In this mini project, i use LSTM and encoder-decoder architecture for image captioning on Flickr8k dataset. I used Keras as the main framework for designing the architecture and PyTorch for using the "bert-base-uncased" model for finding the similarity between true and predicted labels
