Run 35B MoE models in ~2.5 GB of RAM. Weights stream from SSD; a trained router predicts loads a step ahead. Mac, iPhone, Android, Windows supported
Python
These Apache-2.0 templates are copied from the official Edge0 checkpoints for offline multi-turn tool-call regression tests. Model weights are not included.
- Ling: https://huggingface.co/Edge0/Edge0-8B-A1B-preview/blob/269b9a2c4a69d897c50e9f4e125328481d7c0fcf/chat_template.jinja
- Qwen: https://huggingface.co/Edge0/Edge0-35B-A3B-preview/blob/3fe15cbf2bd5bbcdfd611035e1ac88971ff1cdab/chat_template.jinja
The original model cards identify the license as Apache-2.0. Retain the templates' original comments and consult the repository LICENSE.