Please solve some s. 1. What two pieces of information does… Please solve some s. 1. What two pieces of information does the derivative of the loss function provide?the direction and frequency with which to update the network parameters the amount and frequency with which to update the network parameters the direction and amount with which to update the network parameters 2. The validation set is used to: estimate how well a model will generalize to real-world unseen data learn the values of the network paramters that minimize the error choose between different values of hyperparamaters for a network 3. When fine-tuning a model, the learning rate should be: much higher than when transfer learning so the model can learn much quicker much lower than when transfer learning so we don’t destroy what has already been learned the same as that used when transfer learning because we are really training a new model 4. The `loss` for a neural network is: the method for calculating the error between predictions and the targets the recipe that is used to update the weights and biases is what we care about monitoring as a business owner 5. What two pieces of information does the derivative of the loss function provide? the direction and frequency with which to update the network parameters the amount and frequency with which to update the network parameters the direction and amount with which to update the network parameters 6. In order to prepare text for a deep learning model the proper sequence of steps is: index, standardize, tokenize, vectorize standardize, tokenize, index, vectorize tokenize, index, vectorize, standardize 7. If I restrict my sequences to length “max_len” then, sequences: shorter or longer than “max_len” will be ignored will be combined so that all are of length “max_len” will be truncated or padded to be of length “max_len” 8. Parameters/hyperparameters can be defined as: 9. Without limiting its size, what token is at index = 49967 in the bigram vocabulary for the text in file = Sections_1_and_2.txt? leave all leave also leave a 10. import jsonwith open(“Sections_1_and_2.txt”, “r”) as f: my_text = json.load(f) 11. Without limiting its size, what token is at index = 49967 in the bigram vocabulary for the text in file = Sections_1_and_2.txt? 12. Without limiting the size of the unigram vocabulary for the text in file = Sections_1_and_2.txt, what is the integer encoding of my_sent = ‘i am not speaking of its prudence merely its probability’? 13. What is the output of the last time step of a basic RNN, given the following information: Wxh = np.array([[-1, -3, -3, 4, -2, 4, -2], [-3, 0, 3, 0, 1, -3, 3], [-1, 1, -4, -4, 0, 3, 3], [ 2, -2, 1, 1, -1, 4, -3]]), Whh = np.array([[-3, 3, 4, 3], [ 3, -4, -3, -2], [-2, 3, -4, 3], [ 0, -2, 2, -2]]), bh = np.array([[-5], [-5], [ 3], [-5]]), Why = np.array([[ 2, 0, 1, -4], [ 3, -5, 1, -3], [-3, -3, 1, -4]]), by = np.array([[2], [1], [4]]), and input sequence = np.array([[-5, -5, -3, -4, -1, -1, -4], [-5, 0, 0, -3, 0, -4, -2], [-1, -2, -1, 1, -5, -1, -3]])? [[-0.99896612] [ 5.00094309] [-0.99896612]] [[-0.74922459] [ 3.75070732] [-0.74922459]] [[-1.49844918] [ 7.50141463] [-1.49844918]] Computer Science Engineering & Technology Artificial Intelligence Share (0)

Don't use plagiarized sources. Get Your Custom Essay on

Please solve some s. 1. What two pieces of information does…

Just from $13/Page