Dickoabc123
Dickoabc123
Hello @zhen-yan did you manage to solve this? I too would like to know the training and validation losses for each Epoch along with the mAP for each Epoch. @AlexeyAB...
@sslupsky thank you very much for your reply. Can you please tell me how I would use the on-board PMIC or point me to a resource that explains what to...
> > @heartInsert Yes, I can't run test.py.I still don't know how to solve it. > > RuntimeError: Error(s) in loading state_dict for RetinaNet: > > Missing key(s) in state_dict:...
Thank you very much for your reply. I’ll try to do this. Plotting the validation losses against the training losses allows us to determine if the model is overfitting the...
Ah I see. Yes I know what you mean. I’m using this Faster RCNN to verify results of an old journal paper. Just not fluent enough to be able to...
Ah okay, yeah that sounds like the way to go. Not sure I’ll be able to do it but I’ll have a look. Really need the validation losses so I...
@trzy , So I tried to add to the following script to calculate the validation losses: ```# # Faster R-CNN in PyTorch and TensorFlow 2 w/ Keras # tf2/FasterRCNN/__main__.py #...
I've replaced mode.evaluate with model.test_on_batch in the validate section, lets see what happens.
No, that didn’t work either, been looking at it all day.
Of course :) I made this function to calculate the mAP: ``` def evaluate(model, eval_data = None, num_samples = None, plot = False, print_average_precisions = False): if eval_data is None:...