jondurbin/airoboros-70b-3.3

jondurbin/airoboros-70b-3.3 is a 70 billion parameter instruction-tuned causal language model developed by jondurbin, fine-tuned from Meta's Llama-3-70B-Instruct. This model is primarily trained on synthetic data generated by airoboros, alongside various other datasets, and is optimized for context-obedient question answering, summarization, and complex instruction following, including agent-style execution planning and function calling. It utilizes the Llama-3 instruct chat template and has an 8192 token context length, making it suitable for tasks requiring precise adherence to provided context and structured outputs.