Hello,
I'm new to the whole torchdrug pipeline thing. I was checking out the README and saw this part about "Load Trained Model Weight." After loading the weights, how do I go about turning a protein sequence into the input data, feed it into the esm_s model, and then grab the output embeddings? I reviewed the code, and there's this ESM class with a forward function that looks like forward(self, graph, input,..). Can you help me figure out how to put together the graph and input from a protein sequence? Any tips or pointers would be greatly appreciated.