oak-tree

@alvinwan Thanks for the answer, if you can guide me where are the dependencies on PyTorch it can save lots of time :)

Hey @klueska great answer!! I have one question regaring the Jetson ecosystem. In the x86/64 ecosystem the CUDA comes from the `image` of the`container`. So regardless the version of CUDA...

@Dref360 Hey, thanks for answering. ### total_batches_seen What would happen if `seed` is not none? then every worker will get different kind of seed. Moreover, the `total_batches_seen` is meaningless, isnt?...

@Dref360 > Not really. In keras.utils.OrderedEnqueuer we spawn workers processes using a pool. Each worker has a queue of task and the pool resets every epoch. So for each worker,...

am... important to note that each worker get new copy of the instance every async call. so it wont save the changed attributes

hey @robbmcleod @nouiz Thanks a lot for your quick replies and for throwing some light onto the issue with the differences between `numpy` result and `numexpr` result. Fred, thanks for...

Hey, Another question: Looking at `numpy` doc it seems I do have C order arrays. What different does it make on performance when summing on the transpose matrix? Thanks!

is the `inner product` between x and y where x and y are vectors. A good way I saw online to calculate fast the distance between those vector is to...

is there a public roadmap of TensorRT?

hey there, we are working on a version to do the calculation with webworker. here is a result of 1000 nodes. Its not the best and probably for this kind...