I wanted to the add Donut LLM model which seems to be not supported at the moment

after cloning: https://huggingface.co/docs/transformers/en/model_doc/donut I have tried to run `docker run --rm -v .:/model ollama/quantize -q q8_0 /model` but it fails with: `unknown architecture VisionEncoderDecoderModel` I think one can never have enough vision models, so please add support for Donut models and its fine tunings.