DeepHermes-3-Llama-3-8B-Preview is an 8B parameter language model fine-tuned by NousResearch from the Llama 3 architecture. Hosted on Hugging Face, it features specialized chat templates for general use and tool calling, along with support for structured outputs and JSON mode. The model is designed for local or self-hosted inference and is intended for developers building applications that require reasoning, tool use, and code capabilities.
In the Foundation models & chat space, DeepHermes 3 Llama 3 8B Preview takes a focused approach. It focuses on needing a capable, open-weight language model for local inference, tool use, and specialized reasoning tasks. DeepHermes 3 Llama 3 8B Preview is an open-source project aimed at developers and AI researchers. DeepHermes 3 Llama 3 8B Preview is open source under the MIT license. DeepHermes 3 Llama 3 8B Preview is available on the web, the command line, and API, and it can be self-hosted.
Behind DeepHermes 3 Llama 3 8B Preview is NousResearch, and it first shipped in 2024. Development happens publicly on GitHub with 1.4k stars. Key capabilities include Chat Template, Tool Use Support, and JSON Mode.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do