Introduction To Tensorflow

I have made 3 notebooks on Introduction to Tensorflow:

1. Tensorflow_Hello_World.ipynb

I have picked a parabolic log function and tried to train a shallow deep network on a few samples to generate the same function. The goal of this notebook is to give you an idea that how to build a simple network in Tensorflow.

https://colab.research.google.com/github/rayyan17/Introduction-To-Tensor-Flow/blob/master/Tensorflow_Hello_World.ipynb

2. Tensorflow_Intro_NLP.ipynb

I have picked a dataset from kaggle related to fake-news and tired to show the steps you need to go through for solving an NLP based problem

3. Tensorflow_Computer_Vision.ipynb

I have taken the motivation to build this notebook from "TensorFlow in Practice Specialization - Coursera". It has covered important concepts of CV like convolution and pooling. It has also shown how to visualize images in the layers. https://colab.research.google.com/github/rayyan17/Introduction-To-Tensor-Flow/blob/master/Tensorflow_Computer_Vision.ipynb

Presentation

You can find link to the presentation related to this at: https://www.slideshare.net/rayyankhalid35/introduction-to-tensorflow-213058272

Instructions for setup

1. Download and install python from python.org Version: minimum python version required 3.5 URL: https://www.python.org/downloads/

2. Get the notebooks codebase

   git clone https://github.com/rayyan17/Introduction-To-Tensor-Flow.git <path-to-project-dir>

3. Create and activate virtual environment.

   cd <path-to-project-dir>
   python3 -m venv .python3_venv       # creates the virtual env
   source .python3_venv/bin/activate   # activates the virtual env

   After activating the virtual env. Any installed python packages would be installed here without affecting the main python binaries. To de-activate the virtual env, just type "deactivate"

4. Install Required Packages

   pip install -r requirements.txt

Run your Notebook

Run you Notebook using the following command:

```bash
jupyter notebook NOTEBOOK_NAME
```