search

deep-floyd / IF

Other
7.64k stars 497 forks source link

Need much clearer instructions about what needs to be done for torch>=2.0.0 #57

Closed phalexo closed 1 year ago

phalexo commented 1 year ago

This is not clear. " If you are using torch>=2.0.0, make sure to delete all enable_xformers_memory_efficient_attention() functions. "

phalexo commented 1 year ago

I can see only a single call to this function. Should that call be commented out?

minskicat commented 1 year ago

There are 3, one per stage. There is an embedded comment that says to remove this line

phalexo commented 1 year ago

There are 3, one per stage. There is an embedded comment that says to remove this line

Hm. I grepped for it, saw only one. Do you know which files?

kanttouchthis commented 1 year ago

There are 3, one per stage. There is an embedded comment that says to remove this line

Hm. I grepped for it, saw only one. Do you know which files?

You dont need to change anything in the repo. You just need to change it in the code you actually use to run it, like the examples in the readme or the notebooks

SpaceCowboy850 commented 1 year ago

It would just be nice to update to pytorch 2.0 generally, as the base pytorch instructions don't even have 1.13 (I'm sure we can easily find it), and if we don't install manually and just pip install deepfloyd 1.2-rc0, we get a CPU version of torch 1.13, so then that needs to be updated to GPU, but then because it's pip, you can start accidentally screwing up requirements for torchvision and other modules.