Thomas Kelly comments

Repositories
Issues
Comments

Results 5 comments of


                                            Thomas Kelly

How to visualize attention if the sizes of input and output sequence are different? ValueError.

Using generate returns the auto-regressive attentions. Try using teacher forcing (ie providing labels/decoder_ids) through the forward pass

RuntimeError: Found dtype Double but expected Float

This is a common and well-known error for the Delaney dataset.. Most of the time it is solved by ensuring at train time, by ensuring that both input and labels...

CUDA OOM with DeepSpeed ZeRO Stage 3 Offload

Seems related #17043

Segmentation fault

Similar issue [#42 ](https://github.com/gcorso/DiffDock/issues/42#issuecomment-1312430151)

ValueError: No edges and no nodes

I think the error comes from a cropping function https://github.com/gcorso/DiffDock/blob/6d065cd78c11a610a0f3eba30f836a7e3a04358f/utils/sampling.py#L108