Stability AI Releases Open-Source Large Language Model StableLM
Synthetic media startup Stability AI shared the first of a new collection of open-source large language models (LLMs) named StableLM this week. StableLM widens Stability’s portfolio beyond its popular Stable Diffusion text-to-image generative AI model and into producing text and computer code, setting itself up as a direct rival for OpenAI, Google, and other generative AI model developers.
Stability AI has been quietly working on LLMs with EleutherAI, but StableLM is far more comprehensive and widely available. The company released both a 3 billion and a 7 billion-parameter version of StableLM on GitHub and promised the imminent release of 15 billion and 65 billion-parameter options soon. Those interested can experiment with a fine-tuned version of StableLM as a chatbot through Hugging Face as well. You can see an example of what it can do on the right.
The number of parameters comes across as relatively limited in some ways. For comparison, OpenAI built GPT-3 with 175 billion parameters and GPT-4 with an unknown but larger number than that. StableLM’s condensed parameter size is counterbalanced by its training, according to the company. Stability used the Pile open-source dataset and its amalgamation of Wikipedia, Stack Exchange, and PubMed to train StableLM, but an expanded version with triple the number of tokens, 1.5 trillion tokens all told. The result gives what Stability AI calls a “surprisingly high performance in conversational and coding tasks” relative to its parameter size.
“With the launch of the StableLM suite of models, Stability AI is continuing to make foundational AI technology accessible to all. Our StableLM models can generate text and code and will power a range of downstream applications. They demonstrate how small and efficient models can deliver high performance with appropriate training,” Stability AI web designer Anel Islamovic wrote in a blog post about Stable LM. “Language models will form the backbone of our digital economy, and we want everyone to have a voice in their design. Models like StableLM demonstrate our commitment to AI technology that is transparent, accessible, and supportive.”
Stability AI is launching StableLM on the heels of Stable Diffusion XL (SDXL), its latest text-to-image model. SDXL is better at photorealistic images and can expand the scope of an image or fill in details missing from the text prompt. SDXL is being pitched at enterprise customers, the inverse of StableLM’s current experimental phase. Stability can afford to pursue both approaches after raising $101 million last year. StableLM and SDXL are part of Stability’s broader strategy for generative AI, including its acquisition of the company behind AI image manipulation service Clipdrop and its partnership of with digital collectible platform Revel.xyz on an image-to-animation tool called Animai.