lizhicheng
lizhicheng
> Yes, this is a bug. I'll fix this later today. hi，May I ask when this bug is expected to be fixed？
> hi, I change batch_size to 2 in config.py, but it will occur to " File "/data1/zjc/openfold-install/openfold/model/model.py", line 189, in embed_templates t = t * (torch.sum(batch["template_mask"], dim=-1) > 0) RuntimeError:...
> Is training on bigger batches supported in OpenFold-1.0? experimentally resolved loss and violation loss complain with a bigger batch size that the dimensionality is wrong. > > I'm also...
Is it true that during training, no matter whether the input size changes or not, it always goes through the same calculation path？thanks in advance
So can I assume that during training, all training data have the same computational logic?
AMGGPU MI50 with 4 SE, each SE 15 CUS