TechMakesArt
Region: US
Tuesday 07 April 2026 14:57:04 GMT
4344
164
21
2
Music
Download
Comments
bengod :
Great work man you are a inspiration, thank you
2026-04-07 21:25:25
0
Velanox :
that hardware is a beast and good that u use it for AI
2026-04-08 08:49:03
1
Pascual Cora :
That is awesome! Looking forward to your test results!
2026-04-07 15:15:20
1
Enrico Wolf39 :
A new llama.cpp came out fixing the a70 with q8 speed makes it 3x faster
2026-04-07 17:17:22
1
Type: Null :
Congratulations in the PR!
2026-04-08 05:02:14
0
Kaleex Laerten :
are you able to make them use tools? like mcp etc
2026-04-07 16:28:01
0
Dan L :
cool
2026-04-07 16:28:28
2
Epipoiesis :
Nice! Congratulations!
2026-04-08 08:48:45
1
orangehipo :
Thank you for sharing
2026-04-08 09:33:07
0
Paul Thomas :
I wish I had the money for a GPU. stuck trying to shoe horn stuff on CPU only and almost got a workable system but hit a RAM limit. Now I put more ram in it, nothing wants to play nice 😞
2026-04-08 15:32:09
0
FinnishFox Tech :
awesome stuff 🔥
2026-04-07 19:48:01
0
🥳Andy🫡 :
gdr7 are out
2026-04-07 15:43:48
0
getklarted :
Congratulations bro. ✨
2026-04-07 15:29:16
1
Oblation :
congratulations
2026-04-07 16:10:26
1
L0G4N_FN :
Early
2026-04-07 15:14:34
0
user3612762561377 :
Thank you for your work it’s probably efforts like yours that we’ll eventually be able to run truly large models locally. I’d like to ask a question to specialists in local AI. Right now, what are the solutions that allow us to get closer locally to the kind of reasoning we see in large cloud models (the ones with “billions of parameters”, but locally)? I’m mainly looking for strong reasoning capabilities applied to software architecture, complex projects, contracts, and similar domains. Is an approach where you run two or three mid‑sized local models in a loop, each one auditing and improving the other’s proposal until they converge, actually viable? Or are there better strategies today to achieve deep reasoning locally? Thanks in advance for your insights. I’m especially interested in feedback from developers who really know local AI.
2026-04-08 14:19:18
0
Juan P. Zepeda :
🔥🔥🔥🔥
2026-04-07 15:40:12
0
HyperFocus LyNdz :
💚💚💚
2026-04-07 17:44:42
0
To see more videos from user @techmakesart, please go to the Tikwm
homepage.
