Adriano
Adriano
My issue is the same as https://github.com/fastai/fastai/issues/3523 It is closed but from what I could tell the problem still persists. I am not sure if I am doing something wrong...
I get CUDA memory blowing when doing larger datasets ``` return torch.pairwise_distance(x1, x2, p, eps, keepdim) RuntimeError: CUDA out of memory. Tried to allocate 724.00 MiB (GPU 0; 11.00 GiB...
I was trying to export those models to ONNX but I am failing. I am getting the following error ``` SymbolicValueError: Unsupported: ONNX export of operator adaptive_avg_pool2d, output size that...
In this demo notebook [hete](examples/segment_anything.ipynb) the call `inputs = processor(raw_image, input_boxes=input_boxes, return_tensors="pt").to(device)` raises the error: `ValueError("Input boxes must be a list of list of list of floating points.")` `input boxes`...