Giters

affinelayer / pix2pix-tensorflow

Tensorflow port of Image-to-Image Translation with Conditional Adversarial Nets https://phillipi.github.io/pix2pix/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Best way to deploy pix2pix in production.

opened this issue · comments

commented

I successfully trained a deep net using pix2pix. The net is able to modify the content of a square 512x512 px image.

Now I would like to deploy the net in a server and create a basic cloud service where a user (client) can send a picture and the server respond with a modified picture.

Googling around I noticed that the most used way to do this are:

  • AWS Lambda
  • Serverless
  • Tensorflow Serving

The problem with Lambda, is that the model data is very big: 700 MB, and seems to exceed the Lambda limitation (https://docs.aws.amazon.com/lambda/latest/dg/limits.html).

While Tensorflow Serving seems too complicated for me and all the tutorial are for Classification or Recogntion purpose, and not for Image to Image Translation.

What do you suggest?

EDIT: I found also the deeplarnjs.org implementation ready in this repo for pix2pix. But this run the generator on the client browser. Correctly?