Boreas Mistral 7B
Collection
Mistral optimized for Dutch/English: continued pretraining on bilingual data, fine-tuned for conversations.
•
4 items
•
Updated
See Boreas-7B-chat for more info. This repository contains several quantized models and their associated model cards for ollama. Click on 'files' to get a sense of the file sizes.
You can download a single GGUF file with the Huggingface Web UI widgets for quantized files. The Modelcard for ollama can be downloaded from the 'files' section.
Then:
ollama create Boreas-7B-chat-q6_k -f Modelfile_Q6_K
ollama run Boreas-7B-chat-q6_k
Enjoy!