Amazing work! Would love to learn from the training implementation details #28
Description
I've been using/studying your code and paper, and I’m particularly inspired by the way you train the model.
The inference/evaluation code provided is already very clear and easy to run. However, as a learner and researcher, I’m also very curious about how the training was conducted — details like data preprocessing, training strategy would be incredibly helpful for understanding and reproducing your results.
If possible, would you consider releasing the training code as well? It would greatly benefit the community and help others learn from your excellent work! Of course, I fully understand if there are constraints on time, resources, or other considerations, so no pressure at all. Regardless, thank you again for your valuable contribution.
Once again, huge respect for what you’ve built — I’ve starred the repo and will continue to follow its progress. Looking forward to learning more from your work!
Best regards,