mmq

Hi, I am the author of Polyaxon, a bit late to notice, but thanks for sharing Polyaxon here. I will be around to answer questions, and any feedback is welcome.

Thank you for the great feedback, a lot of things are indeed planned to enhance the dashboard as well as the cli in terms of search, filtering, and ordering of the experiments based on some rule, e.g. parameters or metrics.

Also, some of your, and other users', feedback were also very helpful to bring some changes to the infrastructure for the next release, to maximize the usage of the cluster's resources.

You can use scikit-learn, or any other library, to train your models.

Polyaxon, https://github.com/polyaxon/polyaxon, is an alternative project to MLFlow that is also open source.

disclaimer: I am the author of Polyaxon.

@ah- thanks for mentioning Polyaxon and congrats to the CometMl team for building this nice tool, it's good to see that many projects are trying to solve problems related to reproducibility in ML/DL, many people had to build an internal tool for the companies they work for to solve this issue, and many got frustrated after joining a new team and were not able to reproduce any results.

I would like to outline a couple of differences between CometML and Polyaxon, as mentioned before, we are also trying to solve issues related to technical debt in ML, but not only, Polyaxon tries also to simplify training and scheduling parallel and distributed learning. there are also a couple of differences, I see CometML as dashboard, Polyaxon does not have an extensive dashboard as CometML, but it leverages Tensorboard for most of the visualisations. We use the CLI or the API for programatic access to the platform. Most importantly, Polyaxon aims to be an open source and to be installed on premise or in the cloud, it solves the issue related to code tracking based on an internal git and a docker registry, and as someone else mentioned that resources for running an experiment could be an issue for future reproducibility, Polyaxon restarts the experiments with the same resources and dockerfiles, it also tracks hyper params as part of the configuration.

For hyper params tuning and suggestion, Polyaxon can also do hyper params search based on a couple of algorithms, and for the next release, it will include also a service similar to vizier for suggesting more experiments/group of experiments based on a given search space.

Disclaimer: I am the author of Polyaxon

It's a great product, I stopped using google translate when I discovered this one. The translation is much better, more accurate IMO.

I think it's a maintained by google, this is one of the main contributors of skflow https://github.com/ilblackdragon.

Also the repo is under tensorflow organisation.

How is tflearn different than skflow[1] or prettytensor[2], both these modules are developed by google's teams.

[1] https://github.com/tensorflow/skflow/

[2] https://github.com/google/prettytensor

I think that being alone and loneliness is two different things. One can be alone without feeling loneliness, the same way, one can feel loneliness while being surrounded by people [other people mentioned the quality and health of relationships]. When one feels that the Sisyphean task is easier because other people go trough it is comforting. But I think what would make this experience easier is, being aware that it is more absurd to end it. One may put oneself in the crowd to feel the warmth, or one can worm one’s mind and stand for what one believes.

If you are a Python developer, have a look at this video http://pyvideo.org/video/1780/transforming-code-into-beautif...

Also, you can check the Google Python Style Guide:

http://code.google.com/p/soc/wiki/PythonStyleGuide

http://google-styleguide.googlecode.com/svn/trunk/pyguide.ht...

Any idea about Canada? Vancouver seems very interesting tech hub.

http://bit.ly/11rxC84 Zocdoc clone written in Python/Django, angular.js, moment.js.

Works fine for me!

Probably, because he said that he doesn't want to go into his family details. He wants to keep his identity secret. Also the fact that he didn't tell his company about his illness would be a justification.

Sounds good. Although, my experience with service isn't very happy. I guess probably because I am trying to use twilio outside the US. In fact, I sent two mails to the sales team and no answer. And, I just opened a ticket an hour ago so they can help me verify my no US number. I am still using the free plan to integrate their service with my web app.

Some links I found helpful while trying to make a sudoku game with angularjs (https://github.com/mouradmourafiq/angular_sudoku):

http://deansofer.com/posts/view/14/AngularJs-Tips-and-Tricks...

http://www.yearofmoo.com/2012/10/more-angularjs-magic-to-sup...

http://blog.petermolgaard.com/tag/angularjs/

http://onehungrymind.com/angularjs-sticky-notes-pt-1-archite...

http://www.raymondcamden.com/index.cfm/2011/11/29/My-first-A...

A google discussions about the difference between factory and service:

https://groups.google.com/forum/?fromgroups=#!searchin/angul...

"That which does not kill us makes us stronger." -Friedrich Nietzsche

How is your design affected?

For the case when the user isn't willing to give access to his/her twitter account.

The error returned is KeyError, since you are trying to access :

user = User.objects.get(username = authorized_tokens['screen_name'])

I see that you have put a try, except blocks. But you are not catching the KeyError exception. You are only making an exception when the user doesn't exist.

you should add a block : except KeyError. And figure out the logic for this particular case. (probably just redirecting the user to the home page)

You should probably set DEBUG at False. This is very undesirable in a production environment as it will probably leak sensitive information that attackers can use against your site. Unless, of course, if you use only the Django ORM.

I don't think that memorizing is the perfect approach to learn mathematical proofs. Generally, you should be able to try the proof based on your understanding of the concepts, and hence if you can't proof them, it only means that you didn't get the concept well. When I used to be in the CPGE [1], we had one exam after two years intensive courses of maths, physics philosophy and engineering sciences. And we had pretty much everything demonstrated. The only way we, other students and I, dealt with it is by understanding the concept very well, rereading the courses and going through problems. It's been 6 years now that I left CPGE, And I am still able to proof the Bolzano-weirstrass theorem, and I am sure that most of the other students could also.

[1]I don't know if you are familiar with the french educational system, and specifically the CPGE( classes préparatoires aux grandes écoles). It consists of two very intensive years of mathematics, physics, philosophy and engineering science.

plangr.com

Too many websites to visit each time, I read from this small toy feed reader : http://bit.ly/10SNaFJ

Seeking Work - Remote/Onsite ---------- Mourad Mourafiq

Email : - mouradmourafiq [at] gmail

----------

Web Developer, RESTful API integration and development, also interested by data analysis.

I have experience with current technologies:

-Python

  * numpy
  * scipy
  * pandas
  * django

-Hadoop

-javascript

  * jquery
  * angularjs (learning)
  * momentjs

-Posgresql

-mysql

-sphinx search

-elasticsearch

-twitter bootstrap

----------

Github - http://www.github.com/mouradmourafiq

I hope that these guys won't pay attention to your offensive comment. Your experience or your personal taste shouldn't make you, in anyway, say something like that. I think that your first project wasn't perfect as well.

Also, if you happen to notice something that you don't like or you don't agree with, try to make more constructive comments, so that people can respond or enhance.

To my humble understanding of angularjs, the MVC patter is not exactly the one you would understand coming from a JEE background. The fact that angularjs use this pattern means that you should first make your server as a RESTful API, and build everything else on the client side using the framework.

In this way the backend is responsible for deep validations and making the data as json format for later use.

So in the end :

  *Model : data comes from the server json format, becomes jaavscript objects.

  *View: HTML, and directives when you need to manipulate the DOM

  *Controller (also your services):

      ->Queries to the API
      ->Validation (probably duplicate some of your server's side)
      ->Business logic.
      ->Rooting of templates.

You might think that there's a duplication, in terms of controls and validation, but the thing is, you should think of it as two separate level : One for your API (could be used for many purposes), Second a robust, fast client side that deals with data.

-Python

  * numpy
  * scipy
  * pandas
  * django

-Hadoop

-javascript

  * jquery
  * angularjs (learning)
  * momentjs

-Posgresql

-twitter bootstrap

Actually, I was just curious about other people's experience, especially in the worst case scenario (start up fails). Since I quit my day job to free my mind totally for entrepreneurial experience.

As for me I had a side project, geo-social planning tool, that had 5000 users, but not actively using the app. I shut it down, because it was unlikely to succeed, it was strange to the way people meet and go out. And also, because I noticed that other similar projects were also stuck in the chicken and egg problem.

As you made all you can do, pivoted and especially burnt all the money you had.