ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
vivekp has quit [Ping timeout: 250 seconds]
vivekp has joined #mlpack
< davida>
zoq: I take it that this means it is more challenging than first expected. That is a pity as I was hoping to complete the RNN and LSTM exercises of the DeepLearning.ai courses and move on to some personal projects (which unfortunately are also RNNs needing different sized sample inputs).
< zoq>
davida: THe solution I have is pretty much useable if you manually construct the dataset (arma::field).
< davida>
zoq: what is the format required for the field input?
< davida>
zoq: Does the field option also apply for the targets? (e.g. using an RNN to learn translation would require samples of different lengths of English text with targets of different lengths of translated text - lengths of input and output not being the same as well as every sample being different length)
< rcurtin>
I think I am not as creative as Conrad, but that's the best that came to mind today :)
< rcurtin>
just for fun, I posted ensmallen.org to hackernews, and it appears to have made the front page for the past few hours and has a handful of comments
< rcurtin>
surprised it actually got to the front page... my last several attempts to post something haven't gone anywhere
< rcurtin>
zoq: I noticed that this morning too. it seems you have to log in to see it
< zoq>
ah, okay
< zoq>
that works
< rcurtin>
I thought that I had the jobs set up so they could be anonymously viewed, let's see...
< rcurtin>
ah, I looked into it for a little bit but I don't see an easy way to do it. I think the workaround of logging in is fine for now unless you want to dig further