Quantized GGUF builds of EsDrac-v1-7B — the first language model that speaks Mallorcan (mallorquí) — for running locally with llama.cpp, Ollama or LM Studio, no GPU required.
File
Quant
Size
Notes
EsDrac-v1-7B-Q8_0.gguf
Q8_0
~8 GB
near-lossless, best quality
Run with llama.cpp
llama-cli -m EsDrac-v1-7B-Q8_0.gguf -cnv
Run with Ollama
ollama run hf.co/natzx94/EsDrac-v1-7B-GGUF:Q8_0
See the base model card for capabilities (Mallorcan, agentic/tool use, uncensored on Chinese politics) and license (Apache-2.0).
