StableCode
v1.0
The StableCode large model is a tool specifically used to develop and implement artificial intelligence algorithms and applications. It can run on multiple deep learning frameworks and is suitable for tasks such as image classification, target detection and semantic segmentation.
Stability AI is best known for its Stable Diffusion text-generated image model, but that's not its only area of interest, as the company is now branching out into code generation as well. Today, Stability AI released its new open large language model (LLM) StableCode, which is designed to help users generate programming language code and is based on the transformer neural network.
StableCode will offer three different levels: a base model for general use, an instruction model, and a long context window model capable of supporting up to 16,000 tokens.
The StableCode model benefits from the initial programming language data set of the open source BigCode project, and also undergoes additional filtering and fine-tuning by Stability AI. StableCode will initially support development in programming languages such as Python, Go, Java, JavaScript, C, markdown and C++.
Training any LLM relies on data, and the data for StableCode comes from the BigCode project. HuggingFace and ServiceNow also launched open StarCoder LLM based on BigCode in May. Nathan Cooper, chief research scientist at Stability AI, said StableCode's training involves extensive filtering and cleaning of BigCode data.
StableCode's long context window has a context window of 16,000 tokens, which Stability AI claims is larger than any other model. The longer context window enables the use of more professional and complex code generation prompts. It also means that users can let StableCode view A medium-sized code base containing multiple files to help understand and generate new code.
StableCode is still in its early stages, and the official goal now is to see how developers will accept and use the model.
