-
-
Notifications
You must be signed in to change notification settings - Fork 1.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
SpikeSlab RBM tests implementation #1207
Comments
@lozhnikov I had actually started working on some of Kris' PRs, after the discussion on the IRC. I have a branch here. Again it contains only minor conflict resolutions. I'll file a PR soon. |
@ShikharJ ok, feel free to ask questions |
@lozhnikov @ShikharJ Can I work on this issue? |
@lozhnikov I'm looking to work on the Essential Deep learning modules project. Is it okay if I work on this issue? |
Hello, |
@lozhnikov @ShikharJ @Rajiv2605 I am willing to work on the "Essential Deep learning modules" project. How do I get started? |
Actually, this was implemented by @ShikharJ. But if anybody likes to work on additional tests, please feel free. The tests can be found here: https://github.com/mlpack/mlpack/blob/master/src/mlpack/tests/rbm_network_test.cpp. About how to get started, see http://mlpack.org/gsoc.html for a general answer. For the "Essential Deep learning modules" idea, it's helpful to get familiar with the network codebase, run and understand the tests, the models repo https://github.com/mlpack/models might be interesting as well. |
We currently have an additional open PR for SSRBM test implementation at #1477. If someone wishes to take it up, please feel free to do so. |
Hey, I am actually thinking of taking this on as my GSoC project, but I think the implementation is already done. So why is this still listed in GSoC idealist? Does this need reimplementation? |
In summer 2017 @kris-singh wrote an implementation of Binary RBM and Spike and Slab RBM. The first one is implemented according to this article. And the second one is implemented according to this paper.
We reproduced the deeplearning.net MNIST generation test. Here are some samples generated by the binary RBM:
Moreover we tried these models for classification on the Digit dataset.
Unfortunately, we didn't implement extensive tests for the spike and slab RBM implementation. It would be nice to reproduce some tests from the original paper e.g. see section 6.3. "Learning Features for classification".
Ongoing work: Kris's RBM PR, some thoughts on the ssRBM classification test.
The text was updated successfully, but these errors were encountered: