eaplatanios/tensorflow_scala

jhegedus42

@jhegedus42

i mean any of the tensorflow objects?

which are python/c++ i assume, how is the interfacing with tf happening?

Anthony Platanios

@eaplatanios

@biboudis you should be able to use tf.whileLoop. The package is private to force users into using the tf namespace for all ops. Not sure if this is the best approach but it's parallel to the Python API of TensorFlow and I was hoping it would allow for a smoother transition.

@/all I've been putting TF Scala aside for a while now because I had to prioritize other projects. I've also been planning a big re-design that would allow for eager execution and better support for AutoDiff. I hope to soon post updates here regarding these plans.

@jhegedus42 TF Scala communicates with TF through its C API (and JNI). Only references to TF objects are stored in the JVM. This causes some issues with respect to garbage collection (i.e., the GC is not aware of how much memory some of these objects actually occupy and so memory pressure is not computed correctly which can lead to some issues). I haven't found a nice workaround to this issue yet.

jxtps

@jxtps

Looking forward to hearing more @eaplatanios !

jhegedus42

@jhegedus42

@eaplatanios interesting ...

how about scala native ?

http://www.scala-native.org/en/v0.3.9-docs/

http://www.scala-native.org/en/v0.3.9-docs/user/sbt.html#garbage-collectors

http://www.scala-native.org/en/v0.3.9-docs/user/interop.html

AlgorithmDog

@algorithmdog

Hi, I am using the LinearRegression example from https://github.com/eaplatanios/tensorflow_scala/blob/master/modules/examples/src/main/scala/org/platanios/tensorflow/examples/LinearRegression.scala

And I want to make the example to distributed train. Could anyone know how to use distributed training in TF4S

cluster-ai

@ai_cluster_twitter

Hey guys

Are there any docs that explain how to import a custom operation into Scala and then export the graph?

Trying to export the SentencePiece Model and load it in Scala, but I get some missing OP errors. Anybody got some advice? Strongly appreciated

Op type not registered 'SentencepieceEncodeDense' in binary running on loan-W540. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) `tf.contrib.resampler` should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed.

cluster-ai

@ai_cluster_twitter

I figured it out its
TensorFlow.loadLibrary(customOperationsSoFilePath)

Where customOperationsSoFilePath is the path to some Operations.so file

Anthony Platanios

@eaplatanios

@ai_cluster_twitter that's correct. Sorry for not responding earlier! Keep in mind that I haven't updated the library in a while and haven't verified all works fine with TF 2.x. I hope to get a chance to look into this soon.

Izmar Verhage

@izzyreal

Does anyone know how to call Image.decodePng against an Array[Byte]? I'd like to load an image from memory rather than from file.

Izmar Verhage

@izzyreal

With TF Python it's a breeze, just file.read() and pass the result to convert_to_tensorand that to decode_png.

But any combination I tried in Scala doesn't give expected results. The best I got was

Expected image (JPEG, PNG, or GIF), got unknown format starting with '\302\211PNG\r\n\032\n\000\000\000\rIHD'
     [[{{node DecodePng}}]]

and

got unknown format starting with '?PNG\r\n\032\n\000\000\000\rIHDR'

Izmar Verhage

@izzyreal

This is the code that gives the last result

    ImageIO.write(bufferedImage, "png", outputStream)
    val data = Output.constant(new String(outputStream.toByteArray))
    val image = Image.decodePng(data, 3)

James McCabe

@oranda

Wonderful project @eaplatanios. One question. Running MNIST example ends in accuracy of only 0.942/0.931. Is that expected? v0.4.2

Corey O'Connor

@coreyoconnor

Having a great time using this library for numerical processing: https://www.youtube.com/watch?v=lDPT55pmrzc

One question tho: How do I pronounce the author's name Platanios? I trip over it ever time :\

James McCabe

@oranda

@xtordoir pronounces it, though I don't know if he is correct: https://www.youtube.com/watch?v=83iprWIhjMM

Corey O'Connor

@coreyoconnor

@oranda thanks!

James McCabe

@oranda

May I ask the community a general question? How much are the giants (Google, Facebook, etc) using Scala for ML at the current time? Maybe behind the scenes? It's hard to believe that with such large codebases they are still 100% on Python. If they are, do you think the next couple of years could see a breakthrough? Any anecdotal reports of Scala with ML?

Anthony Platanios

@eaplatanios

@coreyoconnor glad you've been enjoying the library! Regarding my name haha many people have trouble pronouncing it, but @xtordoir does a good job. :)

@oranda it depends a lot on the company, but most of them do not use Scala much for ML. Oftentimes models are developed using python and then deployed using TF, Torch, etc, in C++, Java, ... There are also many companies that use Python which is also surprising to me. However, there are a few places that also use Scala for ML, including Microsoft (where I work). I'm also aware of Netflix doing some ML stuff using Scala (@jeremyrsmith may be able to say more) and Twitter. AFAIK it's not used much in Google.

Mandar Chandorkar

@mandar2812

@oranda I implemented my entire PhD research in Scala, loved using Scala for ML! I work in industry now and one thing I learned is that Scala has a very bad reputation. My CTO told me that he feels that one needs a PhD to understand Scala code. The Scala community and its emphasis on a lot of the advanced syntactic features of the language have not helped one bit.

Sometimes its better to hide those capabilities and emphasize how easy it is to get started with a language if we want to attract new talent. But the community has a bias for Monads/DSLs/Type Theory etc.

All of those advanced capabilities and extended freedoms mean that I can probably write maintainable yet simple (whatever that means) Scala code, but if I take a peek into someone else's library, I might be out of my depth very very quickly and even after years of experience I might still have to invest hours/days to understand those advanced design patterns.

So I don't see many companies adopting Scala for ML, and I am not surprised.
Sometimes excess freedoms are not a good thing.

James McCabe

@oranda

@eaplatanios and @mandar2812 thanks for your replies. I have some code that uses tensorflow_scala at https://github.com/oranda/ScalaTensorflowPOC/blob/master/src/main/scala/com/oranda/tensorflow/MNISTSimple.scala. I have tried to keep the code as simple as possible, avoiding things like implicits and advanced Scala features, and a Python dev should be able to understand it. I would love to get your opinions on this style, and how this code can be improved.

Corey O'Connor

@coreyoconnor

@coreyoconnor glad you've been enjoying the library! Regarding my name haha many people have trouble pronouncing it, but @xtordoir does a good job. :)

thanks!

I've put out the second video in a series on tensorflow for scala: https://www.youtube.com/watch?v=NwZ5WviQJVQ

Yiğit Ünlü

@argestes

Hey ppl!

I'm trying to use the library with Scala 2.13

So far I've failed

Corey O'Connor

@coreyoconnor

What version are you using?

IIRC you will need to use a SNAPSHOT version

Anthony Platanios

@eaplatanios

Yes I would suggest 0.5.2-SNAPSHOT. I haven’t published a release yet because I added support for Windows and I’m currently testing that.

danoliv

@danoliv

Hello everybody, I work as an engineer in a data science team, we use 100% scala in production at the moment. We use mostly spark ml, xgboost. We plan to deploy some deeplearning models in production, is tensorflow_scala production ready ? I've used it successfully with tensorflow 1.0 for some personal project but we are planing to use tensorflow 2.2. Thanks

Corey O'Connor

@coreyoconnor

I've put out the second video in a series on tensorflow for scala: https://www.youtube.com/watch?v=NwZ5WviQJVQ

Third video in the series: https://www.youtube.com/watch?v=WpqH9aqR3Mo

Tho I don't actually talk about tensorflow but ZIO library design shrug

Anthony Platanios

@eaplatanios

@danoliv 0.5.2-SNAPSHOT is built for TF 2.2 but still relies on building graphs and does not yet support model training in eager mode. Regarding stability, we’re in the process of evaluating whether we can use it in production. No issues so far, but I’ll let you know if something pops up. In either case, I’m glad to hear you’re considering it :)

@coreyoconnor thanks for the nice videos you’re making! I’ll take a look at this one as I’m not very familiar with ZIO.

danoliv

@danoliv

@eaplatanios thanks for your answer, I am looking forward to test it. Is it available on some public repo ?

Raghav Bharadwaj

@raghavbharadwaj

Hello!! Are 0.5.x-SNAPSHOT releases available with prebuilt binaries? I can't seem to locate them. Thanks for the help!

Where communities thrive

People

Repo info

Activity