def main():
try:
#capture the command line arguments from the interface script.
args = get_args()
#parse the configuration parameters for the cnn model
config = ConfigurationParameters(args)
except:
print('Missing or invalid arguments !')
exit(0)
#load the dataset from the library and print the details
dataset = LoadDataCifar10(config, dataset)
#construct , build, compile and train the cnn model
model = CNNCifar10Model(config, dataset)
#save the model to the disk
#model.save_model()
#generate graphs classfication report, confusion matrix
report = Report(config, model)
report.plot()
report.model_classification_report()
report.plot_confusion_matrix()
def main():
try:
# Capture the command line arguments from the interface script.
args = get_args()
# Parse the configuration parameters for the ConvNet Model.
config = ConfigurationParameters(args)
except:
print('Missing or invalid arguments !')
exit(0)
# Load the dataset from the library, process and print its details.
dataset = FashionMnistLoader(config)
# Test the loaded dataset by displaying the first image in both training and testing dataset.
dataset.display_data_element('train_data', 1)
dataset.display_data_element('test_data', 1)
# Construct, compile, train and evaluate the ConvNet Model.
model = FashionMnistModel(config, dataset)
# Save the ConvNet model to the disk.
# model.save_model()
# Generate graphs, classification report, confusion matrix.
report = Report(config, model)
report.plot()
report.model_classification_report()
report.plot_confusion_matrix()
def main():
try:
args = get_args()
#Parse the configuration ConfigurationParameters
config = ConfigurationParameters(args)
except:
print('Missing or invalid arguments !')
exit(0)
dataset = LoadDataCifar10(config)
g_search = GridSearchBase(config,dataset)
#create a Scikit learn wrapper
g_search.model_wrapper = KerasClassifier(build_fn = g_search.create_model, verbose=0)
#define the grid search parameters
batch_size = [1,2]
epochs = [30,50]
g_search.param_grid = dict(batch_size = batch_size,
epochs = epochs)
g_search.grid = GridSearchCV(estimator=g_search.model_wrapper,\\
param_grid = g_search.param_grid,\\
n_jobs = g_search.n_jobs)
g_search.grid_result = g_search.grid.fit(dataset.train_data,\\
dataset.train_label_one_hot)
#summarize results
print("best :%f using %s" %(g_search.grid_result.best_score_,\\
g_search.grid_result.best_params_))
means = g_search.grid_result.cv_results_['mean_test_score']
stds = g_search.grid_result.cv_results_['std_test_score']
params = g_search.grid_result.cv_results_['params']
for mean,stdev,param in zip(means, stds, params):
print("%f (%f) with: %r" %(mean, stdev, param))