Skip to content

Latest commit

 

History

History
37 lines (29 loc) · 1.55 KB

README.md

File metadata and controls

37 lines (29 loc) · 1.55 KB

A tutorial for using 21 and AWS for deep learning

See the tutorial at https://21.co/learn/deep-learning-aws

Quick deploy

Deploy

Important: You will not be able to quick-deploy without setting up an AWS account and the needed security profiles, etc. See the tutorial at https://21.co/learn/deep-learning-aws for an in-depth description. Here is a short list of the parameters you will need

  • AWS_ACCESS_KEY_ID
  • AWS_SECRET_ACCESS_KEY
  • AWS_DEFAULT_REGION
  • EC2_SSH_KEYPAIR_ID
  • EC2_IAM_INSTANCE_PROFILE_ARN
  • EC2_SECURITY_GROUP_NAME
  • S3_BUCKET_NAME
  • IMGUR_CLIENT_ID
  • IMGUR_CLIENT_SECRET
  • TWO1_USERNAME
  • TWO1_WALLET_MNEMONIC

You can get imgur API keys here.

Finally, if you are going to use GPU instances, you should request a instance limit increase from AWS. Currently they only allow for 2 simultaneously running GPU instances. We have the environment variable EC2_MAX_NUM_INSTANCES to gracefully fail when the limit is reached, and it's set to 1 by default.

If you clone this repository directly (and don't use one of the tagged releases) note that ALLOWED_HOSTS is set to a wildcard to enable the quick-deploy button above. For security purposes you should change this to your specific domain, and note that it's only enforced when DEBUG mode is off.