sub10 addendum + toolcall-v1 (2026-10-25): dataset/toolcall_v1.csv

dataset/toolcall_v1.csv

@@ -0,0 +1,19 @@
+slug,model,quant,file_bytes,leg,s1_correct,s2_correct,s3_correct,s4_correct,single_turn_correct,single_turn_total,single_turn_pct,unique_correct,unique_requests,s5_correct,s5_total,force_quit,python_bool_replies,capped,server_bin,regime
+qwen3-8-27b-q6-k,Qwen3.8-27B,Q6_K,22884408288,st-on,384,57,279,79,799,906,88.2,794,899,,,,77,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-q6-k,Qwen3.8-27B,Q6_K,22884408288,st-off,357,61,292,74,784,906,86.5,779,899,,,,78,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-q6-k,Qwen3.8-27B,Q6_K,22884408288,s5-on,,,,,,,,,,81,100,0,,,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-ud-iq3-xxs,Qwen3.8-27B,UD-IQ3_XXS,11913559104,st-on,394,57,284,80,815,906,90.0,810,899,,,,64,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-ud-iq3-xxs,Qwen3.8-27B,UD-IQ3_XXS,11913559104,st-off,357,54,285,78,774,906,85.4,770,899,,,,57,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-ud-iq3-xxs,Qwen3.8-27B,UD-IQ3_XXS,11913559104,s5-on,,,,,,,,,,80,100,0,,,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+underdog-saluki-27b-iq2-mix,Underdog-Saluki-27B-1.0,IQ2-mix,7898369152,st-on,373,54,273,81,781,906,86.2,777,899,,,,66,2,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+underdog-saluki-27b-iq2-mix,Underdog-Saluki-27B-1.0,IQ2-mix,7898369152,st-off,379,61,289,77,806,906,89.0,800,899,,,,53,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+underdog-saluki-27b-iq2-mix,Underdog-Saluki-27B-1.0,IQ2-mix,7898369152,s5-on,,,,,,,,,,71,100,1,,,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-gsq-rco-iq2-xs,Qwen3.8-27B,GSQ-RCO IQ2_XS,8422841472,st-on,387,57,281,80,805,906,88.9,801,899,,,,49,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-gsq-rco-iq2-xs,Qwen3.8-27B,GSQ-RCO IQ2_XS,8422841472,st-off,271,60,294,69,694,906,76.6,689,899,,,,55,0,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+qwen3-8-27b-gsq-rco-iq2-xs,Qwen3.8-27B,GSQ-RCO IQ2_XS,8422841472,s5-on,,,,,,,,,,76,100,0,,,ggml-org/llama.cpp mainline b10371 (5d16e81dd),"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-pq2-0,Ternary-Bonsai-2-27B,PQ2_0,7206168928,st-on,389,57,283,79,808,906,89.2,803,899,,,,81,1,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-pq2-0,Ternary-Bonsai-2-27B,PQ2_0,7206168928,st-off,390,60,289,71,810,906,89.4,805,899,,,,76,0,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-pq2-0,Ternary-Bonsai-2-27B,PQ2_0,7206168928,s5-on,,,,,,,,,,73,100,0,,,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-ptq1-0,Ternary-Bonsai-2-27B,PTQ1_0,5946648928,st-on,394,58,282,79,813,906,89.7,807,899,,,,79,1,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-ptq1-0,Ternary-Bonsai-2-27B,PTQ1_0,5946648928,st-off,389,60,289,70,808,906,89.2,803,899,,,,76,0,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-off, max_tokens 16384, ctx 24576, greedy (temperature 0), preserve_thinking default (undefined = keep)"
+ternary-bonsai-2-27b-ptq1-0,Ternary-Bonsai-2-27B,PTQ1_0,5946648928,s5-on,,,,,,,,,,73,100,0,,,PrismML-Eng/llama.cpp prism 4fbda12,"raw /completion, official Qwen3.8 template client-side, think-on, max_tokens 16384, ctx 40960, greedy (temperature 0), preserve_thinking default (undefined = keep)"