David Soušek
sousekd
AI & ML interests
Running AI on-prem with agents processing confidential data. Supporting government, private sector, and all kinds of weirdos.
Recent Activity
liked a model about 3 hours ago
AesSedai/MiMo-V2.5-Pro-GGUF new activity 8 days ago
AesSedai/GLM-5.1-GGUF:Q5_K_M inference speed on llama.cpp vs ik_llama.cpp liked a model 8 days ago
AesSedai/GLM-5.1-GGUFOrganizations
None yet
Q5_K_M inference speed on llama.cpp vs ik_llama.cpp
❤️👍 2
#1 opened 8 days ago
by sousekd
Great model for single GPU use cases.
🔥 4
16
#1 opened 17 days ago
by 
phakio
No think tags.
10
#4 opened 21 days ago
by 
DrRos
Q4_X on Blackwell + EPYC + DDR 5
👍🔥 4
6
#3 opened 21 days ago
by sousekd
Working configuration for Nvidia Blackwell
13
#4 opened 29 days ago
by 
luismiguelsaez
The Best Kimi Quant!
🔥 4
13
#1 opened 21 days ago
by 
anikifoss
IQ5_K 136.891 GiB
🔥 2
30
#9 opened 3 months ago
by 
Hunterx
IQ3_KS on EPYC 9355 + 1x RTX 5090
🔥 2
5
#5 opened 2 months ago
by sousekd
1x RTX 5090 on EPYC, CPU-only speeds
❤️ 3
11
#5 opened 3 months ago
by sousekd
EPYC 9355 CPU-only sweep-bench
5
#3 opened 3 months ago
by sousekd
EPYC 9355 CPU-only sweep-bench
🔥 1
#6 opened 3 months ago
by sousekd
EPYC, RTX 5090 vs RTX 6000
🔥 1
7
#7 opened 5 months ago
by sousekd
Benchmarks (Q8_0-Q4_0, RTX 5090, EPYC)
🔥 1
5
#5 opened 6 months ago
by sousekd
Going to fit it into RTX PRO 6000!
👍 1
12
#1 opened 5 months ago
by 
mtcl
IQ2_KS passes the Moonshot K2 Vendor Verifier test
🔥 3
4
#8 opened 5 months ago
by 
whoisjeremylam
Output not respecting lines breaks?
12
#10 opened 6 months ago
by 
justj0sh
Testing smol-IQ4_KSS
🚀 1
9
#8 opened 6 months ago
by 
shewin
Missing <think> tags
👀 1
1
#3 opened 6 months ago
by sousekd
Benchmarks :)
3
#1 opened 7 months ago
by sousekd