-
Notifications
You must be signed in to change notification settings - Fork 0
/
KNIME_Replication_3D_Midpoint_DM.py
186 lines (150 loc) · 7.32 KB
/
KNIME_Replication_3D_Midpoint_DM.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
import tensorflow as tf
import numpy as np
import ml_functions as mli
import datetime as dt
import os
from sklearn.datasets import fetch_openml
from sklearn.model_selection import train_test_split
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import matplotlib as mplt
import sys
# Check TF uses GPU
if tf.test.gpu_device_name():
print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))
else:
print("Please install GPU version of TF")
#Management of parameters in execution
n = len(sys.argv)
print(f"\033[1;32m\nExecuting: {sys.argv[0]}")
print("Total arguments passed:", n-1)
assert n <= 2
if n == 1:
l_space = int(3)
if n == 2:
l_space = int(sys.argv[1])
assert l_space <= 3, "Dimension should be 2 or 3"
print(f"Latent space dimension: {l_space}\033[0;0m")
# Dataset download
STORE_PATH = f"/home/isra/PycharmProjects/CodingTensorFlowV2/StoredResults/KNIME/3D_Midpoint_DM/{dt.datetime.now().strftime('%d%m%Y%H%M')}"
mnist = fetch_openml('mnist_784', version=1, as_frame=True)
df_data = mnist.data.iloc[:]
df_data = df_data.assign(digit_class=mnist.target.iloc[:].astype(int))
np_data = df_data.to_numpy()
np_data_images = np_data[:, 0:784]
np_data_images = np.reshape(np_data_images, (70000, 28, 28))
np_data_labels = np_data[:, 784]
#Partitioning, Test, Train, Validation
x_train,x_validation, y_train, y_validation = train_test_split(np_data_images, np_data_labels, test_size=0.25, random_state=42)
x_train,x_test, y_train, y_test = train_test_split(x_train, y_train, test_size=0.20, random_state=42)
# Datasets capturados
# Función de entrenamiento
def run_training_enc(model_inst: mli.ModelGr, sub_folder: str, iterations: int = 2500,
batch_size: int = 32, log_freq: int = 200, lim_accuracy: float = 0.99,
lim_loss: float = 0.05, graph_name: str = None):
# Directorio para almacenar datos
train_writer = tf.summary.create_file_writer(STORE_PATH + "/" + sub_folder)
# Guardamos el grafo
model_inst.plot_computational_graph(train_writer, x_train[:batch_size, :, :], graph_name)
# Selección del optimizador
optimizer = tf.keras.optimizers.Adam(learning_rate=0.001, beta_1=0.9, beta_2=0.999, epsilon=1e-08, amsgrad=False) # Puede tener un argumento de learning rate
# Vamos a la iteración de entrenamiento
acc = 0
for j in range(iterations):
# Tomamos un batch
image_batch, label_batch = mli.get_batch(x_train, y_train, batch_size)
# image_batch = tf.Variable(image_batch) (Está en el libro pero no funciona)
label_batch = tf.cast(tf.Variable(label_batch), tf.int32)
# Calculamos los logits y la perdida
with tf.GradientTape() as tape:
logits = model_inst.forward(image_batch)
loss = model_inst.loss(logits, label_batch)
gradients = tape.gradient(loss, model_inst.nn_model.trainable_variables)
optimizer.apply_gradients(zip(gradients, model_inst.nn_model.trainable_variables))
# Zona de Logs
if j % log_freq == 0:
max_idxs = tf.argmax(logits, axis=1)
# El resultado de lo anterior e un tensor... Por eso necesitamos hacer un numpy()
acc = np.sum(max_idxs.numpy() == label_batch.numpy()) / len(label_batch.numpy())
print(f"Iter: {j}, loss={loss:.3f}, accuracy={acc * 100:.3f}% (Scenario {graph_name})")
with train_writer.as_default():
tf.summary.scalar('loss', loss, step=j)
tf.summary.scalar('accuracy', acc, step=j)
# log the gradients
model_inst.log_gradients(gradients, train_writer, j)
if acc >= lim_accuracy and loss <= lim_loss:
print(f'\033[1;31m \nEnd for limit in Accuracy={acc*100:.3f}% and Loss={loss:.3f} \033[0;0m')
break
#Accuracy and loss in tests
if __name__ == "__main__":
#Variables de definición de la red
#act_funcs = [tf.sigmoid, tf.nn.relu, tf.nn.leaky_relu, tf.nn.tanh, tf.nn.softplus, tf.nn.softsign]
num_layers = 6
act_functions = [tf.nn.tanh,
tf.nn.tanh,
tf.nn.tanh,
tf.nn.tanh,
tf.nn.tanh]
sizes = [300, 100, 10, l_space, 10, 10]
subfolder_name = "Dense_Model"
# Construcción de las subfolders
out_file = f"/TensorFlow_Visualization_{dt.datetime.now().strftime('%d%m%Y%H%M')}"
# Gardado de datos de entrenamiento
print(f"\033[1;32m Running training: \033[0;0m")
model = mli.ModelGr(activations=act_functions, sizes=sizes,
num_layers=num_layers, name=subfolder_name)
run_training_enc(model, sub_folder=subfolder_name, iterations=5500,
lim_accuracy=0.99, lim_loss=0.05,
graph_name=subfolder_name, batch_size= 300)
#Construcción de la media red
image_batch, label_batch = mli.get_batch(x_validation, y_validation, 17500)
label_batch = tf.cast(tf.Variable(label_batch), tf.int32)
Encoder = mli.Encoder(model)
#Forwarde de la media red
logits = Encoder.forward(image_batch)
logits_array = logits.numpy()
#Accuracy and loss for classifier
val_logits = model.forward(image_batch)
max_idxs = tf.argmax(val_logits, axis=1)
val_loss = model.loss(val_logits, label_batch)
val_acc = np.sum(max_idxs.numpy() == label_batch.numpy()) / len(label_batch.numpy())
print(f"\033[1;31m \n End Iter: loss={val_loss:.3f}, accuracy={val_acc * 100:.3f}%. Scenario: {subfolder_name}")
#Color Codes
colors = ['#33a02c', '#e31a1c', '#b15928', '#6a3d9a', '#1f78b4',
'#ff7f00', '#b2df8a', '#fdbf6f', '#fb9a99', '#cab2d6']
if min(sizes) == 2:
labeled_images = pd.DataFrame(logits_array, columns=["X", "Y"])
labeled_images = labeled_images.assign(label=label_batch)
#Ploteamos el diabolo 2D
plt.figure(figsize=(10, 6))
sns.scatterplot(x="X", y="Y",
hue="label",
data=labeled_images,
legend="full",
palette=colors);
plt.legend(bbox_to_anchor=(1.01, 1),borderaxespad=0)
plt.title(f'Scatter Latent Space 2D. Python Dense Midpoint\n784 - {sizes}')
plt.tight_layout()
plt.show()
elif min(sizes) == 3:
labeled_images = pd.DataFrame(logits_array, columns=["X", "Y", "Z"])
labeled_images = labeled_images.assign(label=label_batch)
fig = plt.figure(figsize=(12, 12))
ax = fig.add_subplot(111, projection='3d')
sc = ax.scatter(labeled_images["X"], labeled_images["Y"], labeled_images["Z"],
c=labeled_images["label"],
cmap=mplt.colors.ListedColormap(colors),
marker='o',
s=5)
plt.legend(*sc.legend_elements(), bbox_to_anchor=(1.05, 1), loc=2)
plt.title(f'Scatter Latent Space 3D. Python Dense Midpoint\n784 - {sizes}')
ax.set_xlabel(labeled_images["X"].name)
ax.set_ylabel(labeled_images["Y"].name)
ax.set_zlabel(labeled_images["Z"].name)
plt.show()
print("\033[1;32m \nWill execute TensorBoard for the last training...")
print("\nEnd of execution \033[0;0m")
outfile = STORE_PATH
print("\ntensorboard --logdir={}/{}\n".format(STORE_PATH, subfolder_name))
os.system("tensorboard --logdir={}/{}".format(STORE_PATH, subfolder_name))