Jacopo,
Ryan Perry
12

Hi Ryan,

thanks for taking the time to read the article!

Short answer: yes, it happened to me before… things to check:

  1. tensorflow version you are using, as it changes so quickly that it can be within AWS limits (or not) for a very tiny amount of disk space.
  2. unnecessary folders you can delete

Long(er) answer: I would suggest starting from my pre-compiled package, which you’ll find available in the article. Try and see if everything works as expected (it should).

If there are no problems with my bundle, it’s time to optimize yours, which could be a useful exercise anyway. When happened to me, I followed general rules (like the ones here, for example: https://medium.com/@maebert/machine-learning-on-aws-lambda-5dc57127aee1#.hj8rkzrmh) to save some space, remove example folders and unused TF folders and then upload my zip.

Let me know how it goes!

Show your support

Clapping shows how much you appreciated Jacopo Tagliabue’s story.