Open manojramanathan opened 3 years ago
The problem with loading the trained models into github is that their weight surpasses the maximum allowed. However, I will try to find a solution to this restriction and update the repository as soon as possible.
Nonetheless, if you happen to have a custom dataset you could use the 'train.py' to reproduce the results by using quantization aware training and regular baseline training. Beware that you might have to modify the 'FasterRCNN/staircase_dataset.py' to load the annotations and images from your dataset.
Currently, both trained_model and dataset are available only to people with access to bcv002. Is there a way to get these as I don't have access to this server. Could you please how to get the required dataset and trained_model? As it is not possible to run the code without them.