if __name__ == '__main__':
if tf.gfile.Exists(train_dir):
tf.gfile.DeleteRecursively(train_dir)
tf.gfile.MakeDirs(train_dir)
log = open(train_dir + ".txt", "w", 1)
is_training = tf.get_variable('is_training',
shape=(),
dtype=tf.bool,
initializer=tf.constant_initializer(
True, dtype=tf.bool),
trainable=False)
global_step = tf.Variable(0, trainable=False)
# trainData, testData, numTrainExamples, numTestExamples, testIterator = preprocessing.inputFlows(batch_size)
trainData, testData, numTrainExamples, numTestExamples, testIterator = preprocessing.inputFlows(
batch_size, numTrainExamples=numTrainExamples)
perGPUTrainData = [list([]) for i in range(numGpus)]
for tD in trainData:
split = tf.split(tD, numGpus, axis=0)
for gpu in range(numGpus):
perGPUTrainData[gpu].append(split[gpu])
perGPUTestData = [list([]) for i in range(numGpus)]
for tD in testData[:-1]:
split = tf.split(tD, numGpus, axis=0)
for gpu in range(numGpus):
perGPUTestData[gpu].append(split[gpu])
for gpu in range(numGpus):
with tf.name_scope('tower_%d' % (gpu)) as scope:
count += 1
except tf.errors.OutOfRangeError:
break
return numpy.sqrt(mean/count), std/count
if __name__ == '__main__':
if tf.gfile.Exists(train_dir):
tf.gfile.DeleteRecursively(train_dir)
tf.gfile.MakeDirs(train_dir)
log = open(train_dir+".txt", "w", 1)
is_training = tf.get_variable('is_training', shape=(), dtype=tf.bool, initializer=tf.constant_initializer(True, dtype=tf.bool), trainable=False)
global_step = tf.Variable(0, trainable=False)
trainDataSmall, testDataSmall, numTrainExamplesSmall, numTestExamplesSmall, testIteratorSmall = preprocessing.inputFlowsForFlowPrediction(batch_size)
trainDataBig, testDataBig, numTrainExamplesBig, numTestExamplesBig, testIteratorBig = preprocessing.inputFlows(batch_size)
# trainData, testData, numTrainExamples, numTestExamples, testIterator = preprocessing.inputFlows(batch_size, numTrainExamples=8000)
# numTrainExamples = numTrainExamplesSmall
numTrainExamples = numTrainExamplesBig
perGPUTrainDataBig = [list([]) for i in range(numGpus)]
for tD in trainDataBig:
split = tf.split(tD, numGpus, axis=0)
for gpu in range(numGpus):
perGPUTrainDataBig[gpu].append(split[gpu])
perGPUTestDataBig = [list([]) for i in range(numGpus)]
for tD in testDataBig[:-1]:
split = tf.split(tD, numGpus, axis=0)
for gpu in range(numGpus):
perGPUTestDataBig[gpu].append(split[gpu])