Closed mzytmp closed 3 years ago
In this version we apply 4 GPU with batchsize 4 for training. Actually, benefited from the Group Normalization (GN), our network is not very sensitive to the batchsize. For any question about the setting of parameters, please refer to our released codebase.
Thanks for clarification! (that is my another acoount)
Hi, I'm asking about the implementation details. You mentioned in your paper the network for DTU is trained on 2 GPU with batchsize 6, but in your script train.sh, I don't see this. Also, is the resolution for DTU inference 480 x 640 according to eval_dtu.sh? Maybe it's not mentioned in the paper. Thank you for clarification!