AlexandreLamb
Results
2
comments of
AlexandreLamb
trafficstars
`Line 410: out = self._pnet.predict(img_y, verbose=0)` is on `Line 342`
Delete cuda() `img_crop = torch.tensor(img_crop, dtype=torch.float32).unsqueeze(0)` You also have to change the device arguments to "CPU" when you call `get_model_by_name`