Stuck on an issue?
Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
Consider implementing engine abstraction
See original GitHub issueWhen mixing TF and TFE you can end up writing basically the same code twice for models. By having some abstract that can be either TF or TFE that code could be reused.
Smallest change needed is possibly just to finish #105 and then pass either tf or a protocol instance. An extension of that could be to add functions to tfe that get redirected to the current protocol.
Issue Analytics
- State:
- Created 5 years ago
- Comments:10 (10 by maintainers)
Top Results From Across the Web
Graphics API abstraction - Wicked Engine Net
The key to enable this is to use a good graphics abstraction, so these complicated algorithms only need to be written once.
Read more >The Abstraction Principle
The Abstraction Principle is one of the most fundamental principles in engineering. It states: ... The implementation is the engine and transmission. Think...
Read more >Abstraction Layers in Programming: An Overview
The abstraction layer creates a separation between two things. For programming, this is often splitting tasks into separate entities. For ...
Read more >Renderer API Abstraction | Game Engine series - YouTube
Patreon ▻ https://patreon.com/thechernoGitHub repository ▻ https://github.com/TheCherno/HazelInstagram ...
Read more >Abstraction in game engines - GameDev.net
Hi!! In most game engine books, abstraction of the graphics API and other subsystems is one of the main concerns.
Read more >
Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free
Top Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
@jvmancuso suggest we keep it around to not forget; the issue discussed here isn’t mentioned in the other issues afaict
Less for a user to learn as they could just take their pre-existing knowledge of tf, then pass the built graph into tfe and get back a converted graph. We’d still offer all of the underlying API we have today but this could be the first step for anyone who just wants to get running with “encrypted” machine learning.