Coding a Deep Convolutional Generative Adversarial Network (DCGAN) from scratch (Python Tensorflow)

Показать описание

Deep Convolutional Generative Adversarial Network (DCGAN) - Code-Along (Python Tensorflow)

Рекомендации по теме

Комментарии

Dude! The structuring of the code is on point. I really want to get in touch with you on linkedin

zahash

Is it also possible to change this in order to have as a Dataset a 2d integer array? I also just want to produce integer arrays

gamedevel

Could you upload DCGAN implementation (data augmentation purpose) using dataset insert from the google drive

lalithavanik

Notice you wrote under the train method: if i % self.epochs_for_sample == 0:
for anyone trying this, it will only print sample 0 so it will appear stuck.
you can remove the condition if you want every sample or change the condition.

unactive

This is one of the most beautiful, elegant and simple DCGAN implementations i have seen till now Thanks. I have few questions though i will be grateful if you can answer.
1) The theory and even the implementation of GAN tells that you train D first and then G. G generates totally random noise in the first few hundred epochs. So, when i use your implementation the losses at the very start for few hundred epochs oppose the theory i.e. G loss is small and D loss is bigger why? Is it possible that G will have small loss than D at the beginning?
2) What these losses actually mean for example D loss 1.223344224 and G loss 0.8933454545 ? Does it mean D did 12% bad and G did 8.9% bad? (This may sound stupid but i am little confused here)

ammarulhassan

I like the coding style, but after complete training network didn't converge. Can you show us the final results... for me: Epoch: 14500. Discriminator loss: 3.917084878679589e-09. Generator loss: 20.475692749023438

muhammadsohail

What kind of snippet do you use for Tensorflow?

itsMike

Are you sharing the code somewhere in a repository?

AndiSteilwandi

Is it possible to use this same model but use our own data? ie using images of our own?

theoclark

Coding a Deep Convolutional Generative Adversarial Network (DCGAN) from scratch (Python Tensorflow)

Coding a Deep Convolutional Generative Adversarial Network (DCGAN) from scratch (Python Tensorflow)

Deep Convolutional Generative Adversarial (DCGAN) Crash Course with Python [2022] Part 1

Deep Convolutional Generative Adversarial (DCGAN) Crash Course with Python [2022] Part 2

Deep Convolutional Generative Adversarial Networks Devlog 1

2023 | Deep Convolutional Generative Adversarial Network, DCGAN #deeplearning #algorithm

🤖Convolutional Neural Networks (CNNs) by #andrewtate and #donaldtrump

Deep Convolutional Generative Adversarial (DCGAN) Crash Course with Python [2022] Final Part

108 DCGAN Deep Convolutional Generative Adversarial Networks

Deep Convolutional Generative Adversarial Networks Devlog 5

Deep Convolutional Generative Adversarial Networks Devlog 3

THIS is HARDEST MACHINE LEARNING model I've EVER coded

Deep Convolutional Generative Adversarial Network (DCGAN) Implementation in TensorFlow

Deep Convolutional Generative Adversarial Networks Devlog 2

Deep Convolutional Generative Adversarial Network MNIST Samples [24500 samples in 1 minute]

Deep Convolutional Generative Adversarial Network Tutorial from Tensorflow Website

Machine Learning | Convolutional Generative Adversarial Networks in 100 lines of PyTorch code

Deep Convolutional Generative Adversarial Network (DCGAN) - Learning Visualized

PyTorch or Tensorflow? Which Should YOU Learn!

Deep Convolutional Generative Adversarial Networks Devlog 4

Build a Generative Adversarial Neural Network with Tensorflow and Python | Deep Learning Projects

Implementing a Deep Convolutional Generative Adversarial Network (DCGAN) in Python

DCGAN | Deep Convolutional Generative Adversarial Network

A Practical Introduction to Deep Convolutional Generative Adversarial Network (DCGAN)

DCGANs | Lecture 65 (Part 1) | Applied Deep Learning