After understanding how neural networks work, implementing some basic architectures using deep learning frameworks, and learning about some advanced techniques to help enhance our neural networks' models results, it's time to apply what you learned! So let's start
In this project, you will build a neural network model to classify images from CIFAR 10 dataset.
The CIFAR-10 dataset consists of 60000 32x32 color images of 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images. The dataset is divided into five training batches and one test batch, each with 10000 images. The test batch contains exactly 1000 randomly-selected images from each class. The training batches contain the remaining images in random order, but some training batches may contain more images from one class than another. Between them, the training batches contain exactly 5000 images from each class. source
Unlike the previous projects, there will be no code cells to fill, the only task you have is to build the best possible model using the techniques you learned about in this unit. But we will guide you with some directives.
You will have enough guidance throughout the project and your work will be reviewed and graded by a teacher assistant. You can also reach out to the TA via slack whenever you feel you are stuck.
- Please use text cells to write the questions' answers in a good way.
- Don't forget to save the different models you tested so you will be able to report the different results you got and the impact of the different techniques you tested later.
In case you don't have a GPU, it is recommended that you use google colab. Start by cloning this repository, then open google colab, click on File > Upload notebook, and finally upload the
.ipynbfile from the repository you have just cloned! Don't forget to change the runtime to GPU. If you want to work in your local environment just open it using jupyter notebook.
- Fork this repository into your Github account. To do so, click on the fork button in the upper right hand corner of a repo page.
- Head to the forked version on your github then clone it on your local space.
- Open the notebook.
Feel free to manage your versions as you want.
- Once you finished your project and you are confident about the results, You have to push those details to your forked version of the project The following Git commands should be helpful for you as a reminder :
$git add .$git commit -m "commit msg"$git push origin main
Once you're done working on the project, submit the link to you repository in the platform.