To make the model easier to try out, would it be possible to add: - [ ] License for both the code base and the weights - [ ] Link to weights on Hugging Face (or wherever they are stored) - [ ] Example of how to perform inference with the model from within a Python script (to avoid running the whole evaluation/training pipeline)