Yuanhong Yu

Results 11 comments of Yuanhong Yu

I remember the intrinsics of scannet have same values?

I met the same proplem and i finally solved it by setting the python version correctly which can be edit in the `CMakeLists.txt` like this 👇 ```cmake set(PYTHOH3_VERSION 3.9m) ```...

i guess @Penterakt can solve the problem by installing SD's conda env mainly because the python version of it is 3.8.5 , which is compatible for the py38 version xxx.so

I download the dataset you mentioned "199GB version" by using wget

i have some questions about in/out-painting inplementation details from the paper. you mentioned that the in/out-painting model need Cs+1 extra input channels and you write down the operation method “Concat(m,m*St,St)”...

> Could you add a test ? I added a simple test (in the latest commit) in 'test_data' which will not be passed in current version but fixed by this...

set strategy == "ddp_find_unused_parameters_true" as error log said

> Doesn't setting `strategy == "ddp_find_unused_parameters_true"` make an extra forward pass, using more computation and time? As far I understant the tutorial, there doesn't seem to be any parameters with...

As discussed further in [this](https://github.com/Lightning-AI/pytorch-lightning/issues/20685) issue, I need to correct my suggestion ⤴ that adding up the generator and discriminator losses and calling backpropagation only once is completely wrong, even...

I meet the same problem while i trying to traning my model with dynamic batch sequence length setting and i have a same strategy with @e-yi to maximize the use...