nicknochnack/imageclassification

Jupyter notebook showing how to build an image classifier with Python and Tensorflow

imageclassification's Issues

Tried creating a confusion matrix but kept getting errors

Can you guide me on how to create a confusion matrix for the code?

Issue in Importing Libraires

The error is in installing TensorFlow.I don't know to tackle it.

Google Colab GPU Error

can you patch Getting Started.ipynb so that it can run on google colab
on cpu it trains perfectly (except taking a lot of time)
but on gpu it throws a error of something like this:

Node: 'sequential/conv2d/Conv2D'
DNN library is not found.
[[{{node sequential/conv2d/Conv2D}}]]

whats the problem?🤔

data

Having issues dowloading the data from github

Multiple cases

Hi @nicknochnack is there any way to implement this code for multiple cases, I'm quite new to this and decided to use your tutorial and I remember hearing you talk about image classification for multiple cases for a future video but I don't remember you posting anything.
So any suggestion on how I should proceed?
Thanks in advance!

cannot import name 'builder' from 'google.protobuf.internal'

while running this code
import tensorflow as tf
import os

GETTING THE FOLLOWING ERROR

ImportError Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_12388\510189925.py in
----> 1 import tensorflow as tf
2 import os

~\AppData\Roaming\Python\Python39\site-packages\tensorflow_init_.py in
36 import typing as _typing
37
---> 38 from tensorflow.python.tools import module_util as _module_util
39 from tensorflow.python.util.lazy_loader import LazyLoader as _LazyLoader
40

~\AppData\Roaming\Python\Python39\site-packages\tensorflow\python_init_.py in
35
36 from tensorflow.python import pywrap_tensorflow as _pywrap_tensorflow
---> 37 from tensorflow.python.eager import context
38
39 # pylint: enable=wildcard-import

~\AppData\Roaming\Python\Python39\site-packages\tensorflow\python\eager\context.py in
26 import numpy as np
27
---> 28 from tensorflow.core.framework import function_pb2
29 from tensorflow.core.protobuf import config_pb2
30 from tensorflow.core.protobuf import rewriter_config_pb2

~\AppData\Roaming\Python\Python39\site-packages\tensorflow\core\framework\function_pb2.py in
3 # source: tensorflow/core/framework/function.proto
4 """Generated protocol buffer code."""
----> 5 from google.protobuf.internal import builder as _builder
6 from google.protobuf import descriptor as _descriptor
7 from google.protobuf import descriptor_pool as _descriptor_pool

ImportError: cannot import name 'builder' from 'google.protobuf.internal'

More than happy or sad

i wanted to try it for something bigger like more than 2 car models but the acc was so low, around 0.2. in fact by adding more different car models it gets worse. how can i make it work?

Thank for your code.. but load_model path is wrong ?

Hi, nicknochnack

I tried clone and run your code.
I have Intel Mac Pro <2019> then I have no GPU. So, I edit your code for install libs and GPU option

!pip install tensorflow opencv-python matplotlib sentencepiece keras
!pip install torch transformers text2sql

tf.config.list_physical_devices('CPU')
#tf.config.list_physical_devices('GPU')

AS-IS : error
new_model = load_model('imageclassifier.h5')

TO-BE
new_model = load_model('models/imageclassifier.h5')

Is it right?

Thanks

255.255.255

ImageClassification/Getting Started.ipynb

Line 626 in ecbc2b2

"resize = tf.image.resize(img, (256,256))\n",

I'm running this line but the output is always 255.255.255
Meaning every image is sad. Any ideas please?

Data directory not found

ImageClassification/Getting Started.ipynb

Line 164 in ecbc2b2

"data = tf.keras.utils.image_dataset_from_directory('data')"

NotFoundError Traceback (most recent call last)
in <cell line: 1>()
----> 1 data = tf.keras.utils.image_dataset_from_directory('data')

2 frames
/usr/local/lib/python3.10/dist-packages/tensorflow/python/lib/io/file_io.py in list_directory_v2(path)
766 """
767 if not is_directory(path):
--> 768 raise errors.NotFoundError(
769 node_def=None,
770 op=None,

NotFoundError: Could not find directory data