Philipp
Philipp
Is it possible to bump the numpy version requirement now that torch-points-kernels supports numpy 1.21? https://github.com/torch-points3d/torch-points-kernels/issues/90
Is it possible to support newer PyTorch version than 1.8 as specified in the pyproject.toml? 1.10 works fine for me now, but I have to install torch-points3d with additional parameters...
It isn't documented what kind of data objects SparseConv3D expected, so I assumed PyTorch-Geometric's `Data` objects as in the other models, however the following exception is returned when training the...
I'm trying to use the KPConv model, however I get the following exception: ``` File "[...]/venv/lib/python3.7/site-packages/torch_points3d/applications/kpconv.py", line 142, in forward data = self.down_modules[i](data) File "[...]/venv/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl return...
#### Describe the workflow you want to enable I'm using SFS within GridSearch and as far as I understand it, GridSearch splits the data into train and test, where it...
Started to resolve #107. Separated the value collection into two methods, one for named values (by `fields`) returning a dict and another one without names returning just a list. Also...
Hi, I would like to extract the parsed values together with their field names from the NMEASentence object into a general container, for example into a namedtuple or dictionary. As...
Sadly, the app is not fully usable for me because I'm using dark mode on Ubuntu 20.04 and some of the text is not readable. Here are some examples: ...
I'm using a jekyll layout where I want to include content of pages, depending on whether it exists. Apparently, it is supposed to be done like described in https://github.com/Shopify/liquid/issues/89 In...
When using the VideoReader I get different frames for the same index. For example: ``` import decord vr = decord.VideoReader("some_video.mkv") print(vr[0] == vr[0]) ``` This prints False. How is this...