Re: [MetaOptimize Q+A] New comment on RBM based topic modeling (replicated softmax) implementation

[Timmy Wilson]

Thanks Christian,

I've been chatting w/ Andreas @ Universität Bonn ---
http://www.ais.uni-bonn.de/deep_learning/

They have a RBM implementation that runs on the GPU -- they also have
Annealed Importance Sampling + Partition Function Calculation --
https://github.com/deeplearningais/CUV/tree/master/examples/rbm

I'm excited to see your students implementation, and see if i can get
it working w/ the CUV code

If you're interested in the project, i have a mailing list here --
http://groups.google.com/group/smarttypes

Here's will the code will reside -- https://github.com/smarttypes/SmartTypes

Thanks again Christian,
Timmy Wilson

On Thu, Nov 10, 2011 at 7:05 AM, Christian O. <osen...@gmail.com> wrote:
> Hi Timmy,
>
> he should get in touch with you soon ...
> maybe takes some days, he currently finishes
> his master thesis so is under some work load,
> but I encouraged him to talk to you.
>
> Best,
> Christian
> On Nov 8, 2011, at 6:25 PM, Timmy Wilson wrote:
>
>> Hi Christian,
>>
>>> A student of mine has a python implementation -- more or less
>>> reproducing the paper's result. Drop me a line if you are interested,
>>> I'll arrange the rest.
>>
>> Yes -- i would love to explore/dissemble/use the code
>>
>> If the algo preforms well, Radim may pull it into gensim --
>> http://radimrehurek.com/gensim/
>>
>>
>>
>> ---------- Forwarded message ----------
>> From: MetaOptimize Q+A <q...@metaoptimize.com>
>> Date: Tue, Nov 8, 2011 at 12:01 PM
>> Subject: [MetaOptimize Q+A] New comment on RBM based topic modeling
>> (replicated softmax) implementation
>> To: timmy wilson <tim...@smarttypes.org>
>>
>>
>> ________________________________
>>
>> Hello timmy wilson,
>>
>> osdf has just posted a comment on the question RBM based topic
>> modeling (replicated softmax) implementation
>>
>> A student of mine has a python implementation -- more or less
>> reproducing the paper's result. Drop me a line if you are interested,
>> I'll arrange the rest.
>>
>> Don't forget to come over and cast your vote.
>>
>> Thanks,
>> MetaOptimize Q+A
>>
>> P.S. You can always fine-tune which notifications you receive here.
>>
>> ________________________________
>
>