Gridcoin Stats

Everything should be back up and running. Let me know if you're seeing any weird workunits (i.e., really long run times). I'll also have an application update out tomorrow which should help track down some of those issues and improve reliability.

By citizen science grid at 2017-07-25

I'm currently debugging an issue with the assimilator. I hope to have things up and running soon.

By citizen science grid at 2017-07-24

I've made an update to the application which I'm hoping will allow us to get even better results. Currently we've hit 99.45% on the MNIST dataset, which is actually quite good considering the current version of the application only performs convolutional operations and doesn't do any pooling. Here's a list of some of the best published results on this dataset: [Link] http://rodrigob.github.io/are_we_there_yet/build/classification_datasets_results.html#4d4e495354 And that puts is pretty solidly in the middle of the list. The new application now lets EXACT also evolve CNNs with pooling operations, which I'm hoping will bump us even higher up the list. This is pretty exciting as we're evolving the structure of the convolutional neural networks (CNNs) from scratch, and we're getting competitive with some of the best human designed CNNs. Anywho, let me know if you are having any errors with the new v0.31 applications. On another note, I just returned from this years [Link] Genetic and Evolutionary Computation Conference in Berlin, where I presented a poster on this work. There was quite a bit of interest in this approach so that was really good. I just want to thank everyone again for contributing all your cycles to this work.

By citizen science grid at 2017-07-22

I've started up a new set of runs. I made some tweaks to the backend to improve accuracy of how we're storing things in the database (found a couple bugs that while not fatal were not giving me the information I wanted). Workunit names look like: exact_genome_TIMESTAMP_ID_NUMBER And in this case, the ID will be >= 39 and <= 43. Let me know if you're having any errors with these new workunits. On another note, I recently submitted a paper to this years IEEE eScience conference. You can take a look at the preliminary draft here: [Link] http://tdesell.cs.und.edu/papers/2017_escience_exact.pdf Our results are starting to get quite good. I also have an application update that should improve things even further -- but want to get a few more data points with the current implementation before doing the update.

By citizen science grid at 2017-07-12

Sorry for taking a bit long on this, but I've gone through and done a credit reimbursement for the MNIST app which was having the "monster" workunits. I took everyones credit for that app and multiplied it by 1.15 (i.e., a 15% bonus). On another credit related note, I've dropped the credit awarded from future generated workunits by 10%, as the general consensus is that the credit rate is really high. Let me know what you think about this and I can do some more adjustment if it was not enough (or too much).

By citizen science grid at 2017-06-30

So it looks like the poor assimilator is struggling under the load of the larger workunits (a good problem to have), which means work generation has been slow. I've made a few updates which I hope will help get things moving faster and more work flowing.

By citizen science grid at 2017-05-31