SHEN, TIANQI

The same problem. My input is a tensor whose shape is 8000x36 while my kan network is: `kan_net = KAN(width=[36, 32, 10], grid=5, k=3, seed=0, device=torch.device('cuda'))` Finally, the network will...

The same question.

The reason for this issue is that the new version ninja does not have the command: `ninja -v` Please modify the following file: xxx/miniconda3/envs/xxx/lib/python3.xxx/site-packages/torch/utils/cpp_extension.py You should ensure all `['ninja', '-v']`...

If further encountering problem like me: > running build_ext building 'topologylayer.functional.persistence' extension creating /home/stq/Research/CV/topology-gs/submodules/TopologyLayer/build/temp.linux-x86_64-cpython-310/topologylayer/functional/persistence Emitting ninja build file /home/stq/Research/CV/topology-gs/submodules/TopologyLayer/build/temp.linux-x86_64-cpython-310/build.ninja... Compiling objects... Allowing ninja to set a default number of workers......

After generating ".o" files, then you can use the new setup.py to compile the whole project using ninja: > from setuptools import setup import torch, os import sysconfig from torch.utils.cpp_extension...

Finally, if you follow my solution to pre-generate the ".o" files, please specify the Macro name defined in "topologylayer/functional/persistence/pybind.cpp" before running my script. Change: `PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)` to `PYBIND11_MODULE(persistence, m)` You...