修改train_model_grpo.py文件-开启vLLM 观察能否解决损失率值0并且无变化问题

src/train_model_grpo.py

@@ -154,7 +154,8 @@ class ModelTrainer:
 
     def load_data(self, train_data_path):
         # 加载训练集和测试集
-        train_dataset = load_dataset("json", data_files={"train": train_data_path}, split="train")
+        with open(train_data_path, 'r') as f:
+            train_dataset = load_dataset("json", data_files={"train": train_data_path}, split="train")
 
         # train_data_path: 训练数据路径，格式为 JSONL
         return train_dataset