AWS has a limit for uploading docs bigger than 25 MB, the same as GitHub. Machine learning tasks for computer vision will easily have bigger data sets than this. After researching on the web, I found this option is quite good. Using FileZilla will make the life a lot easier.
First you need to download and install FileZilla. Afterwards following the best answer on Connect to Amazon EC2 file directory using Filezilla and SFTP on StackOverflow. For AWS EC2 instances you should choose ‘ubuntu’ as the ‘user’ instead of ‘ec2-user’. Now connect the SFTP site you set with the EC2 instance. When the connection is successful, browse the left side to the source and the right side to the target directory. You can easily pull the docs from the site just established to the EC2 instance.
If you encounter with ‘permission denied’ warning message.
Following the answer on Stackoverflow Amazon Ec2 FTP Write Permission [closed]
In brief, it works like this:
Open up your terminal, execute ‘sudo -s’. This command give you right as root user.
Give write permission to all users by executing ‘chmod 777 /the path you want to write/‘
Now you can pull the files to EC2 instance.
Have fun!