this repository to:

• Download the model: Dolly 2 is available in several sizes (dolly-v2-12b, dolly-v2-7b, and dolly-v2-3b). You can download the model weights and use them with your preferred machine learning framework.
• Fine-tune the model: You can use the provided dataset or your own data to fine-tune Dolly 2 for specific tasks.
• Learn more about Dolly 2: The repository also includes documentation and tutorials to help you get started with using and fine-tuning Dolly 2.

There are also other resources available online for working with Dolly 2, such as:

• Colab notebooks: Several Colab notebooks are available that allow you to run Dolly 2 on Google’s Colab infrastructure.
• Hugging Face: Dolly 2 is available on Hugging Face, where you can easily download and use it with the Transformers library.