Update app.py

app.py

@@ -1,105 +1,132 @@
-# app.py  ── 2025-06-08 适配 HuggingFace CPU Space
-import os, logging, gradio as gr
+# app.py
+import gradio as gr
+import logging
+
 from langchain_community.vectorstores import Chroma
 from langchain_huggingface import HuggingFaceEmbeddings
 from langchain.chains import RetrievalQA
+from langchain.schema import Document
 from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
 from langchain.llms import HuggingFacePipeline
 
 logging.basicConfig(level=logging.INFO)
 
-# ========= 1. 载入本地向量库 =========
-embedder = HuggingFaceEmbeddings(
+# === 加载向量库 ===
+embedding_model = HuggingFaceEmbeddings(
     model_name="sentence-transformers/paraphrase-multilingual-mpnet-base-v2"
 )
-VEC_DIR = "vector_store"
-if not (os.path.isdir(VEC_DIR) and os.path.isfile(f"{VEC_DIR}/chroma.sqlite3")):
-    raise RuntimeError(f"❌ 未找到完整向量库 {VEC_DIR}，请先执行 build_vector_store.py")
-
-vectordb = Chroma(persist_directory=VEC_DIR, embedding_function=embedder)
+vector_store = Chroma(
+    persist_directory="vector_store",
+    embedding_function=embedding_model,
+)
 
-# ========= 2. 载入轻量 LLM =========
-model_id = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"     # 1.1B CPU 可跑
+# === 加载 LLM 模型（openchat） ===
+model_id = "openchat/openchat-3.5-0106"
 tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
-    device_map="auto",          # 需要 requirements 里有 accelerate
+    device_map="auto",
     torch_dtype="auto",
 )
-generator = pipeline(
+gen_pipe = pipeline(
     "text-generation",
     model=model,
     tokenizer=tokenizer,
-    max_new_tokens=256,
+    max_new_tokens=512,
     temperature=0.7,
     top_p=0.9,
 )
-llm = HuggingFacePipeline(pipeline=generator)
+llm = HuggingFacePipeline(pipeline=gen_pipe)
 
-# ========= 3. 构建 RAG 问答链 =========
+# === 构建问答链 ===
 qa_chain = RetrievalQA.from_chain_type(
     llm=llm,
     chain_type="stuff",
-    retriever=vectordb.as_retriever(search_kwargs={"k": 3}),
+    retriever=vector_store.as_retriever(search_kwargs={"k": 3}),
 )
 
-# ========= 4. 业务函数 =========
-def simple_qa(user_query: str):
+# === 智能问答函数 ===
+def simple_qa(user_query):
     if not user_query.strip():
         return "⚠️ 请输入学习问题，例如：什么是定积分？"
     try:
-        return qa_chain.run(user_query)
+        answer = qa_chain.run(user_query)
+        return answer
     except Exception as e:
-        logging.error(f"[QA ERROR] {e}")
-        return f"⚠️ 问答失败：{e}"
+        logging.error(f"问答失败: {e}")
+        return f"⚠️ 问答失败，请稍后再试。\n[调试信息] {e}"
 
+# === 大纲生成函数 ===
 def generate_outline(topic: str):
     if not topic.strip():
-        return "⚠️ 请输入章节或主题", ""
+        return "⚠️ 请输入章节或主题，例如：高等数学 第六章 定积分", ""
     try:
-        docs = vectordb.as_retriever(search_kwargs={"k": 3}).get_relevant_documents(topic)
-        snippet = "\n---\n".join(d.page_content for d in docs)
+        docs = vector_store.as_retriever(search_kwargs={"k": 3}).get_relevant_documents(topic)
+        if not docs:
+            return "⚠️ 没有找到相关内容，请换个关键词试试。", ""
+        snippet = "\n".join(doc.page_content for doc in docs)
         prompt = (
-            f"请基于以下资料，为「{topic}」生成结构化学习大纲，格式：\n"
-            f"一、章节标题\n  1. 节标题\n    （1）要点…\n\n"
-            f"资料：\n{snippet}\n\n大纲："
+            f"根据以下内容，为“{topic}”生成大学本科层次的结构化学习大纲，格式如下：\n"
+            f"一、章节标题\n  1. 节标题\n    （1）要点描述\n...\n\n"
+            f"文档内容：\n{snippet}\n\n学习大纲："
         )
-        outline = llm.invoke(prompt).strip()
-        return outline, snippet
+        result = llm.generate(prompt).generations[0][0].text.strip()
+        return result, snippet
     except Exception as e:
-        logging.error(f"[OUTLINE ERROR] {e}")
-        return "⚠️ 生成失败", ""
+        logging.error(f"大纲生成失败: {e}")
+        return "⚠️ 抱歉，生成失败，请稍后再试。", ""
 
-def placeholder(*_):
-    return "功能待开发…"
+# === 占位函数 ===
+def placeholder_fn(*args, **kwargs):
+    return "功能尚未实现，请等待后续更新。"
+
+# === Gradio UI ===
+with gr.Blocks() as demo:
+    gr.Markdown("# 智能学习助手 v2.0\n— 大学生专业课学习助手 —")
 
-# ========= 5. Gradio UI =========
-with gr.Blocks(title="智能学习助手") as demo:
-    gr.Markdown("# 智能学习助手 v2.0\n💡 大学生专业课 RAG Demo")
     with gr.Tabs():
+        # --- 模块 A：智能问答 ---
         with gr.TabItem("智能问答"):
-            chatbot = gr.Chatbot(height=350)
-            msg = gr.Textbox(placeholder="在此提问…")
-            def chat(m, hist): 
-                ans = simple_qa(m)
-                hist.append((m, ans))
-                return "", hist
-            msg.submit(chat, [msg, chatbot], [msg, chatbot])
+            gr.Markdown("> 示例：什么是函数的定义域？")
+            chatbot = gr.Chatbot()
+            user_msg = gr.Textbox(placeholder="输入您的学习问题，然后按回车或点击发送")
+            send_btn = gr.Button("发送")
+
+            def update_chat(message, chat_history):
+                reply = simple_qa(message)
+                chat_history.append((message, reply))
+                return "", chat_history
+
+            send_btn.click(update_chat, inputs=[user_msg, chatbot], outputs=[user_msg, chatbot])
+            user_msg.submit(update_chat, inputs=[user_msg, chatbot], outputs=[user_msg, chatbot])
 
+        # --- 模块 B：生成学习大纲 ---
         with gr.TabItem("生成学习大纲"):
-            topic = gr.Textbox(label="章节主题", placeholder="高等数学 第六章 定积分")
-            outline = gr.Textbox(label="学习大纲", lines=12)
-            debug = gr.Textbox(label="调试：检索片段", lines=6)
-            gen = gr.Button("生成")
-            gen.click(generate_outline, [topic], [outline, debug])
+            gr.Markdown("> 示例：高等数学 第六章 定积分")
+            topic_input = gr.Textbox(label="章节主题", placeholder="请输入章节名")
+            outline_output = gr.Textbox(label="系统生成的大纲", lines=12)
+            snippet_output = gr.Textbox(label="[调试] 检索片段展示", lines=6)
+            gen_outline_btn = gr.Button("生成大纲")
+            gen_outline_btn.click(fn=generate_outline, inputs=topic_input, outputs=[outline_output, snippet_output])
 
+        # --- 模块 C：自动出题（占位） ---
         with gr.TabItem("自动出题"):
-            placeholder(label="待开发")
+            gr.Markdown("（出题模块，待开发）")
+            topic2 = gr.Textbox(label="知识点/主题", placeholder="如：高数 第三章 多元函数")
+            difficulty2 = gr.Dropdown(choices=["简单", "中等", "困难"], label="难度")
+            count2 = gr.Slider(1, 10, step=1, label="题目数量")
+            gen_q_btn = gr.Button("开始出题")
+            gen_q_btn.click(placeholder_fn, inputs=[topic2, difficulty2, count2], outputs=topic2)
 
+        # --- 模块 D：答案批改（占位） ---
         with gr.TabItem("答案批改"):
-            placeholder(label="待开发")
+            gr.Markdown("（批改模块，待开发）")
+            std_ans = gr.Textbox(label="标准答案", lines=5)
+            user_ans = gr.Textbox(label="您的作答", lines=5)
+            grade_btn = gr.Button("开始批改")
+            grade_btn.click(placeholder_fn, inputs=[user_ans, std_ans], outputs=user_ans)
 
-    gr.Markdown("---\nPowered by LangChain • TinyLlama • Chroma")
+    gr.Markdown("---\n由 HuggingFace 提供支持 • 版本 2.0")
 
 if __name__ == "__main__":
     demo.launch()