llama.cpp merges support for DFM Mimir 1B, GGUF weights on Hugging Face

noctrex · reddit · 2026-09-16

A community contribution by Reddit user noctrex has been merged into llama.cpp release b11003, adding support for the open-source DFM Mimir 1B model. Quantized GGUF weights are available on Hugging Face (noctrex/DFM-Mimir), with additional optimizations planned.

Original post →

More from Models

Models channel →