Modelpublic

exo-jhop/ministral3-gdpr-dpo-gguf

sourceHugging Faceotherupdated 13d agoView on Hugging Face

0likes82downloads

Model Card

ministral3-gdpr-dpo-gguf

GGUF build of exo-jhop/ministral3-gdpr-dpo-fp16 for llama.cpp / llama-server deployment.

—assistant-Q5KM.gguf -- Q5KM quantized (~5.7 GiB, 5.70 BPW), recommended for serving.

bash

./llama-server --model assistant-Q5_K_M.gguf --alias eu-legal-assistant --port 8080

See the deploy runbook shipped alongside this project for full setup, health checks, and client config.

Built via converthftogguf.py (bf16) + llama-quantize (Q5K_M). Source model card: exo-jhop/ministral3-gdpr-dpo-fp16.