tiny_fold parameters.py holds parameter values that are used across multiple .py files: free energy parameters, priors used in cost function, learning parameters for Adam method, training/testing data for testing code partition.py holds function definitions for calculating the partition function and gradient (and individual derivatives) for both linear and circular RNA sequences recursion modified from NUPACK n^4 algorithm in Dirks and Pierce 2003 log_partition.py holds function definitions for calculating log(partition) and its derivative for both linear and circular RNA sequences random_sequence_generator.py generates training and testing set of random RNA sequences and calculates their partition functions sequences_train.txt & sequences_test.txt store the RNA sequences (and partition function) that were used in that run to train/test the model stochastic_gradient_descent.py implementation of stochastic gradient descent using Adam method train_toy.py holds the cost function that toy.py minimizes toy.py trains model using methods in scipy.optimize convergence.py graphs the value of the cost function after each iteration of SGD to see how it converges