Results not reproducible with Keras and TensorFlow in Python

Question

My problem is that I cannot reproduce my results with Keras and ThensorFlow.

It seems like a workaround for this problem has been posted recently on the Keras documentation site , but somehow it doesn't work for me.

What am I doing wrong?

I am using a Jupyter laptop on MBP Retina (without Nvidia GPU).

# ** Workaround from Keras Documentation **

import numpy as np
import tensorflow as tf
import random as rn

# The below is necessary in Python 3.2.3 onwards to
# have reproducible behavior for certain hash-based operations.
# See these references for further details:
# https://docs.python.org/3.4/using/cmdline.html#envvar-PYTHONHASHSEED
# https://github.com/fchollet/keras/issues/2280#issuecomment-306959926

import os
os.environ['PYTHONHASHSEED'] = '0'

# The below is necessary for starting Numpy generated random numbers
# in a well-defined initial state.

np.random.seed(42)

# The below is necessary for starting core Python generated random numbers
# in a well-defined state.

rn.seed(12345)

# Force TensorFlow to use single thread.
# Multiple threads are a potential source of
# non-reproducible results.
# For further details, see: https://stackoverflow.com/questions/42022950/which-seeds-have-to-be-set-where-to-realize-100-reproducibility-of-training-res

session_conf = tf.ConfigProto(intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)

from keras import backend as K

# The below tf.set_random_seed() will make random number generation
# in the TensorFlow backend have a well-defined initial state.
# For further details, see: https://www.tensorflow.org/api_docs/python/tf/set_random_seed

tf.set_random_seed(1234)

sess = tf.Session(graph=tf.get_default_graph(), config=session_conf)
K.set_session(sess)


# ** Workaround end **

# ** Start of my code **


# LSTM and CNN for sequence classification in the IMDB dataset
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import LSTM
from keras.layers.embeddings import Embedding
from keras.preprocessing import sequence
from sklearn import metrics
# fix random seed for reproducibility
#np.random.seed(7)

# ... importing data and so on ...

# create the model
embedding_vecor_length = 32
neurons = 91
epochs = 1
model = Sequential()
model.add(Embedding(top_words, embedding_vecor_length, input_length=max_review_length))
model.add(LSTM(neurons))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='mean_squared_logarithmic_error', optimizer='adam', metrics=['accuracy'])
print(model.summary())
model.fit(X_train, y_train, epochs=epochs, batch_size=64)
# Final evaluation of the model
scores = model.evaluate(X_test, y_test, verbose=0)
print("Accuracy: %.2f%%" % (scores[1]*100))

Used version of Python:

Python 3.6.3 |Anaconda custom (x86_64)| (default, Oct  6 2017, 12:04:38) 
[GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)]

A workaround is already included in the code (no effect).

Each time I do the training part, I get different results.

When you reboot the core of a Jupyter laptop, the first time corresponds to the first, and the second to the second.

Therefore, after a reset, I always get, for example, 0.7782on the first start, 0.7732on the second start, etc.

.

!

Answer 1

, , tensorflow , . :

#START A NEW TF SESSION
np.random.seed(0)
tf.set_random_seed(0)
sess = tf.Session(graph=tf.get_default_graph())
K.set_session(sess)

embedding_vecor_length = 32
neurons = 91
epochs = 1
model = Sequential()
model.add(Embedding(top_words, embedding_vecor_length, input_length=max_review_length))
model.add(LSTM(neurons))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='mean_squared_logarithmic_error', optimizer='adam', metrics=['accuracy'])
print(model.summary())
model.fit(X_train, y_train, epochs=epochs, batch_size=64)
# Final evaluation of the model
scores = model.evaluate(X_test, y_test, verbose=0)
print("Accuracy: %.2f%%" % (scores[1]*100))

#CLOSE TF SESSION
K.clear_session()

GPU endorflow:

print datetime.now()
for i in range(10):
    np.random.seed(0)
    tf.set_random_seed(0)
    sess = tf.Session(graph=tf.get_default_graph())
    K.set_session(sess)

    n_classes = 3
    n_epochs = 20
    batch_size = 128

    task = Input(shape = x.shape[1:])
    h = Dense(100, activation='relu', name='shared')(task)
    h1= Dense(100, activation='relu', name='single1')(h)
    output1 = Dense(n_classes, activation='softmax')(h1)

    model = Model(task, output1)
    model.compile(loss='categorical_crossentropy', optimizer='Adam')
    model.fit(x_train, y_train_onehot, batch_size = batch_size, epochs=n_epochs, verbose=0)
print(model.evaluate(x=x_test, y=y_test_onehot, batch_size=batch_size, verbose=0))
K.clear_session()

:

2017-10-23 11:27:14.494482
0.489712882132
0.489712893813
0.489712892765
0.489712854426
0.489712882132
0.489712864011
0.486303713004
0.489712903398
0.489712892765
0.489712903398

, tf- ( , ), "" .

Answer 2

, TensorFlow/Keras . Keras CNTK .

TensorFlow 1.2.1 1.13.1. TensorFlow , .

Answer 3

: Keras Tensorflow. for, , , for loop.

for...
   for...
      reset_keras()
      .
      .
      .

def reset_keras():
    sess = tf.keras.backend.get_session()
    tf.keras.backend.clear_session()
    sess.close()
    sess = tf.keras.backend.get_session()
    np.random.seed(1)
    tf.set_random_seed(2)

PS: nvidia ( ), ... .

Answer 4

+ .

1, .

import os
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = ""

2, , , "tenorflow, numpy, random".

import tensorflow as tf
import numpy as np
import random as rn

sd = 1 # Here sd means seed.
np.random.seed(sd)
rn.seed(sd)
os.environ['PYTHONHASHSEED']=str(sd)

from keras import backend as K
config = tf.ConfigProto(intra_op_parallelism_threads=1,inter_op_parallelism_threads=1)
tf.set_random_seed(sd)
sess = tf.Session(graph=tf.get_default_graph(), config=config)
K.set_session(sess)

, , .

Answer 5

I posted an answer to a similar question (link fooobar.com/questions/18230627 / ... ). And that was accepted. The key to reproducing the same result is turning off the GPU. Hope this can solve your problem.

Answer 6

The best solution that currently works with the GPU is to set the determinism flow tensor with the following:

pip install tensorflow-determinism

Then add the following code to your code

import tensorflow as tf
import os
os.environ['TF_DETERMINISTIC_OPS'] = '1'

source: https://github.com/NVIDIA/tensorflow-determinism