Research
Hugging Face paper PrismaCoder releases 9B and 35B open multimodal code models trained on a 2M-sample dataset
The PrismaCoder authors release PrismaCoder-9B and PrismaCoder-35B along with PrismaCoder-2M, a dataset they describe as the largest open multimodal code dataset. They claim the models lead open state of the art and rival leading proprietary systems, based on the authors' own reported results.
Every outlet that ran it 1 article
Hugging Face PapersResearchLead