The Drummond-1b1-Instruct is a 1.1 billion parameter instruction-following language model developed by Corre Social, fine-tuned from TucanoBR/Tucano-1b1-Instruct. Optimized for Portuguese (PT-BR), this model specializes in generating structured chain-of-thought reasoning before providing final answers. It is designed for tasks requiring structured reasoning with low computational overhead, utilizing a 2048-token context window.
No reviews yet. Be the first to review!
Hugging Face