coding

app.py

@@ -8,26 +8,29 @@ from opencc import OpenCC
 
 # 使用 進擊的巨人 数据集
 # 原数据集是是繁体中文，为了调试方便，将其转换成简体中文之后使用
-if "dataset_loaded" not in st.session_state:
-    st.session_state.dataset_loaded = False
+if "data_list" not in st.session_state:
     st.session_state.data_list = []
     st.session_state.answer_list = []
-if not st.session_state.dataset_loaded:
+
+if not st.session_state.data_list:
     try:
         with st.spinner("正在读取数据库..."):
-            converter = OpenCC('tw2s') # 'tw2s.json' 表示繁体中文到简体中文的转换
+            st.session_state.converter = OpenCC('tw2s') # 'tw2s.json' 表示繁体中文到简体中文的转换
             dataset = load_dataset("rorubyy/attack_on_titan_wiki_chinese")
+            data_list = []
+            answer_list = []
             for example in dataset["train"]:
-                converted_answer = converter.convert(example["Answer"])
-                converted_question = converter.convert(example["Question"])
-                st.session_state.answer_list.append(converted_answer)
-                st.session_state.data_list.append({"Question": converted_question, "Answer": converted_answer})
+                converted_answer = st.session_state.converter.convert(example["Answer"])
+                converted_question = st.session_state.converter.convert(example["Question"])
+                answer_list.append(converted_answer)
+                data_list.append({"Question": converted_question, "Answer": converted_answer})
+            st.session_state.answer_list = answer_list
+            st.session_state.data_list = data_list
             st.success("数据库读取完成！")
             print("数据库读取完成！")
     except Exception as e:
         st.error(f"读取数据集失败：{e}")
         st.stop()
-    st.session_state.dataset_loaded = True
 
 # 构建向量数据库 (如果需要，仅构建一次)
 if "vector_created" not in st.session_state:
@@ -117,9 +120,9 @@ with col3:
         random_index = random.randint(0, dataset_size - 1)
         # 读取随机问题
         random_question = st.session_state.data_list[random_index]["Question"]
-        random_question = converter.convert(random_question)
+        random_question = st.session_state.converter.convert(random_question)
         origin_answer = st.session_state.data_list[random_index]["Answer"]
-        origin_answer = converter.convert(origin_answer)
+        origin_answer = st.session_state.converter.convert(origin_answer)
         print('[]' + str(random_index) + '/' + str(dataset_size) + ']random_question: ' + random_question)
         print('origin_answer: ' + origin_answer)