Automatic AIMS feedback filler this repository consist of .py file which automatically fills the feedback form (using selenium) .ipynb file is the code used to train deep learning model (made using pytorch) to read captcha .pth file has the trained weights to read captcha From analyzing the data of captcha I learned that in total there are 51 alphabets/numbers instead of 62 26 lower case alphabets (a-z), 26 uppercase alphabets (A-Z) and 10 numbers (0-9) which sums to 62 I think this might be because of similarity in the shape of some number and alphabets like 0(zero) & o dataset available on kaggle https://www.kaggle.com/datasets/utkarshdoshi/captcha-dataset