我在这里遵循MNIST教程：https://www.tensorflow.org/tutorials/layers。我想注销每100步的准确度。我尝试修改cnn_model_fn中的火车部分，但它不起作用。
以下是我的修改：
if mode == tf.estimator.ModeKeys.TRAIN:
optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.001)
train_op = optimizer.minimize(
loss=loss,
global_step=tf.train.get_global_step())
logging_hook = tf.train.LoggingTensorHook({"accuracy": accuracy}, every_n_iter=100)
return tf.estimator.EstimatorSpec(mode=mode, loss=loss, train_op=train_op, training_hooks=[logging_hook])
我在accuracy
if
accuracy = tf.metrics.accuracy(labels=labels, predictions=predictions["classes"])
但我收到了以下错误
Extracting MNIST-data/train-images-idx3-ubyte.gz
Extracting MNIST-data/train-labels-idx1-ubyte.gz
Extracting MNIST-data/t10k-images-idx3-ubyte.gz
Extracting MNIST-data/t10k-labels-idx1-ubyte.gz
2018-05-04 18:54:05.819366: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2018-05-04 18:54:05.819388: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2018-05-04 18:54:05.819396: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2018-05-04 18:54:05.819402: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2018-05-04 18:54:05.819408: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
INFO:tensorflow:Create CheckpointSaverHook.
Traceback (most recent call last):
File "cnn_mnist.py", line 119, in <module>
tf.app.run()
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "cnn_mnist.py", line 102, in main
steps=2000)
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/estimator/estimator.py", line 217, in train
loss = self._train_model(input_fn=input_fn, hooks=hooks)
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/estimator/estimator.py", line 577, in _train_model
config=config_pb2.ConfigProto(allow_soft_placement=True)) as mon_sess:
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/monitored_session.py", line 333, in MonitoredTrainingSession
stop_grace_period_secs=stop_grace_period_secs)
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/monitored_session.py", line 627, in __init__
stop_grace_period_secs=stop_grace_period_secs)
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/monitored_session.py", line 449, in __init__
h.begin()
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/basic_session_run_hooks.py", line 162, in begin
for (tag, tensor) in self._tensors.items()}
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/basic_session_run_hooks.py", line 162, in <dictcomp>
for (tag, tensor) in self._tensors.items()}
File "/Users/caitlinwen/miniconda2/lib/python2.7/site-packages/tensorflow/python/training/basic_session_run_hooks.py", line 688, in _as_graph_element
"to current graph %s." % (obj, graph))
ValueError: Passed (<tf.Tensor 'accuracy/value:0' shape=() dtype=float32>, <tf.Tensor 'accuracy/update_op:0' shape=() dtype=float32>) should have graph attribute that is equal to current graph <tensorflow.python.framework.ops.Graph object at 0x18139a8310>.
我的完整代码是：
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
# Imports
import numpy as np
import tensorflow as tf
# Our application logic will be added here
def cnn_model_fn(features, labels, mode):
"""Model function for CNN."""
# Input Layer
input_layer = tf.reshape(features["x"], [-1, 28, 28, 1])
# Convolutional Layer #1
conv1 = tf.layers.conv2d(
inputs=input_layer,
filters=32,
kernel_size=[5, 5],
padding="same",
activation=tf.nn.relu)
# Pooling Layer #1
pool1 = tf.layers.max_pooling2d(inputs=conv1, pool_size=[2, 2], strides=2)
# Convolutional Layer #2 and Pooling Layer #2
conv2 = tf.layers.conv2d(
inputs=pool1,
filters=64,
kernel_size=[5, 5],
padding="same",
activation=tf.nn.relu)
pool2 = tf.layers.max_pooling2d(inputs=conv2, pool_size=[2, 2], strides=2)
# Dense Layer
pool2_flat = tf.reshape(pool2, [-1, 7 * 7 * 64])
dense = tf.layers.dense(inputs=pool2_flat, units=1024, activation=tf.nn.relu)
dropout = tf.layers.dropout(
inputs=dense, rate=0.4, training=mode == tf.estimator.ModeKeys.TRAIN)
# Logits Layer
logits = tf.layers.dense(inputs=dropout, units=10)
predictions = {
# Generate predictions (for PREDICT and EVAL mode)
"classes": tf.argmax(input=logits, axis=1),
# Add `softmax_tensor` to the graph. It is used for PREDICT and by the
# `logging_hook`.
"probabilities": tf.nn.softmax(logits, name="softmax_tensor")
}
if mode == tf.estimator.ModeKeys.PREDICT:
return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions)
# Calculate Loss (for both TRAIN and EVAL modes)
loss = tf.losses.sparse_softmax_cross_entropy(labels=labels, logits=logits)
accuracy = tf.metrics.accuracy(labels=labels, predictions=predictions["classes"])
# Configure the Training Op (for TRAIN mode)
if mode == tf.estimator.ModeKeys.TRAIN:
optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.001)
train_op = optimizer.minimize(
loss=loss,
global_step=tf.train.get_global_step())
logging_hook = tf.train.LoggingTensorHook({"accuracy": accuracy}, every_n_iter=100)
return tf.estimator.EstimatorSpec(mode=mode, loss=loss, train_op=train_op, training_hooks=[logging_hook])
# Add evaluation metrics (for EVAL mode)
eval_metric_ops = {
"accuracy": tf.metrics.accuracy(labels=labels, predictions=predictions["classes"])}
return tf.estimator.EstimatorSpec(mode=mode, loss=loss, eval_metric_ops=eval_metric_ops)
def main(unused_argv):
# Load training and eval data
mnist = tf.contrib.learn.datasets.load_dataset("mnist")
train_data = mnist.train.images # Returns np.array
train_labels = np.asarray(mnist.train.labels, dtype=np.int32)
eval_data = mnist.test.images # Returns np.array
eval_labels = np.asarray(mnist.test.labels, dtype=np.int32)
# Create the Estimator
mnist_classifier = tf.estimator.Estimator(model_fn=cnn_model_fn, model_dir="/tmp/mnist_convnet_model")
# Set up logging for predictions
init = tf.global_variables_initializer()
tf.logging.set_verbosity(tf.logging.INFO)
with tf.Session() as sess:
sess.run(init)
# Train the model
train_input_fn = tf.estimator.inputs.numpy_input_fn(
x={"x": train_data},
y=train_labels,
batch_size=100,
num_epochs=None,
shuffle=True)
mnist_classifier.train(
input_fn=train_input_fn,
steps=2000)
# Evaluate the model and print results
eval_input_fn = tf.estimator.inputs.numpy_input_fn(
x={"x": eval_data},
y=eval_labels,
num_epochs=1,
shuffle=False)
eval_results = mnist_classifier.evaluate(input_fn=eval_input_fn)
print(eval_results)
# merged = tf.summary.merge_all()
# train_writer = tf.summary.FileWriter('./train', sess.graph)
# test_writer = tf.summary.FileWriter('./test')
# tf.global_variables_initializer().run()
if __name__ == "__main__":
tf.app.run()
答案 0 :(得分：2)
模块tf.metrics.accuracy返回定义为here的两个参数accuracy和update_op。
因此，您需要将代码更改为：
accuracy, update_op = tf.metrics.accuracy(labels=labels, predictions=predictions["classes"])
答案 1 :(得分：0)
尝试：
accuracy = tf.compat.v1.metrics.accuracy(
labels=labels,
predictions= prediction )