News for hackers. Stuff that matters.
Re:Learning Dexterity by 0x8BADF00D on Monday July 30, 2018 @12:53PM (#17645888)
> Rapid used 6144 CPU cores and 8 GPUs to train our policy, collecting about one hundred years of experience in 50 hours.

That seemed an order of magnitude higher than I expected. Is training usually this computationally expensive?

Re:Learning Dexterity by boulos on Monday July 30, 2018 @04:13PM (#17647587)
Disclosure: I work on Google Cloud.

Is it the cores that throws you off? (Those are used for the simulation, not the training). Second, I believe those were preemptible cores, so that's $60/hr for the cores and then $20/hr for V100s (which is what I think they used). $80/hr isn't bad considering how much a (small!) team of researchers costs.

Re:Learning Dexterity by GlenTheMachine on Monday July 30, 2018 @01:01PM (#17645979)
The notable thing about this is how little computation it needed. We have a long way to go, but this is a big step in the right direction.
Re:Learning Dexterity by LeanderK on Monday July 30, 2018 @01:04PM (#17646001)
haha I just read it and thought that's a magnitude less than I expected. Pretty often it is. A lot of papers from high profile institutions have a lot of computing power availiable.

First, it seems to be a lot and be really expensive, but think of it in man-hours. It quickly diminishes.

Re:Learning Dexterity by visarga on Monday July 30, 2018 @01:33PM (#17646280)
Not to mention that evolution has had millions of years of optimizing this stuff.
Re:Learning Dexterity by cyberpunk0 on Monday July 30, 2018 @01:22PM (#17646174)
Not always but a lot of machine learning techniques now are effectively just brute forcing and highly expensive band wasteful to compute. Machine learning is getting close but I don't believe we will get there with our current methods
Re:Learning Dexterity by sgillen on Monday July 30, 2018 @02:06PM (#17646587)
Well if you look at the plot in the "Learning progress" section, you'll see that they did require almost two orders of magnitude more training time due to the randomizations they were adding to the simulation. Without these the policy isn't very robust but also takes a lot less time to train.
Re:Learning Dexterity by sigi45 on Monday July 30, 2018 @02:14PM (#17646643)
Yeah but it doesn't matter; One model to rule them all.