RuntimeError: CUDA error: an illegal memory access was encountered，WHY？

[X1aohe7]

File "/data1/code/SemGauss-SLAM/sem_gauss.py", line 1018, in
dense_semantic_slam(experiment.config)
File "/data1/code/SemGauss-SLAM/sem_gauss.py", line 644, in dense_semantic_slam
loss.backward()
File "/data1/miniconda3/envs/sem_gauss/lib/python3.10/site-packages/torch/_tensor.py", line 492, in backward
torch.autograd.backward(
File "/data1/miniconda3/envs/sem_gauss/lib/python3.10/site-packages/torch/autograd/init.py", line 251, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: CUDA error: an illegal memory access was encountered
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

[X1aohe7]

and the datasets？ The semantic tag map was not provided.

[X1aohe7]

why just room0? room1 2 ? office 0-4?