Actualités en direct des projets (non traduites)

Détails

Hello everyone, just wanted to give some updates about the machine learning - python jobs that Toni mentioned earlier in the "Experimental Python tasks (beta) " thread. What are we trying to accomplish? We are trying to train populations of intelligent agents in a distributed computational setting to solve reinforcement learning problems. This idea is inspired in the fact that human societies are knowledgeable as a whole, while individual agents have limited information. Also, every new generation of individuals attempts to expand and refine the knowledge inherited from previous ones, and the most interesting discoveries become part of a corpus of common knowledge. The idea is that small groups of agents will train in GPUgrid machines, and report their discoveries and findings. Information of multiple agents can be put in common and conveyed to new generations of machine learning agents. To the best of our knowledge this is the first time something of this sort is attempted in a GPUGrid-like platform, and has the potential to scale to solve problems unattainable in smaller scale settings. Why most jobs were failing a few weeks ago? It took us some time and testing to make simple agents work, but we managed to solve the problems in the previous weeks. Now, almost all agents train successfully. Why are GPUs being underutilized? and why are CPU used for? In the previous weeks we were running small scale tests, with small neural networks models that occupied little GPU memory. Also, some reinforcement learning environments, especially simple ones like those used in the test, run on CPU. Our idea is to scale to more complex models and environments to exploit the GPU capacity of the grid. More information: We use mainly PyTorch to train our neural networks. We only use Tensorboard because it is convenient for logging. We might remove that dependency in the future.

Source

Création : 17 juin 2021

Détails

Hey everyone,

Due to unforeseen circumstances, we at MilkyWay@home are temporarily deprecating our newest version of Nbody (v1.80) and un-deprecating the previous version (v1.76) in its place. Since the lua files associated with each version are incompatible with each other, we have replaced the previous optimizations with a new set:
-de_nbody_06_16_2021_v176_40k__data__1
-de_nbody_06_16_2021_v176_40k__data__2
-de_nbody_06_16_2021_v176_40k__data__3

We have cancelled all jobs pertaining to the previous set of runs. However, there still may be a few which we were not able to cancel in time. These runs will most likely error out if you get them, but should do so rather quickly (about 2 seconds).

If any complications arise from this, please notify us immediately, and we will quickly find a solution.

Thank you all for your time and continued support,

-Eric Mendelsohn

Source

Création : 16 juin 2021

Détails

It is now possible to set a resource share of 0 in yoyo@home. This was requested by some user.

Source

Création : 17 décembre 2020

Détails

We have now a first GPU version for the M Queens project for Windows. It requires an OpenCL 2.0 GPU. The version can be downloaded from github.com. The zip contains an installation instruction, the app and the app_info.xml.

Source

Création : 15 décembre 2020

Détails

I'm enhancing the ecm search range for B1=43e6 up to 2000 digits. Some test workunits are already send out with the name ecm_es_*test* and a short deadline of 1 day. Together with this the credit formula was changed.

Source

Création : 31 juillet 2020

Détails

Congratulations to user Crashtest, Henric and nenym. Their found prime numbers are in the ECMNET top 10 for year 2020.

Source

Création : 29 mars 2020