my gsoc project topic

Csaba Hruska

29 Mar 2010 29 Mar '10

1:01 p.m.

Hi! I'm wondering about the topic of my gsoc project. I'm interested in these areas: - improving cabal (more sophisticated dependency, multi version, compiler support, etc) This could be useful from practical view. - Another idea is to improve GHC's new LLVM backend. (eg: support cross module optimization, integrate with llvm-gcc or clang to support full project link time optimization, even using FFI) I'm really keen on compiler technology, and i know llvm for 3 years. (and i'm familiar with clang too) I also have a perspective to GHC's compile stages and intermediate languages. But I've seen that other people are interested in LLVM backend gsoc project too. - Another LLVM related idea is to create a framework to support writing llvm passes in haskell, this should be based on existing llvm haskell binding. - I'm also interested in computer graphics. While this kind of project could look really good, the community could not benefit from it. eg: *write a pure FRP based 3D game, withs existing haskell libs *improve lambdacube 3D engine, etc IMO the big thing will come in these area when DPH and GPGPU will be stable and supported. What do you think about these? Cheers, Csaba Hruska

Attachments:

attachment.html (text/html — 1.5 KB)

Show replies by date

Edward Kmett

29 Mar 29 Mar

1:43 p.m.

On Mon, Mar 29, 2010 at 9:01 AM, Csaba Hruska wrote:

...

I'm wondering about the topic of my gsoc project. [..]

I'm interested in these areas:

...

- improving cabal (more sophisticated dependency, multi version, compiler support, etc) [..]

I like the idea of improving cabal, and one can definitely see the community benefit of work on such a pervasively used tool.

...

- Another idea is to improve GHC's new LLVM backend. [..]

I like this idea, but like you said, a number of people seem to be interested in working on the LLVM backend this summer, so it may be a crowded space and hard to coordinate. But anything that could expose more information about the shape of thunks and slot usage to LLVM could help a great deal with alias analysis, which is usually in LLVM the biggest barrier to reordering and rewriting.

...

- Another LLVM related idea is to create a framework to support writing llvm passes in haskell, this should be based on existing llvm haskell binding. [..]

I think this would be an amazing value-add for the LLVM backend, and might open the door to a whole new class of Haskell optimizations.

...

- I'm also interested in computer graphics. While this kind of project could look really good, the community could not benefit from it. [..]

While I personally enjoy working on 3d graphics in general, it is hard to sell the community benefit. -Edward Kmett

Max Bolingbroke

4:17 p.m.

On 29 March 2010 14:01, Csaba Hruska wrote:

...

Hi! - Another LLVM related idea is to create a framework to support writing llvm passes in haskell, this should be based on existing llvm haskell binding.

This is a cool idea! For bonus points, you could provide a binding to the new version of the Hoopl library that Simon PJ and co are working on. This is a Haskell library that makes it astonishingly easy to write dataflow-based optimisations. It is designed for use with C-- in GHC, but is actually agnostic about the type of graphs it optimises, so you should be able to plug it into LLVM's data types quite easily! For double bonus points, you could then use the LLVM Hoopl backend to write a LLVM pass that fixes up any problems we observe in the output of the LLVM code generator for programs originating from GHC :-) You can get an old version of the paper: http://www.cs.tufts.edu/~nr/pubs/dfopt-abstract.html at not - but note that a Brand New and much improved version is being worked on for ICFP - hopefully the code will be public soon. Cheers, Max

Ketil Malde

30 Mar 30 Mar

7:57 a.m.

New subject: GSoC and Machine learning

Hi, Once upon a time, I proposed a GSoC project for a machine learning library. I still get some email from prospective students about this, whom I discourage as best I can by saying I don't have the time or interest to pursue it, and that chances aren't so great since you guys tend to prefer language-related stuff instead of application-related stuff. But if anybody disagrees with my sentiments and is willing to mentor this, there are some smart students looking for an opportunity. I'd be happy to forward any requests. -k -- If I haven't seen further, it is by standing in the footprints of giants

Grzegorz C

11:55 a.m.

New subject: GSoC and Machine learning

Hi, Ketil Malde-5 wrote:

...

Once upon a time, I proposed a GSoC project for a machine learning library.

I still get some email from prospective students about this, whom I discourage as best I can by saying I don't have the time or interest to pursue it, and that chances aren't so great since you guys tend to prefer language-related stuff instead of application-related stuff.

But if anybody disagrees with my sentiments and is willing to mentor this, there are some smart students looking for an opportunity. I'd be happy to forward any requests.

I don't know whether this is a good idea for a GSoC project, but I would certainly welcome such a library. I am using Haskell a bit for statistical NLP: in my experience currently Haskell is excellent for the components which deal with data preprocessing and feature extraction, but when it comes to implementing the core training algorithms and running them on large data sets, it's easy to get very poor performance and/or unexpected stack overflows. So if a library could provide some well-tuned and tested building blocks for implementing the performance critical parts of machine learning algorithms, it would improve the coding experience in a major way. Best, -- Grzegorz -- View this message in context: http://old.nabble.com/my-gsoc-project-topic-tp28068970p28081419.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

Alp Mestanogullari

31 Mar 31 Mar

10:19 a.m.

New subject: GSoC and Machine learning

Note that, if any student is interested, the Haskell Neural Network library [1] is being rewritten from scratch. We (Thomas Bereknyi and I) are discussing many core data structure alternatives, with some suggestions from Edward Kmett. There may even be some room for a rewrite or update of fgl, possibly with an alternative conception, to fit well HNN. I am definitely not sure if this is worth a GSoC and if the community would benefit that much from such a work, but it's there. [1] http://haskell.org/haskellwiki/HNN On Tue, Mar 30, 2010 at 1:55 PM, Grzegorz C wrote:

...

Hi,

Ketil Malde-5 wrote:

...
Once upon a time, I proposed a GSoC project for a machine learning library.

I still get some email from prospective students about this, whom I discourage as best I can by saying I don't have the time or interest to pursue it, and that chances aren't so great since you guys tend to prefer language-related stuff instead of application-related stuff.

But if anybody disagrees with my sentiments and is willing to mentor this, there are some smart students looking for an opportunity. I'd be happy to forward any requests.

I don't know whether this is a good idea for a GSoC project, but I would certainly welcome such a library. I am using Haskell a bit for statistical NLP: in my experience currently Haskell is excellent for the components which deal with data preprocessing and feature extraction, but when it comes to implementing the core training algorithms and running them on large data sets, it's easy to get very poor performance and/or unexpected stack overflows. So if a library could provide some well-tuned and tested building blocks for implementing the performance critical parts of machine learning algorithms, it would improve the coding experience in a major way.

Best, -- Grzegorz

-- View this message in context: http://old.nabble.com/my-gsoc-project-topic-tp28068970p28081419.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-- Alp Mestanogullari http://alpmestan.wordpress.com/ http://alp.developpez.com/

Mihai Maruseac

10:39 a.m.

New subject: GSoC and Machine learning

On Wed, Mar 31, 2010 at 1:19 PM, Alp Mestanogullari wrote:

...

Note that, if any student is interested, the Haskell Neural Network library [1] is being rewritten from scratch. We (Thomas Bereknyi and I) are discussing many core data structure alternatives, with some suggestions from Edward Kmett. There may even be some room for a rewrite or update of fgl, possibly with an alternative conception, to fit well HNN. I am definitely not sure if this is worth a GSoC and if the community would benefit that much from such a work, but it's there. [1] http://haskell.org/haskellwiki/HNN

Well, I'd like to tie two of my favourite things together. I'm using neural nets here and there (not for very big tasks though, yet) and I intended to use them in haskell too. The code from [0] was intended to become one day useful for a project on neural nets in Haskell. I would be interested in this project if it will be accepted and there would be mentors. [0]: http://pgraycode.wordpress.com/2010/01/25/a-general-network-module/ -- Mihai Maruseac

Alp Mestanogullari

11 a.m.

New subject: GSoC and Machine learning

Well, you can join #hnn or #haskell-soc to discuss that with us. But don't put too much hope on that, I'm quite sure it isn't GSoC worthy. OTOH, any contribution is always welcome heh. On Wed, Mar 31, 2010 at 12:39 PM, Mihai Maruseac wrote:

...

...
Note that, if any student is interested, the Haskell Neural Network

On Wed, Mar 31, 2010 at 1:19 PM, Alp Mestanogullari wrote: library

...
[1] is being rewritten from scratch. We (Thomas Bereknyi and I) are discussing many core data structure alternatives, with some suggestions from Edward Kmett. There may even be some room for a rewrite or update of fgl, possibly with an alternative conception, to fit well HNN. I am definitely not sure if this is worth a GSoC and if the community would benefit that much from such a work, but it's there. [1] http://haskell.org/haskellwiki/HNN

Well, I'd like to tie two of my favourite things together. I'm using neural nets here and there (not for very big tasks though, yet) and I intended to use them in haskell too. The code from [0] was intended to become one day useful for a project on neural nets in Haskell.

I would be interested in this project if it will be accepted and there would be mentors.

[0]: http://pgraycode.wordpress.com/2010/01/25/a-general-network-module/

-- Mihai Maruseac

-- Alp Mestanogullari http://alpmestan.wordpress.com/ http://alp.developpez.com/

5577

Age (days ago)

5579

Last active (days ago)

List overview

Download

7 comments

7 participants

participants (7)

Alp Mestanogullari
Csaba Hruska
Edward Kmett
Grzegorz C
Ketil Malde
Max Bolingbroke
Mihai Maruseac