• 大小: 12KB
    文件类型: .zip
    金币: 1
    下载: 0 次
    发布日期: 2021-06-17
  • 语言: 其他
  • 标签: TensorFlow  1.2  CIFAR10  

资源简介

TensorFlow1.2版本CIFAR10代码,之前网上的版本基本上都不能用了 MacOS升级TensorFlow1.2: pip install --upgrade tensorflow

资源截图

代码片段和文件信息

# Copyright 2015 Google Inc. All Rights Reserved.
#
# Licensed under the Apache License Version 2.0 (the “License“);
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing software
# distributed under the License is distributed on an “AS IS“ BASIS
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
“““Builds the CIFAR-10 network.
Summary of available functions:
 # Compute input images and labels for training. If you would like to run
 # evaluations use input() instead.
 inputs labels = distorted_inputs()
 # Compute inference on the model inputs to make a prediction.
 predictions = inference(inputs)
 # Compute the total loss of the prediction with respect to the labels.
 loss = loss(predictions labels)
 # Create a graph to run one step of training with respect to the loss.
 train_op = train(loss global_step)
“““
# pylint: disable=missing-docstring
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import gzip
import os
import re
import sys
import tarfile
import tensorflow.python.platform
from six.moves import urllib
import tensorflow as tf
import cifar10_input
FLAGS = tf.app.flags.FLAGS
# Basic model parameters.
tf.app.flags.DEFINE_integer(‘batch_size‘ 128
                            “““Number of images to process in a batch.“““)
tf.app.flags.DEFINE_string(‘data_dir‘ ‘/tmp/cifar10_data‘
                           “““Path to the CIFAR-10 data directory.“““)
# Global constants describing the CIFAR-10 data set.
IMAGE_SIZE = cifar10_input.IMAGE_SIZE
NUM_CLASSES = cifar10_input.NUM_CLASSES
NUM_EXAMPLES_PER_EPOCH_FOR_TRAIN = cifar10_input.NUM_EXAMPLES_PER_EPOCH_FOR_TRAIN
NUM_EXAMPLES_PER_EPOCH_FOR_EVAL = cifar10_input.NUM_EXAMPLES_PER_EPOCH_FOR_EVAL
# Constants describing the training process.
MOVING_AVERAGE_DECAY = 0.9999     # The decay to use for the moving average.
NUM_EPOCHS_PER_DECAY = 350.0      # Epochs after which learning rate decays.
LEARNING_RATE_DECAY_FACTOR = 0.1  # Learning rate decay factor.
INITIAL_LEARNING_RATE = 0.1       # Initial learning rate.
# If a model is trained with multiple GPU‘s prefix all Op names with tower_name
# to differentiate the operations. Note that this prefix is removed from the
# names of the summaries when visualizing a model.
TOWER_NAME = ‘tower‘
DATA_URL = ‘http://www.cs.toronto.edu/~kriz/cifar-10-binary.tar.gz‘
def _activation_summary(x):
  “““Helper to create summaries for activations.
  Creates a summary that provides a histogram of activations.
  Creates a summary that measure the sparsity of activations.
  Args:
    x: Tensor
  Returns:
    nothing
  “““
  # Remove ‘tow

 属性            大小     日期    时间   名称
----------- ---------  ---------- -----  ----
     文件       13976  2017-06-16 13:42  cifar10.py
     文件        5636  2017-06-16 11:11  cifar10_eval.py
     文件        9239  2017-06-16 13:17  cifar10_input.py
     文件        4623  2017-06-16 13:51  cifar10_train.py

评论

共有 条评论