Fix ATL_Python. Now working great.

This commit is contained in:
Marcus Vinicius de Carvalho 2020-02-19 14:14:31 +08:00
parent 8f306f8888
commit db3494855e
2 changed files with 21 additions and 20 deletions

25
ATL.py
View File

@ -99,7 +99,6 @@ def width_evolution(network: NeuralNetwork, x: torch.tensor, y: torch.tensor = N
network.forward_pass(x) network.forward_pass(x)
network.run_agmm(x, y) network.run_agmm(x, y)
network.feedforward(x, y) network.feedforward(x, y)
network.width_adaptation_stepwise(y) network.width_adaptation_stepwise(y)
@ -143,7 +142,7 @@ def test(network: NeuralNetwork, x: torch.tensor, y: torch.tensor = None, is_sou
metrics['reconstruction_target_loss'].append(float(network.loss_value)) metrics['reconstruction_target_loss'].append(float(network.loss_value))
def force_same_size(a_tensor, b_tensor, shuffle=True, strategy='max'): def force_same_size(a_tensor, b_tensor, shuffle=True, strategy='min'):
common = np.min([a_tensor.shape[0], b_tensor.shape[0]]) common = np.min([a_tensor.shape[0], b_tensor.shape[0]])
if shuffle: if shuffle:
@ -174,7 +173,7 @@ def kl(ae: NeuralNetwork, x_source: torch.tensor, x_target: torch.tensor):
ae.reset_grad() ae.reset_grad()
kl_loss = torch.nn.functional.kl_div(ae.forward_pass(x_target).layer_value[1], kl_loss = torch.nn.functional.kl_div(ae.forward_pass(x_target).layer_value[1],
ae.forward_pass(x_source).layer_value[1], reduction='batchmean') ae.forward_pass(x_source).layer_value[1])
kl_loss.backward() kl_loss.backward()
ae.weight[0] = ae.weight[0] - ae.learning_rate * ae.weight[0].grad ae.weight[0] = ae.weight[0] - ae.learning_rate * ae.weight[0].grad
@ -359,6 +358,18 @@ def ATL(epochs: int = 1, n_batch: int = 1000, device='cpu'):
np.mean(metrics['classification_rate_target']) * 100, np.mean(metrics['classification_rate_target']) * 100,
np.min(metrics['classification_rate_target']) * 100, np.min(metrics['classification_rate_target']) * 100,
metrics['classification_rate_target'][-1] * 100)) metrics['classification_rate_target'][-1] * 100))
print(('%s %s %s %s AGMM Source:' + Fore.GREEN + ' %d ' + Fore.YELLOW + '%f' + Style.RESET_ALL + Fore.RED + ' %d' + Fore.BLUE + ' %d' + Style.RESET_ALL) % (
string_max, string_mean, string_min, string_now,
np.max(metrics['agmm_source_size_by_batch']),
np.mean(metrics['agmm_source_size_by_batch']),
np.min(metrics['agmm_source_size_by_batch']),
metrics['agmm_source_size_by_batch'][-1]))
print(('%s %s %s %s AGMM Target:' + Fore.GREEN + ' %d ' + Fore.YELLOW + '%f' + Style.RESET_ALL + Fore.RED + ' %d' + Fore.BLUE + ' %d' + Style.RESET_ALL) % (
string_max, string_mean, string_min, string_now,
np.max(metrics['agmm_target_size_by_batch']),
np.mean(metrics['agmm_target_size_by_batch']),
np.min(metrics['agmm_target_size_by_batch']),
metrics['agmm_target_size_by_batch'][-1]))
print(('%s %s %s %s Classification Source Loss:' + Fore.GREEN + ' %f' + Fore.YELLOW + ' %f' + Fore.RED + ' %f' + Fore.BLUE + ' %f' + Style.RESET_ALL) % ( print(('%s %s %s %s Classification Source Loss:' + Fore.GREEN + ' %f' + Fore.YELLOW + ' %f' + Fore.RED + ' %f' + Fore.BLUE + ' %f' + Style.RESET_ALL) % (
string_max, string_mean, string_min, string_now, string_max, string_mean, string_min, string_now,
np.max(metrics['classification_source_loss']), np.max(metrics['classification_source_loss']),
@ -377,7 +388,7 @@ def ATL(epochs: int = 1, n_batch: int = 1000, device='cpu'):
np.mean(metrics['reconstruction_target_loss']), np.mean(metrics['reconstruction_target_loss']),
np.min(metrics['reconstruction_target_loss']), np.min(metrics['reconstruction_target_loss']),
metrics['reconstruction_target_loss'][-1])) metrics['reconstruction_target_loss'][-1]))
print(('%s %s %s %s Kullback-Leibler loss 1:' + Fore.GREEN + ' %f' + Fore.YELLOW + ' %f' + Fore.RED + ' %f' + Fore.BLUE + ' %f' + Style.RESET_ALL) % ( print(('%s %s %s %s Kullback-Leibler Loss:' + Fore.GREEN + ' %f' + Fore.YELLOW + ' %f' + Fore.RED + ' %f' + Fore.BLUE + ' %f' + Style.RESET_ALL) % (
string_max, string_mean, string_min, string_now, string_max, string_mean, string_min, string_now,
np.max(metrics['kl_loss']), np.max(metrics['kl_loss']),
np.mean(metrics['kl_loss']), np.mean(metrics['kl_loss']),
@ -414,7 +425,7 @@ def ATL(epochs: int = 1, n_batch: int = 1000, device='cpu'):
dm.load_custom_csv() dm.load_custom_csv()
dm.normalize() dm.normalize()
dm.split_as_source_target_streams(n_batch, 'dallas_2', 0.5) dm.split_as_source_target_streams(n_batch, 0.5)
nn = NeuralNetwork([dm.number_features, 1, dm.number_classes]) nn = NeuralNetwork([dm.number_features, 1, dm.number_classes])
ae = DenoisingAutoEncoder([nn.layers[0], nn.layers[1], nn.layers[0]]) ae = DenoisingAutoEncoder([nn.layers[0], nn.layers[1], nn.layers[0]])
@ -445,14 +456,14 @@ def ATL(epochs: int = 1, n_batch: int = 1000, device='cpu'):
metrics['train_time'].append(time.time()) metrics['train_time'].append(time.time())
for epoch in range(epochs): for epoch in range(epochs):
for x, y in [(x.view(1, x.shape[0]), y.view(1, y.shape[0])) for x, y in zip(Xs, ys)]: for x, y in [(x.view(1, x.shape[0]), y.view(1, y.shape[0])) for x, y in zip(Xs, ys)]:
width_evolution(network=nn, x=x, y=y, agmm=agmm_source_discriminative, train_agmm=True if epoch == 1 else False) width_evolution(network=nn, x=x, y=y, agmm=agmm_source_discriminative, train_agmm=True if epoch == 0 else False)
if not grow_nodes(nn, ae): prune_nodes(nn, ae) if not grow_nodes(nn, ae): prune_nodes(nn, ae)
discriminative(network=nn, x=x, y=y, agmm=agmm_source_discriminative) discriminative(network=nn, x=x, y=y, agmm=agmm_source_discriminative)
copy_weights(source=nn, target=ae, layer_numbers=[1]) copy_weights(source=nn, target=ae, layer_numbers=[1])
for x in [x.view(1, x.shape[0]) for x in Xt]: for x in [x.view(1, x.shape[0]) for x in Xt]:
width_evolution(network=ae, x=x, agmm=agmm_target_generative, train_agmm=True if epoch == 1 else False) width_evolution(network=ae, x=x, agmm=agmm_target_generative, train_agmm=True if epoch == 0 else False)
if not grow_nodes(ae, nn): prune_nodes(ae, nn) if not grow_nodes(ae, nn): prune_nodes(ae, nn)
generative(network=ae, x=x, agmm=agmm_target_generative) generative(network=ae, x=x, agmm=agmm_target_generative)

View File

@ -84,19 +84,9 @@ class DataManipulator:
def normalize_image(self): def normalize_image(self):
raise TypeError('Not implemented') raise TypeError('Not implemented')
def split_as_source_target_streams(self, number_fold_elements=0, method=None, sampling_ratio=0.5): def split_as_source_target_streams(self, number_fold_elements=0, sampling_ratio=0.5):
if number_fold_elements == 0: self.number_fold_elements = number_fold_elements if number_fold_elements is not 0 else self.data.shape[0]
self.number_fold_elements == self.data.shape[0] self.__split_as_source_target_streams_dallas_2(self.number_fold_elements, sampling_ratio)
else:
self.number_fold_elements = number_fold_elements
if method == None or method == 'none' or method == 'None':
self.__split_as_source_target_streams_none(self.number_fold_elements, sampling_ratio)
elif method == 'dallas_1' or method == 'dallas1':
self.__split_as_source_target_streams_dallas_1(self.number_fold_elements, sampling_ratio)
elif method == 'dallas_2' or method == 'dallas2':
self.__split_as_source_target_streams_dallas_2(self.number_fold_elements, sampling_ratio)
self.__create_Xs_ys_Xt_yt() self.__create_Xs_ys_Xt_yt()
def get_Xs(self, number_minibatch): def get_Xs(self, number_minibatch):