Databricks has launched Dolly, a low-cost large language model (LLM) with impressive instruction-following abilities similar to ChatGPT.
By using Databricks, any business can train an open-source LLM to replicate ChatGPT’s extraordinary capabilities in as little as 30 minutes on one machine using high-quality training data.
Dolly functions by tweaking an open-source model with 6 billion parameters from EleutherAI to enhance its capacity for tasks like generating text and brainstorming through instruction following, with assistance from Alpaca data.
In a blog post titled ‘Hello Dolly: Democratizing the magic of ChatGPT with open models,’ Databricks expressed its opinion that models such as Dolly will democratise large language models, making them accessible to all companies and allowing them to customise and enhance their products.
Benefits of Open Models for businesses
There are several reasons why a company might opt to create their own model instead of sending its data to a centralised LLM provider that operates a proprietary model via an API.
Some businesses consider their most sensitive and proprietary intellectual property to be the problems and datasets that are most likely to benefit from AI, and handing this over to a third party might be undesirable.
Additionally, companies may have varying priorities when it comes to model quality, cost, and desired functionality. Databricks holds the view that ML users are better served in the long run by directly owning and controlling their models.
“We’re calling the model Dolly — after Dolly the sheep, the first cloned mammal — because it’s an open-source clone of an Alpaca, inspired by a LLaMA,” the company said in a blog post.
“We’re in the earliest days of the democratisation of AI for the enterprise, and much work remains to be done, but we believe the technology underlying Dolly represents an exciting new opportunity for companies that want to build their own instruction-following models cheaply.”
The release of Dolly is the first in a series of announcements Databricks is making that focus on helping every organisation harness the power of large language models. To learn more, see HERE.
Disclaimer from the company: Generative AI is an emerging technology, and we’re in the early stages of research on how to address factual accuracy, bias, offensive responses, general toxicity, and hallucinations in LLMs. Dolly, like other language models, can sometimes exhibit these behaviours, and we urge our users to exercise good judgment in designing applications of this technology.
Keep up to date with our stories on LinkedIn, Twitter, Facebook and Instagram.