Cuda ran out of space and used the recomanded method

[pangchunhei]

torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 118.00 MiB (GPU 0; 8.00 GiB total capacity; 7.02 GiB already allocated; 0 bytes free; 7.19 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 am using RTX3070 8GB with max_prim=1 and img_size=500, which I reduced both but am still not able to run the training, also tried to add gc.collect() and torch.cuda.empty_cache() but still not able to train. May I know if there are any solutions or will you provide the trained model?

[565353780]

Just edit the function in the Dataset class:

def filter_smallset(self):
    ...
    # "training -> train"
    if self.split == "train":
        ...
    else:
        # add self.max_prim >= len(target) for val and test data
        if self.max_prim >= len(target) >= self.filter_num:
        ...