Tim Zaman
Tim Zaman
Just checked. Using the GUI it is fine and gets ellrl into 600kb On Mon, Mar 18, 2019, 21:57 Nostrademous wrote: > You would think they have rotating console file...
Doesn't need to be the exact same hero.. would be awesome if that could be done somehow though. Could we, dependent on the side dire/radiant, expose Think function?
 currently occupied with self-play induced strategy collapse (bots getting both very hostile)
Should we make things like 'get_total_xp' and 'xp needed to reach level' part of the dotaservice? These seem super generic. Things like reward functions are really specific to users models....
On GCP you get $300 for free, which is equal to around 40 cores for a month. (20 agents). But you don't really need it. You can just install k8s...
Oh btw, nice that you got it working! Yeah what you posted is exact how it should work! Nice! And it will take a few hours for it to go...
Here is a model with the latest dota trained last night on top-of-tree (0.3.4) [your code!] [exp1_job2_model_000001576.pt.zip](https://github.com/TimZaman/dotaclient/files/2769388/exp1_job2_model_000001576.pt.zip)
yep lets do a addon. looks dead-easy. 
Wondering if there are applications here for in-memory sharded _datasets_, a use-case not mentioned above. Example: such that you could randomly access into a massive XTB dataset from any node...