mnist_cnn in keras ON AWS

WHAT WE WILL TALK ABOUT
1.how to setup your working envirement.
2.how to transfer your .py file to the AWS
3.do “hello world” on the AWS
the AMI of ami-084820056c48a24a1 have a user named ec2-user and when you first logon it you will see many .sh file you can use source activate …to setup your envirement eg. “source activate tensorflow_p36” and you can activate the envirement that for tensorflow Keras2 python3.6 CUDA9.0 and MKL-DNN.ofcourse you can activate others .
TO transfer your .py file to the CAKE you need to use “scp” and the format is this scp -i “yourpublickey” /where/and/what/you/want/to/transfer ec2-user@whereyourcake:/where/to/save/the/file/on/ASW
of course you can type it in a local teminal
Do the “hello world” work in ML and DL .we prefer mnist to do this and you can find in/your/Keras/example/ there is mnist_cnn.py scp it to your ASW and python mnist_cnn.py
congratuation your fist work on AWS is done .by the way the spot instance is soooo cheap! i think you can try it