选项平衡后的第一次试跑，约70%正确率

2025-06-02 17:18:30 +08:00
parent 7a725bc003
commit 3984ec002e
12 changed files with 16742 additions and 0 deletions
--- a/results/20250602_1706/deepseek-chat_metrics.json
+++ b/results/20250602_1706/deepseek-chat_metrics.json
@@ -0,0 +1,12 @@
+{
+  "timestamp": "2025-06-02T17:13:08.707748",
+  "metrics": {
+    "accuracy": 0.6700507614213198,
+    "precision_micro": 0.676923076923077,
+    "recall_micro": 0.6700507614213198,
+    "f1_micro": 0.673469387755102,
+    "precision_macro": 0.6899114693446089,
+    "recall_macro": 0.6705102040816326,
+    "f1_macro": 0.6754210676562946
+  }
+}