Databricks Launches Dolly 2.0, a Commercially Viable Open-Source ChatGPT Alternative
Databricks releases Dolly 2.0, an open-source generative AI model that businesses can use and modify for commercial purposes. The lightweight model allows companies to run ChatGPT-like capabilities on their own internal servers.
Databricks releases Dolly 2.0, an open-source generative artificial intelligence model that offers ChatGPT-like capabilities for commercial use. This new iteration addresses the overwhelming demand from businesses seeking an accessible AI that bypasses the paywalls and restrictions of centralized services. Unlike the initial version, Dolly 2.0 removes barriers that previously prevented commercial applications.
The model operates on 12 billion parameters, making it remarkably lightweight compared to massive models like GPT-3. Databricks fine-tunes Dolly 2.0 using a high-quality, crowdsourced dataset of 15,000 prompt and response pairs generated by its own employees. This efficient approach allows the AI to generate coherent responses while remaining small enough to run entirely on a company's internal servers.
By fully open-sourcing Dolly 2.0, including its training code and the databricks-dolly-15k dataset under a Creative Commons license, Databricks aims to democratize large language models. The company envisions transforming AI from an expensive luxury into a customizable commodity that any business can own. This strategy ensures organizations maintain complete data privacy without sharing sensitive information with third-party providers.