search

andvg3 / LSDM

Dataset and Code for NeurIPS 2023 paper "Language-driven Scene Synthesis using Multi-conditional Diffusion Model."
https://lang-scene-synth.github.io/
MIT License
47 stars 1 forks source link

Segfault during test time #6

Open LukasH77 opened 1 week ago

LukasH77 commented 1 week ago

python -m run.test_sdm data/protext/proxd_test/ \ --load_model training/sdm/model_ckpt/best_model_cfd.pt \ --model_name sdm \ --fix_ori \ --test_on_valid_set \ --output_dir training/sdm/output

running this as advised in the readme gives me a segfault error this is the output I get

PolyMeshT::add_face: patch re-linking failed PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge ~/miniconda3/envs/external_lsdm/lib/python3.11/site-packages/scipy/sparse/_index.py:108: SparseEfficiencyWarning: Changing the sparsity structure of a csr_matrix is expensive. lil and dok are more efficient. self._set_intXint(row, col, x.flat[0]) PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge PolyMeshT::add_face: complex edge 0%| | 0/20 [00:00<?, ?it/s] Segmentation fault (core dumped)

Originally there also were some deprecation warnings that I fixed on the off chance of it helping. Any experiences with this?

andvg3 commented 1 week ago

Hi, thanks for reporting this bug. I highly suspect that the problem was because of the system. Can you provide me more details about your installed packages version and systems?