Tensor Flow Estimator模板基于模型的保存和恢复

Question

我从tensorflow github中获取了神经网络的简单示例，并尝试将其分为两部分。第一部分是训练+测试，第二部分是分离出需要恢复的测试部分。恢复似乎有效，但无法找到预测功能。

以下是第一部分：

from __future__ import print_function

from tensorflow.python.saved_model import builder as saved_model_builder

# Import MNIST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("/tmp/data/", one_hot=False)

import tensorflow as tf
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
import shutil

matplotlib.use('TkAgg')

# Parameters
learning_rate = 0.1
num_steps = 1000
batch_size = 128
display_step = 100

# Network Parameters
n_hidden_1 = 256 # 1st layer number of neurons
n_hidden_2 = 256 # 2nd layer number of neurons
num_input = 784 # MNIST data input (img shape: 28*28)
num_classes = 10 # MNIST total classes (0-9 digits)

#init = tf.initialize_all_variables()

sess = tf.Session()

# Define the input function for training
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': mnist.train.images}, y=mnist.train.labels,
    batch_size=batch_size, num_epochs=None, shuffle=True)

# Define the neural network
def neural_net(x_dict):
    # TF Estimator input is a dict, in case of multiple inputs
    x = x_dict['images']
    # Hidden fully connected layer with 256 neurons
    layer_1 = tf.layers.dense(x, n_hidden_1, name="layer_1")
    # Hidden fully connected layer with 256 neurons
    layer_2 = tf.layers.dense(layer_1, n_hidden_2, name="layer_2")
    # Output fully connected layer with a neuron for each class
    out_layer = tf.layers.dense(layer_2, num_classes, name="out_layer")
    return out_layer

# Define the model function (following TF Estimator Template)
def model_fn(features, labels, mode):
    # Build the neural network
    logits = neural_net(features)

    # Predictions
    pred_classes = tf.argmax(logits, axis=1)
    pred_probas = tf.nn.softmax(logits)

    # If prediction mode, early return
    if mode == tf.estimator.ModeKeys.PREDICT:
        return tf.estimator.EstimatorSpec(mode, predictions=pred_classes) 

    # Define loss and optimizer
    loss_op = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(
        logits=logits, labels=tf.cast(labels, dtype=tf.int32)))
    optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
    train_op = optimizer.minimize(loss_op, global_step=tf.train.get_global_step())

    # Evaluate the accuracy of the model
    acc_op = tf.metrics.accuracy(labels=labels, predictions=pred_classes)

    # TF Estimators requires to return a EstimatorSpec, that specify
    # the different ops for training, evaluating, ...
    estim_specs = tf.estimator.EstimatorSpec(
      mode=mode,
      predictions=pred_classes,
      loss=loss_op,
      train_op=train_op,
      eval_metric_ops={'accuracy': acc_op})

    return estim_specs

# Build the Estimator
model = tf.estimator.Estimator(model_fn)

# Train the Model
model.train(input_fn, steps=num_steps)

# Evaluate the Model
# Define the input function for evaluating
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': mnist.test.images}, y=mnist.test.labels,
    batch_size=batch_size, shuffle=False)
# Use the Estimator 'evaluate' method
model.evaluate(input_fn)

#model.export_savedmodel(".", input_fn)

init = tf.global_variables_initializer()
sess.run(init)

tf.add_to_collection("nn_model", model)

# Add ops to save and restore all the variables.
#saver = tf.train.Saver()

#save_path = saver.save(sess, "model/model.ckpt")

try:
    shutil.rmtree("model")
except:
    pass

builder = saved_model_builder.SavedModelBuilder("model")
builder.add_meta_graph_and_variables(sess, ["nn"])
builder.save()

print("Model saved in file")

# Predict single images
n_images = 4
# Get images from test set
test_images = mnist.test.images[:n_images]
# Prepare the input data
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': test_images}, shuffle=False)
# Use the model to predict the images class
preds = list(model.predict(input_fn))

# Display
for i in range(n_images):
    plt.imshow(np.reshape(test_images[i], [28, 28]), cmap='gray')
    plt.show()
    print("Model prediction:", preds[i])

上述程序运行正常。它保存模型，不确定正确，因为我看到正在创建的所有目录。虽然它确实给出了一个警告：

警告：tensorflow：序列化nn_model时遇到错误。 类型不受支持，或者项目类型与CollectionDef中的字段类型不匹配。 “Estimator”对象没有“name”属性

这是“apply”程序，它会在predict（）行恢复并尝试应用并失败：

import tensorflow as tf

# Import MNIST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("/tmp/data/", one_hot=False)

sess=tf.Session()
#First let's load meta graph and restore weights
#saver = tf.train.import_meta_graph('model/model.ckpt.meta')
#saver.restore(sess,tf.train.latest_checkpoint('nn_model'))
tf.saved_model.loader.load(sess, ["nn"], "model")

model = tf.get_collection('nn_model')

# Predict single images
n_images = 4
# Get images from test set
test_images = mnist.test.images[:n_images]
# Prepare the input data
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': test_images}, shuffle=False)
# Use the model to predict the images class
preds = list(model.predict(input_fn))

# Display
for i in range(n_images):
    plt.imshow(np.reshape(test_images[i], [28, 28]), cmap='gray')
    plt.show()
    print("Model prediction:", preds[i])

它给出的错误是：

Traceback（最近一次调用最后一次）：   文件“applynn.py”，第35行，in     preds = list（model.predict（input_fn）） AttributeError：'module'对象没有属性'predict'

那么这里缺少什么？

Answer 1

所以这个问题现在已经解决了。以下是我必须采取的措施来解决这个问题。

第一部分是：

from __future__ import print_function

from tensorflow.python.saved_model import builder as saved_model_builder

# Import MNIST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("/tmp/data/", one_hot=False)

import tensorflow as tf
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
import shutil

matplotlib.use('TkAgg')

# Parameters
learning_rate = 0.1
num_steps = 1000
batch_size = 128
display_step = 100

# Network Parameters
n_hidden_1 = 256 # 1st layer number of neurons
n_hidden_2 = 256 # 2nd layer number of neurons
num_input = 784 # MNIST data input (img shape: 28*28)
num_classes = 10 # MNIST total classes (0-9 digits)

#init = tf.initialize_all_variables()

sess = tf.Session()

# Define the input function for training
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': mnist.train.images}, y=mnist.train.labels,
    batch_size=batch_size, num_epochs=None, shuffle=True)

# Define the neural network
def neural_net(x_dict):
    # TF Estimator input is a dict, in case of multiple inputs
    x = x_dict['images']
    # Hidden fully connected layer with 256 neurons
    layer_1 = tf.layers.dense(x, n_hidden_1, name="layer_1")
    # Hidden fully connected layer with 256 neurons
    layer_2 = tf.layers.dense(layer_1, n_hidden_2, name="layer_2")
    # Output fully connected layer with a neuron for each class
    out_layer = tf.layers.dense(layer_2, num_classes, name="out_layer")
    return out_layer

# Define the model function (following TF Estimator Template)
def model_fn(features, labels, mode):
    # Build the neural network
    logits = neural_net(features)

    # Predictions
    pred_classes = tf.argmax(logits, axis=1)
    pred_probas = tf.nn.softmax(logits)

    # If prediction mode, early return
    if mode == tf.estimator.ModeKeys.PREDICT:
        return tf.estimator.EstimatorSpec(mode, predictions=pred_classes) 

    # Define loss and optimizer
    loss_op = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(
        logits=logits, labels=tf.cast(labels, dtype=tf.int32)))
    optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
    train_op = optimizer.minimize(loss_op, global_step=tf.train.get_global_step())

    # Evaluate the accuracy of the model
    acc_op = tf.metrics.accuracy(labels=labels, predictions=pred_classes)

    # TF Estimators requires to return a EstimatorSpec, that specify
    # the different ops for training, evaluating, ...
    estim_specs = tf.estimator.EstimatorSpec(
      mode=mode,
      predictions=pred_classes,
      loss=loss_op,
      train_op=train_op,
      eval_metric_ops={'accuracy': acc_op})

    return estim_specs

# Build the Estimator
estimator = tf.estimator.Estimator(model_fn, model_dir='estimator')

# Train the Model
estimator.train(input_fn, steps=num_steps)

# Evaluate the Model
# Define the input function for evaluating
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': mnist.test.images}, y=mnist.test.labels,
    batch_size=batch_size, shuffle=False)
# Use the Estimator 'evaluate' method
estimator.evaluate(input_fn)

#model.export_savedmodel(".", input_fn)

init = tf.global_variables_initializer()
sess.run(init)

tf.add_to_collection("nn_model", estimator)

# Add ops to save and restore all the variables.
#saver = tf.train.Saver()

#save_path = saver.save(sess, "model/model.ckpt")

try:
    shutil.rmtree("model")
except:
    pass

builder = saved_model_builder.SavedModelBuilder("model")
builder.add_meta_graph_and_variables(sess, ["nn"])
builder.save()

print("Model saved in file")

# Predict single images
n_images = 4
# Get images from test set
test_images = mnist.test.images[:n_images]
# Prepare the input data
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': test_images}, shuffle=False)
# Use the model to predict the images class
preds = list(estimator.predict(input_fn))

# Display
for i in range(n_images):
    plt.imshow(np.reshape(test_images[i], [28, 28]), cmap='gray')
    plt.show()
    print("Model prediction:", preds[i])

第二部分是：

import tensorflow as tf
import matplotlib.pyplot as plt
import numpy as np

# Import MNIST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("/tmp/data/", one_hot=False)

# Network Parameters
n_hidden_1 = 256 # 1st layer number of neurons
n_hidden_2 = 256 # 2nd layer number of neurons
num_classes = 10 # MNIST total classes (0-9 digits)

# Define the neural network
def neural_net(x_dict):
    # TF Estimator input is a dict, in case of multiple inputs
    x = x_dict['images']
    # Hidden fully connected layer with 256 neurons
    layer_1 = tf.layers.dense(x, n_hidden_1, name="layer_1")
    # Hidden fully connected layer with 256 neurons
    layer_2 = tf.layers.dense(layer_1, n_hidden_2, name="layer_2")
    # Output fully connected layer with a neuron for each class
    out_layer = tf.layers.dense(layer_2, num_classes, name="out_layer")
    return out_layer

# Define the model function (following TF Estimator Template)
def model_fn(features, labels, mode):
    # Build the neural network
    logits = neural_net(features)

    # Predictions
    pred_classes = tf.argmax(logits, axis=1)
    pred_probas = tf.nn.softmax(logits)

    # If prediction mode, early return
    if mode == tf.estimator.ModeKeys.PREDICT:
        return tf.estimator.EstimatorSpec(mode, predictions=pred_classes)

    # Define loss and optimizer
    loss_op = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(
        logits=logits, labels=tf.cast(labels, dtype=tf.int32)))
    optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)
    train_op = optimizer.minimize(loss_op, global_step=tf.train.get_global_step())

    # Evaluate the accuracy of the model
    acc_op = tf.metrics.accuracy(labels=labels, predictions=pred_classes)

    # TF Estimators requires to return a EstimatorSpec, that specify
    # the different ops for training, evaluating, ...
    estim_specs = tf.estimator.EstimatorSpec(
      mode=mode,
      predictions=pred_classes,
      loss=loss_op,
      train_op=train_op,
      eval_metric_ops={'accuracy': acc_op})

    return estim_specs


sess=tf.Session()

estimator = tf.estimator.Estimator(model_fn, model_dir='estimator')

# Predict single images
n_images = 4
# Get images from test set
test_images = mnist.test.images[:n_images]
# Prepare the input data
input_fn = tf.estimator.inputs.numpy_input_fn(
    x={'images': test_images}, shuffle=False)
# Use the model to predict the images class
preds = list(estimator.predict(input_fn))

# Display
for i in range(n_images):
    plt.imshow(np.reshape(test_images[i], [28, 28]), cmap='gray')
    plt.show()
    print("Model prediction:", preds[i])

请注意，我已将模型变量称为估算器，因为它实际上是估算器。另外，我传递的是model_dir，因此将估算器与其他变量分开序列化。我还必须明确确保第二个python文件可以访问它们依赖的两个函数和任何变量。在代码中做了几个小的其他修复。