pkmital/tensorflow_tutorials

★ 5,667Jupyter NotebookAudience · researcherComplexity · 2/5Setup · moderate

Mindmap

mindmap
  root((TF Tutorials))
    Topics
      TF basics
      Regression
      Convolutional nets
      Autoencoders
      Residual nets
    Formats
      Python scripts
      Jupyter Notebooks
    Setup
      CPU local install
      GPU on Amazon EC2
    Audience
      ML beginners
      Researchers

mindmap root((TF Tutorials)) Topics TF basics Regression Convolutional nets Autoencoders Residual nets Formats Python scripts Jupyter Notebooks Setup CPU local install GPU on Amazon EC2 Audience ML beginners Researchers

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Run step-by-step Jupyter Notebook examples to learn how TensorFlow handles computation and neural network training

USE CASE 2

Study convolutional network and autoencoder implementations to understand classic deep learning architectures

USE CASE 3

Use the Amazon EC2 setup guide to run GPU-accelerated training on cloud instances

USE CASE 4

Read side-by-side Python scripts alongside notebooks to understand how each concept maps to real code

Tech stack

PythonTensorFlowJupyter NotebookCUDAAmazon EC2

Getting it running

Difficulty · moderate Time to first run · 30min

Uses TensorFlow 1.x APIs from 2016, some code will not run with current TensorFlow without modification.

In plain English

This repository is a collection of tutorials for TensorFlow, a popular library used to build machine-learning models. The tutorials were written in January 2016 by Parag K. Mital and cover a progression of topics from the very basics up to more advanced neural network architectures. Each tutorial comes as both a standalone Python script and as a Jupyter Notebook, which is an interactive document format that lets you run code in small steps and see results inline. The series starts with how TensorFlow is set up and how it handles computation, then moves through regression techniques, and then into various kinds of neural networks: convolutional networks for image recognition, autoencoders for compressing and reconstructing data, a denoising autoencoder, a variational autoencoder, and a residual network. The README includes links to installation guides for TensorFlow on different operating systems, as well as instructions for getting it running on cloud GPU instances on Amazon EC2 for those who need more computing power. Pre-compiled installation files for Ubuntu with Python 3 and CUDA support are included in the repository. This is a learning resource rather than a production library. It is aimed at people who want to understand how these machine-learning concepts work by reading and running real code examples. The repository reflects TensorFlow as it existed in early 2016, so some APIs and practices will differ from current versions of the framework.

Copy-paste prompts

Prompt 1

Using the pkmital/tensorflow_tutorials variational autoencoder notebook, walk me through how the encoder compresses input images and the decoder reconstructs them

Prompt 2

I want to run the convolutional network tutorial from this repo, what do I need to install and how do I launch the Jupyter Notebook locally?

Prompt 3

Translate the autoencoder example from this TensorFlow 1.x tutorial into modern TensorFlow 2.x or PyTorch syntax

Prompt 4

Explain what a residual network is using the example in this tutorial series, in plain language for someone who knows basic Python

Prompt 5

How do I set up TensorFlow with CUDA on Ubuntu to run the GPU-accelerated examples from this tutorial on Amazon EC2?

Open on GitHub → Explain another repo

← pkmital on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.