Search found 156 matches

by zenke
Tue Jan 22, 2019 9:11 pm
Forum: Examples
Topic: sim_background; homeostatic LTD term
Replies: 1
Views: 7858

Re: sim_background; homeostatic LTD term

Hmm. So kappa should definitely be in the denominator. Is your question that all the units cancel leaving A^- unit-less and you would want it to have units to work in Eqn. (17)?
by zenke
Mon Oct 29, 2018 2:50 pm
Forum: Development
Topic: SuperSpikeConnection in parallel
Replies: 1
Views: 4310

Re: SuperSpikeConnection in parallel

Thanks and thanks for reporting this. Time has indeed been a constraint lately. The answer to your question depends a bit on which connection class you are after. The random feedback connection should in principle support parallelism, so if you are seeing a problem there, this might be due to a bug....
by zenke
Tue Sep 18, 2018 4:04 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Just committed a new version to the develop branch which now uses MPI_Allgatherv to sync the different processes with variable sizes. Would be interesting if this version of SyncBuffer works with your code. Not sure it will save us a lot, but potential savings are high if different ranks produce vas...
by zenke
Sat Sep 01, 2018 11:08 am
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Hi Anders, I just merged of BcpnnConnection to branch dev-bcpnn with some change suggestions. I used sim_brunel2k_bcpnn to test that my version and yours produce the same output. I did not merge several modifications you did to the Monitor class and other files because I wasn't sure about them. I le...
by zenke
Fri Aug 31, 2018 8:40 am
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

But for our networks we need to have pairwise delays between minicolumns and I think this cannot be achieved by means of these mechanisms you mention, or ...? If I understand you correctly, then you want to be able to set a different delay for each synapse? You are right, this is currently not poss...
by zenke
Thu Aug 23, 2018 9:32 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Hi, The comparisons look indeed very good. Fingers crossed it will scale well! Delays: Individual delays are something that is still a bit underdeveloped in Auryn because I never played with this in my simulations. Out of the box Auryn supports "axonal" delays on a per SpikingGroup/NeuronGroup basis...
by zenke
Tue Aug 21, 2018 1:42 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Hi Anders, good to hear this is moving forward. If I understand you correctly you are currently facing two problems: i) The p* traces do not return values which resemble spiking probabilities and ii) the weight monitor does not return the same values that you print from the evolve() method. Concerni...
by zenke
Tue Aug 21, 2018 1:22 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Hi Anders, Using your patch and some new code, I just posted a refactored version of SyncBuffer which hopefully fixes the problem you encountered. Still need to do some more testing with it before merging it in the active develop branch The code can be found here https://github.com/fzenke/auryn/tree...
by zenke
Fri Aug 17, 2018 3:16 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Hi Anders, This made my day! Thanks for posting this. Looking very good. Hope you get similarly good scaling with BCPNN enabled. Did you have the chance to check on the activity in these massively parallel simulations? I guess for the performance measurement you turned off spike output. If you actua...
by zenke
Tue Aug 14, 2018 7:31 pm
Forum: Development
Topic: Computational basis for BCPNN learning rule
Replies: 24
Views: 29885

Re: Computational basis for BCPNN learning rule

Thanks, Anders! I got your pull request. I will now review the code, run some tests and then merge it manually. This is really exciting stuff. I had never run a simulation with Auryn on more than 128 cores, but I am happy that after some tinkering this seems to work :D Thanks for helping with improv...