Stability AI, a leading AI technology company, has announced the launch of its StableLM suite of language models, which includes a range of models suitable for commercial or research purposes. The StableLM model is open-source, allowing developers to inspect, use, and adapt StableLM base models freely, subject to the terms of the CC BY-SA-4.0 license.
StableLM is built on the success of earlier language models, including GPT-J, GPT-NeoX, and the Pythia suite, which were trained on The Pile open-source dataset. The StableLM model is trained on a new experimental dataset built on The Pile but is three times larger, with 1.5 trillion tokens of content.
The Alpha version of the model is available in 3 billion and 7 billion parameters, with 15 billion to 65 billion parameter models to follow. StableLM models can generate text and code and will power a range of downstream applications, representing a transparent, open, and scalable alternative to proprietary AI. They demonstrate how small and efficient models can deliver high performance with appropriate training.
Despite its small size of 3 to 7 billion parameters (by comparison, GPT-3 has 175 billion parameters), StableLM performs surprisingly well in conversational and coding tasks, thanks to the richness of the dataset. Stability AI is also releasing a set of research models that are instruction fine-tuned, initially using a combination of five recent open-source datasets for conversational agents: Alpaca, GPT4All, Dolly, ShareGPT, and HH.
In 2022, Stability AI drove the public release of Stable Diffusion, a revolutionary image model that represents a transparent, open, and scalable alternative to proprietary AI. With the launch of the StableLM suite of models, Stability AI is continuing to make foundational AI technology accessible to all. Its open-source approach and flexibility will enable more developers to experiment with and develop advanced AI models, leading to more innovative and effective solutions in the future.
The release of StableLM is a significant milestone in the AI industry, as it marks the availability of yet another open-source language model that can be leveraged by developers worldwide. As an open-source model, StableLM is accessible to everyone, including researchers, startups, and established companies looking to develop new AI solutions.
We research, curate and publish daily updates from the field of AI. Paid subscription gives you access to paid articles, a platform to build your own generative AI tools, invitations to closed events and open-source tools.
Consider becoming a paying subscriber to get the latest!