Hello
If I want to use a model in a docker environment, but also want to lower the size of the image, is it possible to have a lightweight version of the transformer lib that no longer can train and so on, but only can run an already trained model? I am using a language model from Helsinki university to translate text from English to Danish
Link to the model used: https://huggingface.co/Helsinki-NLP/opus-mt-en-da
Thanks, Thomas