DesignCoder / eval /significance_tests_v592.json
xingxm's picture
Add full V5.9.2 evaluation artifacts
51aa93b verified
Raw History Blame Contribute Delete
1.37 kB
{
"27B_AdamW_step400_vs_9B_AdamW_step200": {
"mean_diff": 4.896,
"bootstrap_ci95": [
2.867,
6.789
],
"wilcoxon_W": 5719.5,
"wilcoxon_p": 4.847597288453082e-07,
"significant_at_0.05": true
},
"27B_AdamW_step400_vs_4B_AdamW_step200": {
"mean_diff": 6.978,
"bootstrap_ci95": [
4.814,
9.154
],
"wilcoxon_W": 4877.0,
"wilcoxon_p": 4.47399700563063e-10,
"significant_at_0.05": true
},
"27B_AdamW_step400_vs_4B_Muon_step200": {
"mean_diff": 15.704,
"bootstrap_ci95": [
13.551,
17.92
],
"wilcoxon_W": 1574.5,
"wilcoxon_p": 4.570449701547677e-25,
"significant_at_0.05": true
},
"9B_AdamW_step200_vs_4B_AdamW_step200": {
"mean_diff": 2.082,
"bootstrap_ci95": [
-0.034,
4.176
],
"wilcoxon_W": 8359.5,
"wilcoxon_p": 0.0647704989672306,
"significant_at_0.05": false
},
"9B_AdamW_step200_vs_4B_Muon_step200": {
"mean_diff": 10.808,
"bootstrap_ci95": [
8.818,
12.804
],
"wilcoxon_W": 2929.0,
"wilcoxon_p": 3.661798966685315e-18,
"significant_at_0.05": true
},
"4B_AdamW_step200_vs_4B_Muon_step200": {
"mean_diff": 8.726,
"bootstrap_ci95": [
6.512,
10.939
],
"wilcoxon_W": 4134.5,
"wilcoxon_p": 5.27884641909125e-13,
"significant_at_0.05": true
}
}