Files
MatBench/layer1/ALL-merge/eval-deepseek-v3.log
2025-05-28 10:55:34 +08:00

10 lines
282 B
Plaintext

Accuracy of deepseek-chat: {
'accuracy': 0.8643731392656302,
'precision_micro': 0.8502372479240806,
'recall_micro': 0.9483956334766788,
'f1_micro': 0.8966379984362783,
'precision_macro': 0.7204990997584283,
'recall_macro': 0.9738882746226911,
'f1_macro': 0.8145522821153672
}