Will Vithayapalert issues

Repositories
Issues
Comments

Results 2 issues of


                                            Will Vithayapalert

Make chain and structure optional

Why do we assume batch size of 1 for structure predictions during training?

During training, why do we always set example = batch[0] for 1 instance of structure prediction? https://github.com/RosettaCommons/foundry/blob/676b4a95f3dda7f599ce2c0618033251cc4a170c/models/rfd3/src/rfd3/trainer/rfd3.py#L157

question

RFdiffusion3