I am working on a speech-to-text task with an architecture composed of LSTMs and a CTC on keras. I already understood the theory behind it and now i\'m working on the implem
