# Machine Learning

**URL:** https://discourse.julialang.org/c/domain/ml/24.md?page=55

[Latest](https://discourse.julialang.org/latest.md) · [Categories](https://discourse.julialang.org/categories.md) · [Tags](https://discourse.julialang.org/tags.md)

**Page:** 56

---

## [Problem with error reporting in Mocha](https://discourse.julialang.org/t/problem-with-error-reporting-in-mocha/9958)

<div class="topic-metadata">

**Author:** [@Alejandro](https://discourse.julialang.org/u/Alejandro)\
**Replies:** 3\
**Last updated:** [March 26, 2018, 2:46am UTC](https://discourse.julialang.org/t/problem-with-error-reporting-in-mocha/9958 "2018-03-26T02:46:51Z")

</div>

Hello. My name is Alejandro, nice to meet you all. On to business… I am a total Machine Learning novice beyond the theoretical aspect though now it seems I need to get to work in some applications if I am to actually ma…

---

## [\[ANN\] Flux v0.5](https://discourse.julialang.org/t/ann-flux-v0-5/9620)

<div class="topic-metadata">

**Author:** [@MikeInnes](https://discourse.julialang.org/u/MikeInnes)\
**Replies:** 11\
**Last updated:** [March 15, 2018, 3:53pm UTC](https://discourse.julialang.org/t/ann-flux-v0-5/9620 "2018-03-15T15:53:08Z")

</div>

Hey all, I’m pleased to announce the 0.5 release of the Flux machine learning library. Here’s an incomplete list of things that have changed since the last announcement. Experimental JIT compilation work for models, a…

---

## [Problems with Flux](https://discourse.julialang.org/t/problems-with-flux/9719)

<div class="topic-metadata">

**Author:** [@jacobcvt12](https://discourse.julialang.org/u/jacobcvt12)\
**Replies:** 2\
**Last updated:** [March 14, 2018, 8:10pm UTC](https://discourse.julialang.org/t/problems-with-flux/9719 "2018-03-14T20:10:54Z")

</div>

I’m having trouble fitting a really basic neural network in Flux. I wanted to get a handle on the syntax by simulating simple data from a logistic regression, but the model isn’t converging to anything reasonable. I thin…

---

## [How do I do a binary classification with GLMnet, or Lasso?](https://discourse.julialang.org/t/how-do-i-do-a-binary-classification-with-glmnet-or-lasso/8933)

<div class="topic-metadata">

**Author:** [@oxinabox](https://discourse.julialang.org/u/oxinabox)\
**Replies:** 4\
**Last updated:** [February 9, 2018, 3:40pm UTC](https://discourse.julialang.org/t/how-do-i-do-a-binary-classification-with-glmnet-or-lasso/8933 "2018-02-09T15:40:43Z")

</div>

I wanted to include GLMnet and Lasso in http://white.ucc.asn.au/2017/12/18/7-Binary-Classifier-Libraries-in-Julia.html But I outright could not work out how to use them. I feel like they can be used for this, but ma…

---

## [Replacement for os.path.walk in Python?](https://discourse.julialang.org/t/replacement-for-os-path-walk-in-python/8887)

<div class="topic-metadata">

**Author:** [@amarbl](https://discourse.julialang.org/u/amarbl)\
**Replies:** 1\
**Last updated:** [February 7, 2018, 12:00pm UTC](https://discourse.julialang.org/t/replacement-for-os-path-walk-in-python/8887 "2018-02-07T12:00:14Z")

</div>

Hi Folks, I am new to Julia, and beginning to port some machine learning projects over to Julia. One thing I am missing is the python os library which can walk a directory path quite easily. I am googling around and loo…

---

## [On Machine Learning and Programming Languages](https://discourse.julialang.org/t/on-machine-learning-and-programming-languages/7574)

<div class="topic-metadata">

**Author:** [@Mason](https://discourse.julialang.org/u/Mason)\
**Replies:** 48\
**Last updated:** [January 25, 2018, 3:50pm UTC](https://discourse.julialang.org/t/on-machine-learning-and-programming-languages/7574 "2018-01-25T15:50:52Z")

</div>

New blog post on the Julia website

---

## [Equivalent for tf.keras.layers.Lambda in Tensorflow.jl/MXNet.jl](https://discourse.julialang.org/t/equivalent-for-tf-keras-layers-lambda-in-tensorflow-jl-mxnet-jl/8511)

<div class="topic-metadata">

**Author:** [@v-i-s-h](https://discourse.julialang.org/u/v-i-s-h)\
**Replies:** 1\
**Last updated:** [January 24, 2018, 4:55pm UTC](https://discourse.julialang.org/t/equivalent-for-tf-keras-layers-lambda-in-tensorflow-jl-mxnet-jl/8511 "2018-01-24T16:55:44Z")

</div>

Hi, I’m porting some code from Python using Tensorflow to Julia. I’m unable to find an equivalent for custom layer in TensorFlow.jl or in MXNet.jl. This is the exact code I want to translate: from tensorflow.python.ke…

---

## [TeaPot like thing in Julia](https://discourse.julialang.org/t/teapot-like-thing-in-julia/8561)

<div class="topic-metadata">

**Author:** [@Arshpreet\_Singh](https://discourse.julialang.org/u/Arshpreet_Singh)\
**Replies:** 0\
**Last updated:** [January 24, 2018, 11:04am UTC](https://discourse.julialang.org/t/teapot-like-thing-in-julia/8561 "2018-01-24T11:04:22Z")

</div>

I was looking for Teapot like in Julia. Here is Teapot in Python: To summarize what we’ve learned so far about effective machine learning system design, we should: Always tune the hyperparameters for our models Always…

---

## [Simple Knet.jl Question: save/load model?](https://discourse.julialang.org/t/simple-knet-jl-question-save-load-model/8202)

<div class="topic-metadata">

**Author:** [@Joshua\_Bowles](https://discourse.julialang.org/u/Joshua_Bowles)\
**Replies:** 2\
**Last updated:** [January 6, 2018, 7:35pm UTC](https://discourse.julialang.org/t/simple-knet-jl-question-save-load-model/8202 "2018-01-06T19:35:41Z")

</div>

I’m new to julia (as of last week; toyed a bit long time ago with v0.2). I’ve looked over previous Knet posts, searched docs, and Knet (stable) source but can’t find an example of saving/loading a model. I see what look…

---

## [Knet v0.9.0 supports windows, beats new benchmarks](https://discourse.julialang.org/t/knet-v0-9-0-supports-windows-beats-new-benchmarks/8024)

<div class="topic-metadata">

**Author:** [@denizyuret](https://discourse.julialang.org/u/denizyuret)\
**Replies:** 0\
**Last updated:** [December 27, 2017, 5:11pm UTC](https://discourse.julialang.org/t/knet-v0-9-0-supports-windows-beats-new-benchmarks/8024 "2017-12-27T17:11:58Z")

</div>

Knet v0.9.0 released with significant performance improvements, new benchmarks and windows support. Compatibility Windows GPU support implemented. (tested with VS-2015, Cuda 9.1) MacOS GPU support improved: nvml only u…

---

## [TensorFlow.jl on a cluster](https://discourse.julialang.org/t/tensorflow-jl-on-a-cluster/7718)

<div class="topic-metadata">

**Author:** [@Alexander-Barth](https://discourse.julialang.org/u/Alexander-Barth)\
**Replies:** 0\
**Last updated:** [December 12, 2017, 7:45pm UTC](https://discourse.julialang.org/t/tensorflow-jl-on-a-cluster/7718 "2017-12-12T19:45:19Z")

</div>

We have a cluster with 64 nodes and two 8-core processors. Is it possible to use multiple CPUs with TensorFlow.jl using either 16 cores on a single node or multiple nodes ? Are the corresponding functions from the Pyth…

---

## [TensorFlow.jl, set learning rate as Variable](https://discourse.julialang.org/t/tensorflow-jl-set-learning-rate-as-variable/7848)

<div class="topic-metadata">

**Author:** [@davidavdav](https://discourse.julialang.org/u/davidavdav)\
**Replies:** 0\
**Last updated:** [December 19, 2017, 11:39am UTC](https://discourse.julialang.org/t/tensorflow-jl-set-learning-rate-as-variable/7848 "2017-12-19T11:39:46Z")

</div>

Hello, I’d like to control the learning rate of the optimizer in TensorFlow, so in different stages of the training (actually fixing different parts of the network) I want to be able to set different learning rates. Fr…

---

## [Flux 3, now with 100% more Julia!](https://discourse.julialang.org/t/flux-3-now-with-100-more-julia/5824)

<div class="topic-metadata">

**Author:** [@MikeInnes](https://discourse.julialang.org/u/MikeInnes)\
**Replies:** 18\
**Last updated:** [December 18, 2017, 9:54pm UTC](https://discourse.julialang.org/t/flux-3-now-with-100-more-julia/5824 "2017-12-18T21:54:49Z")

</div>

Flux v0.3 is a big release, with a brand new, pure-Julia backend for machine learning. This lets us immediately support all of Julia’s most powerful features, without having to compile them to a more limited backend like…

---

## [Knet.jl CNN Tutorial Speed](https://discourse.julialang.org/t/knet-jl-cnn-tutorial-speed/7500)

<div class="topic-metadata">

**Author:** [@rickhg12hs](https://discourse.julialang.org/u/rickhg12hs)\
**Replies:** 3\
**Last updated:** [December 12, 2017, 1:59pm UTC](https://discourse.julialang.org/t/knet-jl-cnn-tutorial-speed/7500 "2017-12-12T13:59:43Z")

</div>

Knet’s tutorial.ipynb trains a CNN with MNIST data. I have no GPU, but my quad-core I7 has been pegged for a day stuck on cnnmodels = \[ copy(train!(wcnn, dtrn)) for epoch=1:100 \];. Is this exceptionally long training t…

---

## [TensorFlow.jl MNIST Tutorial fails.](https://discourse.julialang.org/t/tensorflow-jl-mnist-tutorial-fails/7476)

<div class="topic-metadata">

**Author:** [@rickhg12hs](https://discourse.julialang.org/u/rickhg12hs)\
**Replies:** 0\
**Last updated:** [December 3, 2017, 2:30pm UTC](https://discourse.julialang.org/t/tensorflow-jl-mnist-tutorial-fails/7476 "2017-12-03T14:30:44Z")

</div>

Ref: Github Issue Report Both MNIST tutorials (softmax & CNN) fail with TensorFlow.jl. I’m a ML noob and I’m not sure how to further diagnose the problem. Is this a Tensorflow.jl problem or perhaps a tutorial problem? …

---

## [Get\_variable with initializer xavier](https://discourse.julialang.org/t/get-variable-with-initializer-xavier/7042)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 1\
**Last updated:** [November 13, 2017, 3:04pm UTC](https://discourse.julialang.org/t/get-variable-with-initializer-xavier/7042 "2017-11-13T15:04:30Z")

</div>

I’m trying to emulate a python call to get\_variable() in Julia: Python: tf.get\_variable(name,size,initializer=tf.contrib.layers.xavier\_initializer(seed = 0)) Julia: get\_variable(size,…?..) but not getting the xavier…

---

## [MXNet installation fails](https://discourse.julialang.org/t/mxnet-installation-fails/6920)

<div class="topic-metadata">

**Author:** [@v-i-s-h](https://discourse.julialang.org/u/v-i-s-h)\
**Replies:** 1\
**Last updated:** [November 6, 2017, 11:33pm UTC](https://discourse.julialang.org/t/mxnet-installation-fails/6920 "2017-11-06T23:33:23Z")

</div>

My attempt to install MXNet fails on Mac OSX 10.13 with the following error. I tried in both Mac OSX 10.13 and Ubuntu 16.04. julia\> versioninfo() Julia Version 0.6.0 Commit 903644385b (2017-06-19 13:05 UTC) Platform Inf…

---

## [Will cassette.jl (new backend for ReverseDiff.jl) obviate need for ML framework?](https://discourse.julialang.org/t/will-cassette-jl-new-backend-for-reversediff-jl-obviate-need-for-ml-framework/6514)

<div class="topic-metadata">

**Author:** [@outlace](https://discourse.julialang.org/u/outlace)\
**Replies:** 3\
**Last updated:** [October 19, 2017, 12:39pm UTC](https://discourse.julialang.org/t/will-cassette-jl-new-backend-for-reversediff-jl-obviate-need-for-ml-framework/6514 "2017-10-19T12:39:49Z")

</div>

I saw the JuliaCon2017 talk by Jarrett Revels about ReverseDiff.jl and how he’s working on cassette.jl to overhaul the backend for ReverseDiff.jl, which should address most (or all) of it’s current limitations. If casse…

---

## [TensorFlow: Basic example fails on "Variable"](https://discourse.julialang.org/t/tensorflow-basic-example-fails-on-variable/6560)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 5\
**Last updated:** [October 20, 2017, 9:23am UTC](https://discourse.julialang.org/t/tensorflow-basic-example-fails-on-variable/6560 "2017-10-20T09:23:42Z")

</div>

When using the basic example from https://malmaud.github.io/tfdocs/basic\_usage/ the only way I can get it to complete without error is by suppressing the y “Variable”. This version works correctly (but the test fails): u…

---

## [Leaky relu with Flux](https://discourse.julialang.org/t/leaky-relu-with-flux/6377)

<div class="topic-metadata">

**Author:** [@skariel](https://discourse.julialang.org/u/skariel)\
**Replies:** 1\
**Last updated:** [October 10, 2017, 1:09pm UTC](https://discourse.julialang.org/t/leaky-relu-with-flux/6377 "2017-10-10T13:09:03Z")

</div>

Any ideas how to implement LeakyRelu, getting a no method matching error on isless. Also I had to implement my own std function: Thanks! mystd(x) = sqrt(sum((x-mean(x)).^2)/(length(x)-1)) lrelu(x) = isless(x,0.0) ? 0.…

---

## [Porting MLKernels.jl to Julia v0.6](https://discourse.julialang.org/t/porting-mlkernels-jl-to-julia-v0-6/6191)

<div class="topic-metadata">

**Author:** [@juliohm](https://discourse.julialang.org/u/juliohm)\
**Replies:** 2\
**Last updated:** [October 2, 2017, 3:12pm UTC](https://discourse.julialang.org/t/porting-mlkernels-jl-to-julia-v0-6/6191 "2017-10-02T15:12:28Z")

</div>

I am finishing a PR for fixing MLKernels.jl in Julia v0.6. I got stuck with the following issue. Suppose I have code of the form: abstract type B{T} end abstract type C{T} \<: B{T} end struct D1{T} \<: B{T} end struct D2…

---

## [TensorFlow: argument evaluation](https://discourse.julialang.org/t/tensorflow-argument-evaluation/6111)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 0\
**Last updated:** [September 27, 2017, 12:35pm UTC](https://discourse.julialang.org/t/tensorflow-argument-evaluation/6111 "2017-09-27T12:35:00Z")

</div>

In following Lyndon White’s example of TensorFlow at JuliaML and TensorFlow Tuitorial I was able to get it to run successfully after two changes: reduce\_sum uses the now deprecated (and refused on my 1.4 dev TF) reduct…

---

## [Knet questions](https://discourse.julialang.org/t/knet-questions/5716)

<div class="topic-metadata">

**Author:** [@Phil\_Tomson](https://discourse.julialang.org/u/Phil_Tomson)\
**Replies:** 2\
**Last updated:** [September 26, 2017, 7:42pm UTC](https://discourse.julialang.org/t/knet-questions/5716 "2017-09-26T19:42:37Z")

</div>

Unfortunately, the knet.jl intro video from the 2017 Julia Conf seems to be missing audio for at least half of the talk… One thing I noticed was that there was a slide that defined a function using an @knet macro: @kne…

---

## [TensorFlow : CPU warnings](https://discourse.julialang.org/t/tensorflow-cpu-warnings/5993)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 4\
**Last updated:** [September 20, 2017, 2:12pm UTC](https://discourse.julialang.org/t/tensorflow-cpu-warnings/5993 "2017-09-20T14:12:14Z")

</div>

Each time I run a TF programme it runs fine but starts out warning me in about 20 lines of messages that I really should be putting my Intel CPU to better use by using the MKL instructions. So I put several hours into tr…

---

## [TensorFlow : explicit close session?](https://discourse.julialang.org/t/tensorflow-explicit-close-session/5952)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 2\
**Last updated:** [September 19, 2017, 1:54am UTC](https://discourse.julialang.org/t/tensorflow-explicit-close-session/5952 "2017-09-19T01:54:50Z")

</div>

The Python API calls for an explicit close of the session according to example code I have seen. The docs for Julia seem to indicate there is no close() on a session. Would this be a Python/Julia difference to add to the…

---

## [TensorFlow : train.restore](https://discourse.julialang.org/t/tensorflow-train-restore/5948)

<div class="topic-metadata">

**Author:** [@Colin\_Beckingham](https://discourse.julialang.org/u/Colin_Beckingham)\
**Replies:** 2\
**Last updated:** [September 18, 2017, 3:33pm UTC](https://discourse.julialang.org/t/tensorflow-train-restore/5948 "2017-09-18T15:33:45Z")

</div>

I’m trying to familiarize with the sample code which runs correctly. There is a line in the epoch iteration train.save(…) which saves a number of binary files labelled from 95 to 99. I understand that they can be recove…

---

## [ANN: XDiff.jl - an expression differentiation package](https://discourse.julialang.org/t/ann-xdiff-jl-an-expression-differentiation-package/1642)

<div class="topic-metadata">

**Author:** [@dfdx](https://discourse.julialang.org/u/dfdx)\
**Replies:** 25\
**Last updated:** [August 22, 2017, 11:02am UTC](https://discourse.julialang.org/t/ann-xdiff-jl-an-expression-differentiation-package/1642 "2017-08-22T11:02:16Z")

</div>

Although it’s not quite stable yet, people seem to be interested in the topic, so I think it’s time to announce a package I’ve been working on lately. XDiff.jl is an expression differentiation package, supporting fully …

---

## [Consumers() in TensorFlow.jl](https://discourse.julialang.org/t/consumers-in-tensorflow-jl/5221)

<div class="topic-metadata">

**Author:** [@Tomas\_Pevny](https://discourse.julialang.org/u/Tomas_Pevny)\
**Replies:** 0\
**Last updated:** [August 4, 2017, 10:44am UTC](https://discourse.julialang.org/t/consumers-in-tensorflow-jl/5221 "2017-08-04T10:44:07Z")

</div>

I would like to implement this method in TensorFlow.jl https://github.com/ProbabilisticNumerics/cabs which should automatically adjust the size of the minibatch in SGD. I am curious about the method, since in security …

---

## [MXNet.jl - weighted classes in SoftmaxOutput](https://discourse.julialang.org/t/mxnet-jl-weighted-classes-in-softmaxoutput/4925)

<div class="topic-metadata">

**Author:** [@dilumaluthge](https://discourse.julialang.org/u/dilumaluthge)\
**Replies:** 0\
**Last updated:** [July 18, 2017, 12:15pm UTC](https://discourse.julialang.org/t/mxnet-jl-weighted-classes-in-softmaxoutput/4925 "2017-07-18T12:15:40Z")

</div>

I’m using MXNet.jl to write a binary classifier for a problem with class imbalance. I’d like to weight the two classes unequally when calculating the loss (i.e. a false positive would be weighted differently than a false…

---

## [Cannot create a random tensor](https://discourse.julialang.org/t/cannot-create-a-random-tensor/4709)

<div class="topic-metadata">

**Author:** [@Tomas\_Pevny](https://discourse.julialang.org/u/Tomas_Pevny)\
**Replies:** 0\
**Last updated:** [July 7, 2017, 6:56am UTC](https://discourse.julialang.org/t/cannot-create-a-random-tensor/4709 "2017-07-07T06:56:46Z")

</div>

Hello all, Does anyone have experience with random number generation in TensorFlow.jl? At the moment, I am failing to create a random\_normal. If I invoke random\_normal(\[5,5\]) I receive following error message: ERROR…

[Previous page](https://discourse.julialang.org/c/domain/ml/24.md?page=54)

[Next page](https://discourse.julialang.org/c/domain/ml/24.md?page=56)