If I was doing AI NPCs, I would crowdsource. Use an AI just search for previous replies before creating a new one to see if one already fits. Then just download and play back the reply that's already been calculated. You can also implement thinking animations, If it replies, going to take a second or two. And possibly small talk to cover up a long inference. Like elevator loading screens. Npc: "Huh? Yeah that's a good question." NPC continues doing what they were doing for a few seconds before answering.
NPC's in games latency should be in the 10's of ms maximum.. maybe 20ms or so? in fact whatever the benchmarked average latency for the game servers users was (I think that is fair).
Another amazing Rip drop on us. This really taught me a lot. This is what I need. ❤
could you please confirm that to access these nim model container , do we required subscription?????
do you know the answer to this by now?
@@diogosa5840 there is a free trial
Can we render using NIM
If I was doing AI NPCs, I would crowdsource.
Use an AI just search for previous replies before creating a new one to see if one already fits. Then just download and play back the reply that's already been calculated.
You can also implement thinking animations, If it replies, going to take a second or two. And possibly small talk to cover up a long inference. Like elevator loading screens.
Npc: "Huh? Yeah that's a good question."
NPC continues doing what they were doing for a few seconds before answering.
what is the cost ?
Useful, thanks!
But I think you need an Enterprise plan from Nvidia. No idea how much that costs.
How you get ai information
it's funny that they name the containers that they give you "influencers"
Make one for Google Gemini please
NPC's in games latency should be in the 10's of ms maximum.. maybe 20ms or so? in fact whatever the benchmarked average latency for the game servers users was (I think that is fair).
Good teacher should not use word "easy", because its not when you just try to learn stuff. It can be easy after.
OMG!!!!!