Memory

This project uses recurrent neural networks and CNNs to test the task of memory retention.

Instructions to run - first run dataset_creation.py to create the dataset adn then run memory_encoder_mlp.py to evaluate memory retention.

dataset_creation.py - script to create synthetic dataset and save it on disk
memory_encoder_mlp.py - LSTM, RNN, and CNN models to test memory retention task.

Name		Name	Last commit message	Last commit date
__pycache__		__pycache__
dataset		dataset
s2s		s2s
.gitignore		.gitignore
README.md		README.md
RNN.py		RNN.py
balanced_acc_seq_len_CNN.pkl		balanced_acc_seq_len_CNN.pkl
balanced_acc_seq_len_RNN.pkl		balanced_acc_seq_len_RNN.pkl
balanced_acc_seq_len_dist_CNN.pkl		balanced_acc_seq_len_dist_CNN.pkl
balanced_acc_seq_len_dist_RNN.pkl		balanced_acc_seq_len_dist_RNN.pkl
balanced_acc_seq_len_dist_lstm.pkl		balanced_acc_seq_len_dist_lstm.pkl
balanced_acc_seq_len_dist_transformer_no_orthonormal.pkl		balanced_acc_seq_len_dist_transformer_no_orthonormal.pkl
balanced_acc_seq_len_distlstm.pkl		balanced_acc_seq_len_distlstm.pkl
balanced_acc_seq_len_lstm.pkl		balanced_acc_seq_len_lstm.pkl
dataset_creation.py		dataset_creation.py
fra.txt		fra.txt
hyperparam_tuning.py		hyperparam_tuning.py
lstm_seq2seq.py		lstm_seq2seq.py
lstm_seq2seq_synthetic.py		lstm_seq2seq_synthetic.py
lstm_seq2seq_synthetic_encoder_mlp.py		lstm_seq2seq_synthetic_encoder_mlp.py
memory_encoder_mlp.py		memory_encoder_mlp.py
memory_encoder_mlp_optuna.py		memory_encoder_mlp_optuna.py
plots.py		plots.py
seq2seq_synthetic_encoder_mlp.py		seq2seq_synthetic_encoder_mlp.py
test.py		test.py
test_file.py		test_file.py
train.py		train.py
train_evaluate.py		train_evaluate.py
visuals.py		visuals.py

Navigation Menu