deep-text-recognition-benchmark
deep-text-recognition-benchmark copied to clipboard
FineTuning Layer Freezing
How do you exactly freeze layers during FineTuning? I want to freeze feature extraction layers because i my finetuned model stops recognizing the old data the pretrained model recognizes, but i want my fine tuned model to keep recognizing it and also recognize the data it was fine tuned on. do i just set the freeze_FeatureExtraction: False to True?