# Accuracy issues on Flux

**URL:** <https://discourse.julialang.org/t/accuracy-issues-on-flux/92016>\
**Category:** Performance\
**Tags:** question, flux\
**Created:** [December 22, 2022, 6:58pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016 "2022-12-22T18:58:33Z")\
**Posts on this page:** 6\
**Page:** 2

<div class="post-metadata">

**Author:** ![Radu\_Mihai\_Diaconu](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/radu_mihai_diaconu/32/45306_2.png) [@Radu\_Mihai\_Diaconu](https://discourse.julialang.org/u/Radu_Mihai_Diaconu)\
**Post date:** [December 28, 2022, 4:06pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/22 "2022-12-28T16:06:23Z")

</div>

around 98%

---

<div class="post-metadata">

**Author:** ![Radu\_Mihai\_Diaconu](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/radu_mihai_diaconu/32/45306_2.png) [@Radu\_Mihai\_Diaconu](https://discourse.julialang.org/u/Radu_Mihai_Diaconu)\
**Post date:** [December 28, 2022, 4:07pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/23 "2022-12-28T16:07:22Z")

</div>

i will quit this task tbh…i is a waste of time with every architecture i cant get why it is not working. i better concentrate on other deep leanring that acually works.

---

<div class="post-metadata">

**Author:** ![mkitti](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mkitti/32/12459_2.png) [@mkitti](https://discourse.julialang.org/u/mkitti)\
**Post date:** [January 1, 2023, 8:31pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/24 "2023-01-01T20:31:36Z")

</div>

Next time please lead with the working Tensorflow code.

---

<div class="post-metadata">

**Author:** ![Radu\_Mihai\_Diaconu](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/radu_mihai_diaconu/32/45306_2.png) [@Radu\_Mihai\_Diaconu](https://discourse.julialang.org/u/Radu_Mihai_Diaconu)\
**Post date:** [January 1, 2023, 8:52pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/25 "2023-01-01T20:52:18Z")

</div>

# Convolutional Neural Network

# Importing the libraries

import tensorflow as tf  
from keras.preprocessing.image import ImageDataGenerator  
tf. **version**

# Part 1 - Data Preprocessing

# Preprocessing the Training set

train\_datagen = ImageDataGenerator(rescale = 1./255,  
shear\_range = 0.2,  
zoom\_range = 0.2,  
horizontal\_flip = True)  
training\_set = train\_datagen.flow\_from\_directory(‘dataset/training\_set’,  
target\_size = (64, 64),  
batch\_size = 32,  
class\_mode = ‘binary’)

# Preprocessing the Test set

test\_datagen = ImageDataGenerator(rescale = 1./255)  
test\_set = test\_datagen.flow\_from\_directory(‘dataset/test\_set’,  
target\_size = (64, 64),  
batch\_size = 32,  
class\_mode = ‘binary’)

# Part 2 - Building the CNN

# Initialising the CNN

cnn = tf.keras.models.Sequential()

# Step 1 - Convolution

cnn.add(tf.keras.layers.Conv2D(filters=32, kernel\_size=3, activation=‘relu’, input\_shape=[64, 64, 3]))

# Step 2 - Pooling

cnn.add(tf.keras.layers.MaxPool2D(pool\_size=2, strides=2))

# Adding a second convolutional layer

cnn.add(tf.keras.layers.Conv2D(filters=32, kernel\_size=3, activation=‘relu’))  
cnn.add(tf.keras.layers.MaxPool2D(pool\_size=2, strides=2))

# Step 3 - Flattening

cnn.add(tf.keras.layers.Flatten())

# Step 4 - Full Connection

cnn.add(tf.keras.layers.Dense(units=128, activation=‘relu’))

# Step 5 - Output Layer

cnn.add(tf.keras.layers.Dense(units=1, activation=‘sigmoid’))

# Part 3 - Training the CNN

# Compiling the CNN

cnn.compile(optimizer = ‘adam’, loss = ‘binary\_crossentropy’, metrics = [‘accuracy’])

# Training the CNN on the Training set and evaluating it on the Test set

cnn.fit(x = training\_set, validation\_data = test\_set, epochs = 25)

# Part 4 - Making a single prediction

import numpy as np  
from keras.preprocessing import image  
test\_image = image.load\_img(‘dataset/single\_prediction/cat\_or\_dog\_1.jpg’, target\_size = (64, 64))  
test\_image = image.img\_to\_array(test\_image)  
test\_image = np.expand\_dims(test\_image, axis = 0)  
result = cnn.predict(test\_image)  
training\_set.class\_indices  
if result[0][0] == 1:  
prediction = ‘dog’  
else:  
prediction = ‘cat’  
print(prediction)

this is the working tensorflow example if it helps you

---

<div class="post-metadata">

**Author:** ![jmair](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jmair/32/35117_2.png) [@jmair](https://discourse.julialang.org/u/jmair)\
**Post date:** [January 1, 2023, 11:04pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/26 "2023-01-01T23:04:27Z")

</div>

Hi, it’s very helpful to use backticks for quoting code (see [Please read: make it easier to help you](https://discourse.julialang.org/t/please-read-make-it-easier-to-help-you/14757)) and so it’s easier to copy paste & have the correct formatting.

---

<div class="post-metadata">

**Author:** ![Radu\_Mihai\_Diaconu](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/radu_mihai_diaconu/32/45306_2.png) [@Radu\_Mihai\_Diaconu](https://discourse.julialang.org/u/Radu_Mihai_Diaconu)\
**Post date:** [January 3, 2023, 3:43pm UTC](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016/27 "2023-01-03T15:43:17Z")

</div>

```julia
# Convolutional Neural Network

# Importing the libraries

import tensorflow as tf
from keras.preprocessing.image import ImageDataGenerator
tf. **version**

# Part 1 - Data Preprocessing

# Preprocessing the Training set

train_datagen = ImageDataGenerator(rescale = 1./255,
shear_range = 0.2,
zoom_range = 0.2,
horizontal_flip = True)
training_set = train_datagen.flow_from_directory(‘dataset/training_set’,
target_size = (64, 64),
batch_size = 32,
class_mode = ‘binary’)

# Preprocessing the Test set

test_datagen = ImageDataGenerator(rescale = 1./255)
test_set = test_datagen.flow_from_directory(‘dataset/test_set’,
target_size = (64, 64),
batch_size = 32,
class_mode = ‘binary’)

# Part 2 - Building the CNN

# Initialising the CNN

cnn = tf.keras.models.Sequential()

# Step 1 - Convolution

cnn.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation=‘relu’, input_shape=[64, 64, 3]))

# Step 2 - Pooling

cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))

# Adding a second convolutional layer

cnn.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation=‘relu’))
cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))

# Step 3 - Flattening

cnn.add(tf.keras.layers.Flatten())

# Step 4 - Full Connection

cnn.add(tf.keras.layers.Dense(units=128, activation=‘relu’))

# Step 5 - Output Layer

cnn.add(tf.keras.layers.Dense(units=1, activation=‘sigmoid’))

# Part 3 - Training the CNN

# Compiling the CNN

cnn.compile(optimizer = ‘adam’, loss = ‘binary_crossentropy’, metrics = [‘accuracy’])

# Training the CNN on the Training set and evaluating it on the Test set

cnn.fit(x = training_set, validation_data = test_set, epochs = 25)

# Part 4 - Making a single prediction

import numpy as np
from keras.preprocessing import image
test_image = image.load_img(‘dataset/single_prediction/cat_or_dog_1.jpg’, target_size = (64, 64))
test_image = image.img_to_array(test_image)
test_image = np.expand_dims(test_image, axis = 0)
result = cnn.predict(test_image)
training_set.class_indices
if result[0][0] == 1:
prediction = ‘dog’
else:
prediction = ‘cat’
print(prediction)

```

ok

[Previous page](https://discourse.julialang.org/t/accuracy-issues-on-flux/92016.md?page=1)