Pytorch Point Transformer

Results

You can download a trained model from the model registry.

pip install -r ./requirements.torch.gpu.txt
pip install -r ./requirements.cpu.txt -r ./requirements.base.txt

pip install -r ./requirements.torch.gpu.txt
pip install -r ./requirements.gpu.txt -r ./requirements.base.txt

Only for presentation.ipynb

pip install -r requirements.add.txt

dvc repro pipelines/simple_shapes/dvc.yaml:train

dvc repro pipelines/modelnet10/dvc.yaml:train

or

dvc repro pipelines/modelnet40/dvc.yaml:train

Download dataset here
Run: python ./preprocess_partnet.py --data_dir <data_dir> --out_dir <where_to_store_processed>
Edit: partnet.yaml. Set dataset_dir to the <where_to_store_processed>
Run: dvc repro pipelines/partnet/dvc.yaml:train

Name		Name	Last commit message	Last commit date
Latest commit History 141 Commits
.dvc		.dvc
.github/workflows		.github/workflows
configs		configs
log_set		log_set
media		media
pipelines		pipelines
point_transformer		point_transformer
tests		tests
training		training
transforms		transforms
.dvcignore		.dvcignore
.editorconfig		.editorconfig
.gitignore		.gitignore
.pylintrc		.pylintrc
README.md		README.md
colab_training.ipynb		colab_training.ipynb
debug_model.py		debug_model.py
downloader.py		downloader.py
log_settings.yaml		log_settings.yaml
notebook_utils.py		notebook_utils.py
prepare_data.py		prepare_data.py
preprocess_partnet.py		preprocess_partnet.py
presentation.ipynb		presentation.ipynb
requirements.add.txt		requirements.add.txt
requirements.base.txt		requirements.base.txt
requirements.cpu.txt		requirements.cpu.txt
requirements.dev.txt		requirements.dev.txt
requirements.gpu.txt		requirements.gpu.txt
requirements.torch.cpu.txt		requirements.torch.cpu.txt
requirements.torch.gpu.txt		requirements.torch.gpu.txt
resolve_config.py		resolve_config.py
show_partnet.ipynb		show_partnet.ipynb
tox.ini		tox.ini
train.py		train.py