Hacker News Reader

Comment by amznthrowaway5 | Hacker News Reader

Chris Lattner left Google a few days ago. Swift for TensorFlow is likely dead.

I'd be surprised if it's dead. At least two of the people on the Swift for Tensorflow team were building a deep learning library in Swift before Swift for Tensorflow was a thing. Much of their work is a part of Swift for Tensorflow.

What about autodiff? Is that a separate effort?

That's good to hear, but S4TF has failed to gain any real traction among ML users and was started over two years ago. I've heard not many care about it apart from the core team and now that Lattner is gone it's likely Google will send it to the graveyard soon.

https://twitter.com/clattner_llvm/status/1222032740897284097

Jeff Dean seems to disagree.

Also, I think it hasn't picked up steam because it just isn't mature enough yet

That's a very non-committed response from Jeff Dean!

I suspect it will drift along like the Java TensorFlow wrapper. Useful in some specific circumstances, but unused outside that.

https://forums.fast.ai/t/swift-for-tensorflow/40167/72

I think Lattner saying that just confirms there are concerns.

The only people who seem to be saying it will continue are people outside Google. Crucially there's no one within Google saying they use it for internal projects.

And it's a good news - Swift is a niche, vendor-locked language. Using something like Rust for production level and something like Julia for R&D stage is a better fit.

Reply on news.ycombinator.com