Python Neural Networks – Tensorflow 2.0 Tutorial – Creating a Model
[ad_1]
This python neural network tutorial covers how to create a model using tensorflow 2.0 and keras. We will then train the model on our dataset and have it predict the classification of our test data.
Playlist: https://www.youtube.com/watch?v=OS0Ddkle0o4&list=PLzMcBGfZo4-lak7tiFDec5_ZMItiIIfmj
Text-Based Tutorial: https://techwithtim.net/tutorials/python-neural-networks/creating-a-model/
Tensorflow Website: https://www.tensorflow.org/alpha/tutorials/keras/basic_classification
◾◾◾◾◾
💻 Enroll in The Fundamentals of Programming w/ Python
https://tech-with-tim.teachable.com/p…
📸 Instagram: https://www.instagram.com/tech_with_tim
🌎 Website https://techwithtim.net
📱 Twitter: https://twitter.com/TechWithTimm
⭐ Discord: https://discord.gg/pr2k55t
📝 LinkedIn: https://www.linkedin.com/in/tim-rusci…
📂 GitHub: https://github.com/techwithtim
🔊 Podcast: https://anchor.fm/tech-with-tim
💵 One-Time Donations: https://www.paypal.com/donate/?token=…
💰 Patreon: https://www.patreon.com/techwithtim
◾◾◾◾◾◾
⚡ Please leave a LIKE and SUBSCRIBE for more content! ⚡
Tags:
– Tech With Tim
– Neural Network Tutorial
– Python Neural Network
– Python Tutorials
– Creating a model tensorflow
Source
[ad_2]
your explanations are so clear man, makes it so much easy to understand!
I just did exact same process that you done, but my model run on 1875 pieces of data and test 313 of data's. I should mention that I run in Pycharm and Conda environment. I'm total beginner, but I would be thankful if you tell me why is this happening?
how you calculated the hidden layers as you say 128 ?
Tim and Laurence Moroney are the 2 highest-caliber instructors in the CNN arena. Your knowledge share is appreciated!
5:46 wouldn't you only have 10 biases because each output neuron gets a bias
If anyone facing errors, try this:
from tensorflow import keras
from tensorflow import keras as k
from tensorflow.keras import layers as l
ps : I am using pycharm python 3.7 tensor 2.0 and keras 2.2.5
Sir why 10 output? How to find out that? And how we set the number of neuron, what I mean is 128 or 64 how we calculate that?
I feel bad about you ..
The Youtubers that dont teach better than u are more famous!
YOU ARE OP!
Mantap videonya.
Saya juga ada nih rekomendasi channel lain buat belajar neural network siapa tau cocok hehe.
https://youtu.be/vyAsO_fzNF8
I got 93percent accuracy by adding 5 hidden layers of 128 neurons
I have a 0.1 percent accuracy help
I am trying to run this
model = keras.Sequential([
keras.layers.Flatten(input_shape= (28,28))
keras.layers.Dense(128, activation="relu")
keras.layers.Dense(10,activation="softmax")
])
But I am getting a syntax error, help
Hey what is the purpose of hidden layers??
Thanks Tim
what is the application your using to input the program commands?
Great Video. A question :
How to decide the number of neurons in the hidden layer ?
How can we use other image as input?
Thanks alot Tim! This is insanely helpful to get a first hands on dive before starting serious ML courses on udacity. I am still wondering though why you picked a hidden layer of 128 neurons? Sadly you didn't explain the architecture and why you chose it, is there some more information about that?
Can anyone help me to know why we are using the hidden layer…? Will our network don't work without it ?
DLL load failed
shit vid
10:36
This is an awesome video. Helped so much. Thank you!
I ran the 5, 10 and 15 epochs, and always I got accuracy: 0.1. What is happening?
Hey Tim I am just now going through your 7 hours course and have just a little question on this part. I would like to save the trained model to use things like that in programms, but sadly you didn't explain how to safe those models. I already tried to pickle.dump() them in a file but i am getting some kind of error. would be verry glad to get help on that.
It says it has no loss to optimize 🙁
Thanks Tim, this is really helpful. Much appreciated 🙂
i got this error:
10000/1 [================================…….
test_loss, test_acc = model.evaluate(test_images, test_labels)
TypeError: 'numpy.float64' object is not iterable
and accuracy around: 35%
quick question. I tested my neural network and it had fine accuracy but when it auto tested it became significantly less. For example my accuracy for the last epoch was .90 but when it said Tested acc then it was .09. Any Help?
Could you please give a hint on how you run this through cmd? If I run it through the tensorflow interface, it only ends up spitting out errors.
and i had been wondering who you look like for all this time….
now i realise that it's Jake from two and a half men!
I ran the 5, 10 and 15 epochs, and always I got accuracy: 0.1. What is happening?
Why hidden layer is 128 neurons in this case and how you chose it based on input 784? 🙄forgive me if this is a dumb question😂but I need answer please
Hi Tim! I tried to save the model like you did in a ML video with the pickle module but i couldnt do it. The error says that i cant pickle "_thread.RLock objetcs". If you could help me, that would be awesome. Thank you and your tutorials are great!
This is actually a great video, but my question is: the y_train value is going to be the train_label (the output) in this case?
the error i get is saying Failed to load the native TensorFlow runtime.
this is crazy. My mind is blowing. Thank you so much Tim
Nice video man..
Good job.❤
Running it on an old pc and it takes very long :')
When we choose optimizer and loss function during compile our model, is it very important what we choose? Some kind may not match to our model or be unappropriate? I looked into keras documentation and I saw so many different options.
You are a very underrated youtuber
Hey Tim, your channel is truly great! You are very pedagogical and I'm learning so much.
I'm using TF 2.0 with a GTX 1070 and every epoch is solved in about 3.5s, but you are doing the same in 2.3s on a CPU. Do you have some beastly processor or is there a problem with my install? Thanks
hi Tim! thanks for the tutorial
sorry if you have explained this question, but how you determine to use 'adam' for your optimizer and 'sparse …' for your loss?? is it any consideration?
Tim : "I never saw tutorials that explained layers'
Me: "that's why I am here"
I think: Sparse_cato… is for arrays with one int and just catoro… are for arrays with more than one int so for this project just catorori…cross.. would be better?! Can someone correct me if im wrong?