peizhou26 / few-shot-3d-point-cloud-classification Goto Github PK

Hi, I always get this error for DGCNN and PointNet2

CUDA out of memory. Tried to allocate 1.56 GiB (GPU 0; 14.75 GiB total capacity; 14.43 GiB already allocated; 154.81 MiB free; 14.46 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF)

I have tried multiple GPUs 24GB for about 08 machines etc, but the same issue.
Also reduced batch size, workers, pin to memory, etc. Can you recommend

I need your further guidance. Thank you

Hi,
Thank you for sharing your work. I’m a bit confused with the novel.json and base.json files.
For both files, the labels are not aligning. If it's possible, can you please help me? Do the labels for the same class in novel.json and base.json need to be the same? (most probably I'm getting it wrong.)

For example:
label for airplane in base.json is 0
label for bathtub in novel.json is 0

much appreciated. Thank you