verne.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
< ShikharJ>
zoq: I had read on Yann LeCun's AMA on Reddit, that he doesn't see much potential in sampling based techniques (like RBMs). So many top people making so many different claims, make me wonder whether anybody knows what's happening at all :/
< ShikharJ>
zoq: Sorry to repeat this but can we merge the RBM PR now? I'll push the SSRBM test in the evening?
caiojcarvalho has joined #mlpack
cjlcarvalho has quit [Ping timeout: 260 seconds]
witness_ has joined #mlpack
< zoq>
ShikharJ: No worries, will merge the code in the next hour (as I said on the PR, we should wait 3 days to give anyone else a chance to look over the code).
travis-ci has joined #mlpack
< travis-ci>
manish7294/mlpack#81 (tree - c3d4c9b : Manish): The build is still failing.
< Atharva>
sumedhghaisas: zoq: Currently, the MeanSquaredError is returns average loss over all the batch and also over the datapoint. I think, we will need to do the same with KL loss(currently it's only divided by the n_cols) and as a result I will also need to change it in ReconstructionLoss. Should I go ahead with this?
vivekp has quit [Ping timeout: 264 seconds]
vivekp has joined #mlpack
ImQ009 has joined #mlpack
< ShikharJ>
zoq: Since SSRBM Test requires CIFAR-10 dataset (178MB), I'll push the test as a block comment for now.
Samir has quit [Quit: Leaving]
yaswagner has joined #mlpack
< ShikharJ>
rcurtin: Is there a benchmark system available for testing for the next few days? I wish to run the SSRBM test code and update changes in the PR.
< zoq>
ShikharJ: Sounds reasonable, we can integrate the test later into the models repo.
< zoq>
ShikharJ: About the benchmark system, they are still running, so please feel free to run some experiments.
< zoq>
Atharva: I see, we can also adjust the MeanSquaredError function, your choice.
< ShikharJ>
zoq: Can you delete the current mlpack repo inside Savannah? I made a slight issue, that I'm too lazy to fix, I'd rather rm -r the directory and clone the latest master?
< ShikharJ>
zoq: Also, Kris' RBM PR can be closed now. I've commented on the PR.
< zoq>
ShikharJ: You can't rm -r your own folder?
< ShikharJ>
I'm not one of the sudoers, as it once returned.
< Atharva>
zoq: Oh, if its okay to adjust the MeanSquaredError, that's great!
< zoq>
ShikharJ: Are we talking about: /home/ShikharJ/mlpack?