Here is the simple instructions to set up a EC2 instance to run machine learning and deep learning on AWS
1. Run an EC2 instance from AWS
Login the AWS account and switch to Oregon. Click on the Community AMI on the left panel.
Search ami-813110b1 and lunch the instance
Choose t2.micro the free tier
2. Install and configure Anaconda
Download Anaconda
$ wget https://repo.continuum.io/archive/Anaconda2-4.2.0-Linux-x86_64.sh
- Install with bash
$ bash Anaconda2-4.2.0-Linux-x86_64.sh
- Configure Anaconda
Start python
>>> from notebook.auth import passwd
>>> passwd()
Copy the passwrod here and paste to the configuration file
$ jupyter notebook --generate-config
Open ./jupyter/jupyter_notebook_config.py
c.NotebookApp.ip = '*'
c.NotebookApp.password = u'sha1:67c9e60bb8b6:9ffede0825894254b2e042ea597d771089e11aed'
$ pip install tensorflow
$ pip install keras
4. Start jupyter notebook
$ jupyter notebook --no-browser --port=8888
Open the browser and enter DNS the address
http://ec2-35-161-215-37.us-west-2.compute.amazonaws.com:8888