def performBatchTest(weightsFilename):
tokenizer, wordsToIndex = loadWordMapping()
datasetDirectory = "aclImdb/test/"
sequenceLength = 300
vecSpaceSize = 8
print("Evaluating model performance...")
#Get the testing dataset
reviews, ratings = readDataset(datasetDirectory, sequenceLength,
vecSpaceSize)
#Load the model and its weights
model = loadModel(len(wordsToIndex) + 1, sequenceLength, vecSpaceSize)
model.load_weights(weightsFilename)
X = tokenizer.texts_to_sequences(reviews)
X = pad_sequences(X, maxlen=sequenceLength, padding='post')
loss, accuracy = model.evaluate(X, ratings)
print("\tLoss: {}".format(loss))
print("\tAccuracy: {}".format(accuracy))
#!/usr/bin/env python
import tensorflow as tf
from tensorflow.keras import datasets, layers, models
import matplotlib.pyplot as plt
import dataset as ds
import numpy as np
import sys
bin = np.arange(256)
if not ds.check_folder():
ds.makeDataset(bin, type="s")
x, y = ds.readDataset(bin, type="s")
#train_images, train_labels,test_images,test_labels =ds.readDataset(bin,type="s")
order = np.random.permutation(y.shape[0])
x = x[order, :, :, :]
y = y[order, :]
div = int(0.7 * y.shape[0])
train_images, test_images = x[0:div, :, :, :] / 1.0, x[div:, :, :, :] / 1.0
train_labels, test_labels = y[0:div, :], y[div:, :]
"""
if "-v" in sys.argv:
class_names = ['Healthy', 'Parkinson']
plt.figure(figsize=(10,10))
for i in range(26):
plt.subplot(5,6,i+1)
return paddedEncoding'''
"""def readGloveVectors():
wordToVecMap = {}
loadedVectors = np.load("embeddings.npy", mmap_mode="r")
with open("embeddings.vocab", "r", encoding="utf8") as fileRead:
for index, word in enumerate(fileRead):
wordToVecMap[word.strip()] = loadedVectors[index]
return wordToVecMap"""
datasetDirectory = "aclImdb/train/"
sequenceLength = 300
vecSpaceSize = 8
reviews, ratings = readDataset(datasetDirectory, sequenceLength, vecSpaceSize)
#Get embedded matrix representing vocabulary
wordsToIndex, tokenizer = generateWordMapping(reviews)
#Generate model and output summary
model = loadModel(len(wordsToIndex) + 1, sequenceLength, vecSpaceSize)
model.summary()
#Define weights checkpoint
filepath = "data/weights-{epoch:d}.hdf5"
checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, mode='min')
#Train the model
X = tokenizer.texts_to_sequences(reviews)
X = pad_sequences(X, maxlen=sequenceLength, padding='post')
#print(encodedReviews.shape)
model.fit(X,