Instructions to turn off equivarinace?

[Ala1s]

could you please share how to turn off equivariance, I would like to apply this model for my use case, where I also don't need equivariance

Originally posted by @Ala1s in #25 (comment)

[NishanthARao]

Any updates on this? Were you able to figure out how to turn off the equivariance?
Thanks!

[Ala1s]

Any updates on this? Were you able to figure out how to turn off the equivariance? Thanks!

unfortunately no updates from the author, but you can use set a 'model' argument to "gnn_dynamics" instead of "egnn_dynamics", then you will be using a normal GCN as denoising networks, which works quite well in my usecase. However, I still dont completely understand the purpose of some loss terms calculated in the diffusion model so maybe something there is also ought to be removed/changed in order to improve results for non-equivariant usecase. If you figure something out regarding the loss terms, please share!