The release of StableLM, a collection of open source large language models, was announced by Stability AI.
Stability AI, the company behind the well-known image-generation tool Stable Diffusion, has introduced a set of open source language-model tools, adding to the growth of the large-language-model market.
The publicly accessible alpha versions of the StableLM suite, which has models with 3 billion and 7 billion parameters, are now available. Models with 15 billion, 30 billion, and 65 billion parameters are listed as “in progress,” while a model with 175 billion parameters is slated for future development.
In contrast, OpenAI’s GPT-4 has a parameter count estimated to be 1 trillion, which is 6 times greater than that of GPT-3. However, as Stability AI pointed out in its blog post announcing the release of StableLM: “The parameter count may not be a suitable indicator of large-language-model (LLM) effectiveness.
“StableLM is trained on a new experimental dataset built on The Pile, but three times larger with 1.5 trillion tokens of content. […] The richness of this dataset gives StableLM surprisingly high performance in conversational and coding tasks, despite its small size of 3 to 7 billion parameters.”
The StableLM models’ level of robustness is still unknown. On the Stability AI team’s GitHub website, they promised to release further details regarding the LLMs’ capabilities, including model definitions and training parameters. The introduction of a potent, open-source alternative to OpenAI’s ChatGPT could be fascinating for the cryptocurrency trading community, should the models test well enough.
People are developing sophisticated trading bots on top of the GPT API and new variants that include third-party tool access, such as BabyAGI and AutoGPT, as Cointelegraph reported.
Open-source models being included in the mix may be advantageous for tech-savvy traders who don’t want to pay OpenAI’s access fees. Interested parties can try out a live interface for the HuggingFace-hosted 7 billion parameter StableLM model. However, attempts to do so as of the time of this article’s publication revealed the website overloaded or at capacity.
