question?
Hello,
Thank you for sharing this great work! 🙌
I'm trying to load and run inference with the provided .pt checkpoint , but I noticed that the repository currently does not include the model architecture definition or a minimal script showing how to properly instantiate and load the weights.
Could you please clarify:
When will the model definition code be released?
In the meantime, could you provide the exact configuration needed to correctly load these checkpoints?
Additionally, I observed that some of the sample datasets are not sampled at 200Hz, while the inference pipeline in the documentation seems to assume a fixed 200Hz input .
Would you recommend resampling all input data to 200Hz before inference? If so, is there an official resampling script we should use?
Any guidance would be greatly appreciated!
Thanks again for your contribution to the community.