thanks, and what you will do if you want deterministic engine? that will stoped the probability from treining. Is better to change directly CUDA math? I have some idea and last days I thinking about what to try if you have metacontrol over inference based on thermodynamic for nonlinear systems. It is experimental system3 that is working on big models and I try to build one at home. But may be HDD is not good GPU and CPU RTX 4080 just 12GB. thanks
2026-09-10 01:11:01
0
dionpanag :
thanks for the info, very impirtsnt
2026-09-10 04:34:06
1
Peace :
Your videos are very easy to understand.. please keep making like this…
2026-09-11 06:17:24
2
Kiznavierr :
thanks dude, your video is very easy to understand
2026-09-09 23:35:16
2
Property Jinn :
why do we have to train it , how to make it understand from the token and answer , rather than having training ?
2026-09-11 04:23:08
0
noodler83 :
Excellent content. Thanks!
2026-09-11 03:03:47
2
SHAH JEE :
easy explanation
2026-09-11 03:01:03
2
Maxin :
👍👍👍
2026-09-09 22:29:13
1
To see more videos from user @easybits.tech.room, please go to the Tikwm
homepage.