Skafos Documentation

Smarter Apps Run on Skafos.

  • Development Tools
  • Integration Guide
  • FAQs
  • Frequently Asked Questions

    1. I’m getting a “Module Does Not Exist” error in Xcode after installing the Skafos iOS Framework..

    Make sure you’ve followed the integration guide, using either CocoaPods or Carthage as your dependency manager. After you’ve run the install commands in your terminal, open up a fresh project workspace and clean house (cmd + k, cmd + shift + k).

    Sometimes you get the Module Does Not Exist error if you haven’t done a project build for the first time. Try building and see if the error persists (if you installed the framework correctly you shouldn’t see it anymore).

    2. I cloned an example app and am getting some errors…

    Run through the following series of steps to make sure you have all the necessary dependencies and are in the right place.

    Sometimes you might see errors like this:

    2019-05-20 13:09:23 App Transport Security has blocked a cleartext HTTP (http://) resource load since it is insecure. Temporary exceptions can be configured via your app's Info.plist file.
    2019-05-20 13:09:23 Cannot start load of Task <B05C79B0-571D-4629-BA8D-490013343849>.<1> since it does not conform to ATS policy.
    

    Fortunately it’s an easy fix. Just add this setting to your app’s Info.plist file:

    <key>NSAppTransportSecurity</key>
    <dict>
      <key>NSAllowsArbitraryLoads</key>
      <true/>
    </dict>
    

    Use a text editor like Vim and insert the above snippet within the <plist>...</plist> tags.

    1. Where do I generate an API Token?

    You can generate and revoke API tokens for your account under the Account Settings page of the dashboard.

    2. My model is not updating after I deployed with Skafos…

    There are a couple things to check to make sure you see the model update after a deployment:

    3. Can I use a non-CoreML artifact as my machine learning model?

    Yes. Apple built CoreML to play nicely in Swift and Xcode. Skafos integrates with CoreML seamlessly (see all of our example apps). However, Skafos can be used to deploy any model format to the edge like TensorFlow Lite or DLib.

    Here are a list of example ML apps that demonstrate Skafos integration and delivery with non-CoreML models:

    More coming soon

    4. Does Skafos handle model training?

    Skafos does not handle model training directly. Because each ML app and use-case is unique, Skafos is built to support any training environment you choose. We provide a Python SDK for you to upload models to Skafos from wherever is most convenient for you.


    If you’ve gone through each of these and are still having issues, please email help@skafos.ai and we will provide assistance.