The following projects are based on the programming assignments of deeplearning.ai, course Sequence models. They were implemented in Keras and I reimplemented them using TensorFlow.
-
Character level text generation with LSTM: Build a character-level text generation LSTM network to generate new dinosaur names.
-
Generate music with LSTM: Build an LSTM network to generate music.
-
Translate dates using Attention mechanism: Translate dates from human readable to machine readable format using the attention mechanism.
-
Trigger word detection using GRU: Use GRUs for trigger word detection, thus a device can wake up upon hearing a certain word.
-
Word embeddings with LSTM: Implement an LSTM model which inputs a sentence in the form of word embeddings and finds the most appropriate emoji to be used with this sentence.