4 ms·
TF 2.0 in general (SavedModel included) has been a giant clusterfuck so far. As we used to say at Google "the old thing is deprecated, the new thing is not read
by m0zg 7y ago
TF 2.0 in general (SavedModel included) has been a giant clusterfuck so far. As we used to say at Google "the old thing is deprecated, the new thing is not ready yet" - internally Google is pretty much entirely like this. I get that they had to ship it sooner or later, but IMO they shipped 2.0 a bit too soon. Shipping sooner is fine if you fix the bugs quickly, but that's not really happening either - some rather potent bugs remain open for half a year or more.
So at the moment you have a choice of working with a version that's officially EOL (1.15) or using a version that has several critical features broken or missing, and on which TF Model Zoo doesn't work. Oh, and EOL version is also busted beyond belief. Worse yet, Google does not reward maintenance work, and since this is considered "launched", you can count on the bugs being there for an extended period of time until Google itself moves to 2.0, which I'd wager is unlikely to happen anytime soon. Facepalm.
- missosoup 7y agoSo basically everyone should move to pytorch and ONNX?
- amelius 7y agoThe best thing would be if everybody converged to the same framework.
- m0zg 7y agoONNX is a turd in is present state as well. Too many unsupported ops and you can't really run ONNX on e.g. a phone or really anywhere. You could theoretically convert it into something that does run where you need it to run, but then you have to deal with the converter's supported ops as well, and design your net around the intersection of the two sets of ops and their parameters. If you're doing anything more complicated than basic classifiers, you're going to run into issues. That said, I think in a few years with enough elbow grease ONNX could become pretty exciting. As to what people should use for research: just use PyTorch. I switched to it when it first came out, and never looked back, _except_ when I need my stuff to run on e.g. a phone or something, which is something PyTorch won't help you with.
- pjmlp 7y agoThe best part is the whole Tensorflow-Swift, with the halfbacked support for anything that isn't an Apple device. Meanwhile Julia works just fine on Windows.
- m0zg 7y agoProblem with Julia is that approximately nobody wants to learn yet another language just to do pretty much the same things that Python can do with the help of e.g. PyTorch or NumPy. Us programmers often overestimate the level of enthusiasm non-programmers have for learning programming languages. Heck, I'm kind of also running out of enthusiasm in this area myself.
- pjmlp 7y agoStill better to learning a language that only works properly on Apple platforms, without any kind of eco-system for scientific computing. I don't see Tensorflow-Swift going anywhere.
- m0zg 7y agoI agree. I looked into it briefly (since I sorta already know Swift anyway and like the language) but I can't even run this stuff on Linux easily, so it's kind of useless for me so far. This doesn't in any way mean that Julia has a snowball's chance in hell of taking off though. Besides the language, Python had a decade of head start in terms of "batteries included". And people have gotten used to that particular brand of batteries.
- pjmlp 7y agoJulia will add pressure for Python folks to actually take PyPy seriously. There is a large subcommunity that doesn't enjoy being forced to drop down to C for anything performance. Julia already counts some banks on their list, that tend to use ML languages instead of Python due to performance.
- 7y ago