RaulPPelaez
commented
3 months ago Hi, that article came out before torchmd-net had a conda package. It requires a version of this project from around 3-4 years ago. Back then there was not even a versioning system in place. I was not involved in torchmd-cg, so I cannot say exactly which commit/tag that project is expecting, but I would start by trying out https://github.com/torchmd/torchmd-net/releases/tag/0.0.1
We are working on Chignolin Coarse Grain Tutorial from this repository: https://github.com/torchmd/torchmd-cg.git
We are working on a Machine Learning project pertaining to Coarse-graining of protein molecules. As a part of this project, we are currently trying to run the torchmd-cg tutorial, which involves training and simulation of Chignolin data (as an example). We ran the following command to train the Chignolin data: torchmd-train -c train.yaml The train.yaml file contains certain parameters like test_ratio, val_ratio, inference_batchsize, distribured_backend among others which resulted in errors upon running the code. I have attached the screenshot of the error message displayed. In the train.yaml file, we commented those parameters out and the code was executed without errors. The train.yaml file with commented out parameters has been attached. It will be really helpful if you inform us about the significance of these parameters and ways to include them without having to face errors.