11 ms·
Tensorflow 2.0 Beta 0
- dannykwells 7y agoI am very happy that Google has realized the importance of usability. Hopefully that comes with concomitant improvements in the tf documentation, which, while thorough is completely unusable and lacks good examples for complex things.
- levesque 7y agoMight give it another try, but my latest incursion in the Tensorflow universe did not end pleasantly. I ended up recoding everything in Pytorch, took me less than a day to do the stuff that took me more than a week in TF. One problem is that there are too many ways to do the same thing in TF and it's hard to transition from one to the other.
- m0zg 7y agoYeah, the only reason to use TF is really its deployment friendliness. If PyTorch addressed that more comprehensively, there'd be no good reason to use TF at all. For research PyTorch blows TF out of the water completely, and it's been that way for years, ever since it came out.
- forgotmyhnacc 7y agoWhat are you looking for in deployment friendliness? There's TorchScript to run your code faster (which is a work in progress)
- jeffshek 7y agoOne of the major benefits of TF 2.0 is apparently the capability to quickly deploy to TPU units with a single parameter change. (I haven't tried it, just followed the marketing). AFAIK, This is still being worked on PyTorch via XLA, but not quite there yet.
- m0zg 7y agoI've found that with TF in general you can only go "quickly" if everything works. If anything is busted you're more or less screwed because it's so opaque. In contrast, PyTorch lets you inspect whatever you want by setting a pdb breakpoint, and when it gives you errors you most of the time can figure out what's wrong without debugging. The importance of this cannot be overstated.
- m0zg 7y agoFor me it's deploying to mobile, mostly. There's ONNX but it doesn't seem to be terribly mature and it doesn't support some of the common ops, and e.g. FB's own Caffe2 doesn't run it natively. There's also no mature tooling to produce quantized models. TF remains the only real option to do quantization aware training or even easy post-training quantization. Specifically, my life would be a lot easier if I could save a mobilenet-style model to e.g. ONNX or some other static graph format that does not require model code in order to load weights. I would like then to be able to load this saved model directly into something on Android and iOS that can use GPU and DSP present on the chip, with minimal extra futzing.
- chadmeister 7y agoSeriously, I've been waiting for a long time now for this to come about. This would make pytorch a much more powerful platform.
- sandGorgon 7y agoWhat do you think of Keras in this space ? Because TF 2.0 is entirely keras based. https://medium.com/tensorflow/standardizing-on-keras-guidance-on-high-level-apis-in-tensorflow-2-0-bad2b04c819a https://medium.com/tensorflow/standardizing-on-keras-guidanc...
- sytelus 7y agoThis is one thing that confuses me. Why Keras is still a separate brand? Why everything isn't under just tensorflow namespace instead of having to do tf.keras all the time. I really wish tf just had one API and just one thing to learn.
- cheez 7y agoKeras is a high level API that can use multiple backends. So it makes sense for them to remain separated.
- p1esk 7y agoI haven't used TF much lately, but the last time I looked at TF2 it felt like they are making it harder to build low level api models.
- cheez 7y agoI doubt it, it's more likely that they are creating higher level abstractions atop the lower level ones and are advertising/documenting the higher level ones
- levesque 7y agoKeras is part of the problem for me. It is rather rigid and it's hard to get around. Works super well for the regular use case. On the other hand, when you want to start doing custom stuff, it's hell.
- NickHoff 7y agoMaybe I'll give TF another try, but right now I'm really liking PyTorch. With TensorFlow I always felt like my models were buried deep in the machine and it was very hard to inspect and change them, and if I wanted to do something non-standard (which for me is most of the time) it was difficult even with Keras. With PyTorch though, I connect things however how I want, write whatever training logic I want, and I feel like my model is right in my hands. It's great for research and proofs-of-concept. Maybe for production too.
- p1esk 7y agoI just hope they don't screw it up in the process of integrating PT with Caffe2.
- strebler 7y agoTF's deprecation velocity was way too high for my taste. Things we wrote would stop working randomly with their updates. I feel very similar to you about the models being "buried too deep" in their (ever-changing) machine. I much preferred how easy it was to hack Caffe V1 (once you got past the funky names, etc). These days, I really like mxnet. Torch was a disaster, but Pytorch is much better. It's not bad in production, definitely my #2.
- panpanna 7y ago> TF's deprecation velocity That's Google on a nutshell. In fact, they may drop TF altogether next month. You never know ...
- chipotle69 7y agoTest
- iamcreasy 7y agoI am curious, what do you like about mxnet?
- panpanna 7y agoCan't explain it but for some reason Tensorflow never felt "right" to me, even work keras. Pytorch on the other hand feels so much more natural...
- flensortow 7y agoAny news on Swift for Tensorflow? I’m skeptical of how much practical benefit it will provide but still willing to take a look at it. There doesn’t seem to be any mention of it here.
- panpanna 7y agoFrom what I understand this is mostly because they hired the Swift guy. I understand the benefits compared to Python (although I would have preferred Go or Kotlin). But what happens when the guy eventually moves on in a year or two?
- favorited 7y agoGoogle's making a big investment in Swift, so if Chris left and they were interested in continuing to support it they shouldn't have a problem. I've gone to Swift on the Server conferences hosted/sponsored by Google, their (non-TF) Swift teams are building some cool Swift tools, etc.
- panpanna 7y agoGoogle is a Go shop, specially when it comes to servers. Swift is not properly supported on linux, which is Google's main platforms.
- favorited 7y agoGoogle is not at all a Go shop. Google is a C++ shop, especially when it comes to servers.
- panpanna 7y ago(that contradicts your earlier comment) Google has been gradually moving away from C++ and Java since 2012. See this quora post with multiple references from Google employees. https://www.quora.com/How-is-Go-used-at-Google-What-could-be-areas-specific-systems-applications-in-which-Go-could-replace-other-languages-used-nowadays-inside-Google-and-why-would-it-make-sense-to-introduce-Go-in-place-of-another-language https://www.quora.com/How-is-Go-used-at-Google-What-could-be...
- nickserv 7y agoNice to see the project moving along, I'm just getting started with the basics for a way finding application and will probably start off with version 2 then. Hopefully by the time stable comes around I'll be near production ready as well. A bit off-topic, but does TF or pyTorch work nicely with AMD GPUs? I'd rather not have to deal with Nvidia's blob drivers if at all possible.
- p1esk 7y agodoes TF or pyTorch work nicely with AMD GPUs? No.
- hn2017 7y agoDoes anyone know when there'll be updated Coursera (or other sites) courses with TF 2.0?
- wodenokoto 7y agoThey have one on udacity