編集履歴

質問編集履歴

タイトル変更

2019/11/28 06:36

投稿

スコア5

test CHANGED Viewed

	@@ -1 +1 @@
1	- 各クラスごとの正解率を出力させ~~たいです、！~~
1	+ cifar-10 各クラスごとの正解率を出力させる方法

test CHANGED Viewed

File without changes

説明の修正

2019/11/28 06:36

投稿

sinugo

スコア5

test CHANGED Viewed

File without changes

test CHANGED Viewed

@@ -2,7 +2,7 @@
-ここに質問の内容を詳しく書いてください。
 今大学でtensorflowを使った雲画像の自動認識を勉強しています。
@@ -14,6 +14,20 @@
+*現在の出力
+epoch ●● duration = ●● sec, prediction = ●●
+*実現したい出力
+epoch ●● class : 0 duration = ●● sec, prediction = ●●
+epoch ●● class : 1 duration = ●● sec, prediction = ●●
+epoch ●● class : 2 duration = ●● sec, prediction = ●●

文の修正

2019/11/26 09:05

投稿

sinugo

スコア5

test CHANGED Viewed

File without changes

test CHANGED Viewed

@@ -316,325 +316,23 @@
 ```
-```python3
-model.py
-# coding: UTF-8
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-import tensorflow.compat.v1 as tf
-tf.disable_v2_behavior()
-NUM_CLASSES = 3
-def _get_weights(shape, stddev=1.0):
-  var = tf.get_variable('weights', shape,
-                        initializer=tf.truncated_normal_initializer(stddev=stddev))
-  return var
-def _get_biases(shape, value=0.0):
-  var = tf.get_variable('biases', shape,
-                        initializer=tf.constant_initializer(value))
-  return var
-def inference(image_node):
-  # conv1
-  with tf.variable_scope('conv1') as scope:
-    weights = _get_weights(shape=[3, 3, 3, 64], stddev=0.05)
-    conv = tf.nn.conv2d(image_node, weights, [1, 1, 1, 1], padding='SAME')
-    biases = _get_biases([64], value=0.1)
-    bias = tf.nn.bias_add(conv, biases)
-    conv1 = tf.nn.relu(bias, name=scope.name)
-  # pool1
-  pool1 = tf.nn.max_pool(conv1, ksize=[1 ,2, 2, 1], strides=[1, 2, 2, 1],
-                         padding='SAME', name='pool1')
-  # norm1
-  norm1 = tf.nn.lrn(pool1, 4,
-                    bias=1.0, alpha=0.001 / 9.0, beta=0.75, name='norm1')
-  # conv2
-  with tf.variable_scope('conv2') as scope:
-    weights = _get_weights(shape=[3, 3, 64, 64], stddev=0.05)
-    conv = tf.nn.conv2d(norm1, weights, [1, 1, 1, 1], padding='SAME')
-    biases = _get_biases([64], value=0.1)
-    bias = tf.nn.bias_add(conv, biases)
-    conv2 = tf.nn.relu(bias, name=scope.name)
-  # norm2
-  norm2 = tf.nn.lrn(conv2, 4,
-                    bias=1.0, alpha=0.001 / 9.0, beta=0.75, name='norm2')
-  # pool2
-  pool2 = tf.nn.max_pool(norm2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1],
-                         padding='SAME', name='pool2')
-  # conv3
-  with tf.variable_scope('conv3') as scope:
-    weights = _get_weights(shape=[3, 3, 64, 128], stddev=0.05)
-    conv = tf.nn.conv2d(pool2, weights, [1, 1, 1, 1], padding='SAME')
-    biases = _get_biases([128], value=0.1)
-    bias = tf.nn.bias_add(conv, biases)
-    conv3 = tf.nn.relu(bias, name=scope.name)
-  # pool3
-  pool3 = tf.nn.max_pool(conv3, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1],
-                         padding='SAME', name='pool3')
-  # norm3
-  norm3 = tf.nn.lrn(pool3, 4,
-                    bias=1.0, alpha=0.001 / 9.0, beta=0.75, name='norm3')
-  reshape = tf.reshape(norm3, [1, -1])
-  dim = reshape.get_shape()[1].value
-  # fc3
-  with tf.variable_scope('fc3') as scope:
-    weights = _get_weights(shape=[dim, 256], stddev=0.05)
-    biases = _get_biases([256], value=0.1)
-    fc3 = tf.nn.relu(tf.matmul(reshape, weights) + biases, name=scope.name)
-  # fc4
-  with tf.variable_scope('fc4') as scope:
-    weights = _get_weights(shape=[256, 128], stddev=0.05)
-    biases = _get_biases([128], value=0.1)
-    fc4 = tf.nn.relu(tf.matmul(fc3, weights) + biases, name=scope.name)
- # output
-  with tf.variable_scope('output') as scope:
-    weights = _get_weights(shape=[128, NUM_CLASSES], stddev=1 / 128.0)
-    biases = _get_biases([NUM_CLASSES], value=0.0)
-    logits = tf.add(tf.matmul(fc4, weights), biases, name='logits')
-  return logits
-```
-```python3
-reader.py
-# coding: UTF-8
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-import os
-import numpy as np
-class Cifar10Record(object):
-  width = 32
-  height = 32
-  depth = 3
-  def set_label(self, label_byte):
-    self.label = np.frombuffer(label_byte, dtype=np.uint8)
-  def set_image(self, image_bytes):
-    byte_buffer = np.frombuffer(image_bytes, dtype=np.int8)
-    reshaped_array = np.reshape(byte_buffer,
-                                [self.depth, self.height, self.width])
-    self.byte_array = np.transpose(reshaped_array, [1, 2, 0])
-    self.byte_array = self.byte_array.astype(np.float32)
-class Cifar10Reader(object):
-  def __init__(self, filename):
-    if not os.path.exists(filename):
-      print(filename + ' is not exist')
-      return
-    self.bytestream = open(filename, mode="rb")
-  def close(self):
-    if not self.bytestream:
-      self.bytestream.close()
-  def read(self, index):
-    result = Cifar10Record()
-    label_bytes = 1
-    image_bytes = result.height * result.width * result.depth
-    record_bytes = label_bytes + image_bytes
-    self.bytestream.seek(record_bytes * index, 0)
-    result.set_label(self.bytestream.read(label_bytes))
-    result.set_image(self.bytestream.read(image_bytes))
-    return result
-```
 ### 試したこと
+_eval関数をもう１つ作って、
+ prediction = _eval(sess, top_k_op,
-いくつかサイトを調べて自分のプログラムに合うように調整してみたのですが、
+                         train_placeholder, label_placeholder)
+      print('epoch %d duration = %d sec, prediction = %.3f'
+            % (epoch, duration, prediction))
-なかなかうまく出力できませんでした、、、
+をfor文で回そうとしたのですがうまく行きませんでした。
 ### 補足情報（FW/ツールのバージョンなど）

2019/11/22 08:38

投稿

sinugo

スコア5

test CHANGED Viewed

	@@ -1 +1 @@
1	- 各クラスごとの正解率を出力させたいです、、！
1	+ 各クラスごとの正解率を出力させたいです、！

test CHANGED Viewed

File without changes