logongas

Herramientas de usuario

Herramientas del sitio

Ocultar Menú
Traza: • Semillas
clase:iabd:pia:experimentos:semillas

¡Esta es una revisión vieja del documento!

Semillas

Comprobar el valor de los 3 generadores de números aleatorios y comprobar si varían con las versiones de python,numpy o tf

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
#!/usr/bin/env python3
import sys
import random
import numpy as np
import tensorflow as tf
 
print("Versión de Python:", sys.version)
print("Versión de NumPy:", np.__version__)
print("Versión de TensorFlow:", tf.__version__)
 
random.seed(5)
np.random.seed(5)
tf.random.set_seed(5)
  
 
print("Con random:",random.uniform(0.0, 1.0))
print("Con Numpy:",np.random.uniform())
print("Con Tensorflow:",tf.random.uniform(shape=(), minval=0, maxval=1).numpy())
 
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
  
np.random.seed(5)
tf.random.set_seed(5)
random.seed(5) 
  
model=Sequential()
model.add(Dense(5, activation='relu',input_dim=2))
model.add(Dense(5, activation='relu'))
model.compile(loss='mean_squared_error')
for layer in model.layers:
    print(layer.get_weights()[0].reshape(-1))
for layer in model.layers:
    print(layer.get_weights()[1])
 

Mi ordenador

Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.23.5
Versión de TensorFlow: 2.8.4
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.23403454  0.05525887  0.47856975  0.01571751 -0.28857118 -0.3340401
  0.35486686 -0.43231097 -0.8567835  -0.5518664 ]
[ 0.3124839  -0.70741147  0.09365994 -0.5855427  -0.34088686 -0.65839696
 -0.15128028  0.7582902   0.14519715  0.3895806  -0.689723   -0.6508303
  0.4060253   0.45104933 -0.30246195  0.40485013  0.15446562  0.06680018
  0.59774756 -0.02054155 -0.75420505  0.13622719 -0.2671212  -0.40525684
  0.6609149 ]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]


Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.23.5
Versión de TensorFlow: 2.12.0
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.44648862  0.39008212 -0.2500074   0.5530908   0.8441111   0.42121875
  0.48914194  0.1446724  -0.6092199   0.27305746]
[ 0.66853154  0.10833853  0.48234296  0.41093683  0.05519938  0.50477564
 -0.7613942   0.61522007 -0.5131155   0.34656572  0.5320357  -0.48540664
 -0.5437882   0.58495295  0.42670572  0.02728176 -0.34232992  0.5339979
 -0.26441842 -0.13371903 -0.14562988  0.7259475   0.37215436  0.6933198
 -0.45787498]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]



Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.26.1
Versión de TensorFlow: 2.14.0
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.44648862  0.39008212 -0.2500074   0.5530908   0.8441111   0.42121875
  0.48914194  0.1446724  -0.6092199   0.27305746]
[ 0.66853154  0.10833853  0.48234296  0.41093683  0.05519938  0.50477564
 -0.7613942   0.61522007 -0.5131155   0.34656572  0.5320357  -0.48540664
 -0.5437882   0.58495295  0.42670572  0.02728176 -0.34232992  0.5339979
 -0.26441842 -0.13371903 -0.14562988  0.7259475   0.37215436  0.6933198
 -0.45787498]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]


Google colab


#CPU
Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.23.5
Versión de TensorFlow: 2.14.0
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.44648862  0.39008212 -0.2500074   0.5530908   0.8441111   0.42121875
  0.48914194  0.1446724  -0.6092199   0.27305746]
[ 0.66853154  0.10833853  0.48234296  0.41093683  0.05519938  0.50477564
 -0.7613942   0.61522007 -0.5131155   0.34656572  0.5320357  -0.48540664
 -0.5437882   0.58495295  0.42670572  0.02728176 -0.34232992  0.5339979
 -0.26441842 -0.13371903 -0.14562988  0.7259475   0.37215436  0.6933198
 -0.45787498]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]

#T4 GPU
Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.23.5
Versión de TensorFlow: 2.14.0
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.44648862  0.39008212 -0.2500074   0.5530908   0.8441111   0.42121875
  0.48914194  0.1446724  -0.6092199   0.27305746]
[ 0.66853154  0.10833853  0.48234296  0.41093683  0.05519938  0.50477564
 -0.7613942   0.61522007 -0.5131155   0.34656572  0.5320357  -0.48540664
 -0.5437882   0.58495295  0.42670572  0.02728176 -0.34232992  0.5339979
 -0.26441842 -0.13371903 -0.14562988  0.7259475   0.37215436  0.6933198
 -0.45787498]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]

#TPU
Versión de Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Versión de NumPy: 1.23.5
Versión de TensorFlow: 2.12.0
Con random: 0.6229016948897019
Con Numpy: 0.22199317108973948
Con Tensorflow: 0.6263931
[ 0.44648862  0.39008212 -0.2500074   0.5530908   0.8441111   0.42121875
  0.48914194  0.1446724  -0.6092199   0.27305746]
[ 0.66853154  0.10833853  0.48234296  0.41093683  0.05519938  0.50477564
 -0.7613942   0.61522007 -0.5131155   0.34656572  0.5320357  -0.48540664
 -0.5437882   0.58495295  0.42670572  0.02728176 -0.34232992  0.5339979
 -0.26441842 -0.13371903 -0.14562988  0.7259475   0.37215436  0.6933198
 -0.45787498]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]

clase/iabd/pia/experimentos/semillas.1698480561.txt.gz · Última modificación: 2023/10/28 10:09 por admin

Herramientas de la página

Excepto donde se indique lo contrario, el contenido de este wiki esta bajo la siguiente licencia: GNU Free Documentation License 1.3
GNU Free Documentation License 1.3 Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki