Phi-3.5-mini is a lightweight, state-of-the-art open model built upon datasets used for Phi-3 - synthetic data and filtered publicly available websites - with a focus on very high-quality, reasoning dense data. The model belongs to the Phi-3 model family and supports 128K token context length. The model underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures.

This dataset is the compilation of structured output and function calling data used in the Hermes 2 Pro series of models.

This repository contains a structured output dataset with function-calling conversations, json-mode, agentic json-mode and structured extraction samples, designed to train LLM models in performing function calls and returning structured output based on natural language instructions. The dataset features various conversational scenarios where AI agents are required to interpret queries and execute appropriate single or multiple function calls.

The synthetic data generation was led by @interstellarninja in collaboration with @NousResearch, @teknium, @THEODOROS and many others who provided guidance.

Hermes Function-calling Standard enables creation of LLM agents that are capable of executing API calls directly from user instructions. For instance, when asked to "find a flight from New York to Los Angeles for next Friday," a function-calling agent can interpret the request, generate the necessary function call (e.g., search_flights), and return the results. These agents significantly enhance the utility of AI by enabling direct interactions with APIs, making them invaluable in digital assistants across various domains.

For a complete useage guide of models trained on this data, see our github repo: https://github.com/NousResearch/Hermes-Function-Calling