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
| Name | Last commit message | Last commit date |
|---|---|---|
| .. | ||
| adapters | ||
| attention | ||
| backends | ||
| engine | ||
| models | ||
| moe | ||
| prerouter | ||
| server | ||
| streaming | ||
| __init__.py 1.1 KB | ||
| __main__.py 171 B | ||
| cli.py 11.0 KB | ||
| config.py 1.2 KB | ||
| registry.py 4.3 KB | ||
| sampling.py 2.6 KB | ||