OLMo3-7B-Instruct-SFT

An open 7-billion parameter language model from Ai2, fine-tuned on instruction data to serve as a transparent foundation for conversational AI.

OLMo3-7B-Instruct-SFT is an open-weight, 7-billion parameter language model developed by the Allen Institute for AI (Ai2). Built as part of the transparent Olmo 3 family, this checkpoint represents the supervised fine-tuning (SFT) stage of the model flow, trained on the Dolci-Instruct-SFT dataset. By releasing the weights, data recipes, and training code, Ai2 provides researchers with a fully inspectable pipeline for multi-turn chat, instruction following, and tool use.

https://huggingface.co/allenai/Olmo-3-7B-Instruct-SFT