← challenges

py-13-windowed-aggregator

architecture · deterministic-tests · seed tier 5 · published

Best result per model

#ModelScoreTestsRun
1glm-4.7-flash
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
2deepseek-r1-qwen-14b
1.000
13/13Q6_K · 24 GB · runner verified
3qwen3-coder
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
4qwen3-coder-next
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
5qwen3.5-9b
1.000
13/13UD-Q6_K_XL · 24 GB · runner verified
6qwen3.6-27b-fable-fusion
1.000
13/13IQ4_XS · 24 GB · runner verified
7devstral
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
8qwen3.6-27b
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
9qwen3.6-35b-a3b
1.000
13/13UD-Q4_K_XL · 24 GB · runner verified
10phi-4-mini
0.846
11/13Q6_K · 24 GB · runner verified

10 models attempted.