Update app.py with full DAN-L3-R1-8B support

app.py

@@ -1,70 +1,87 @@
+# app.py
+import os
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
 import gradio as gr
-from huggingface_hub import InferenceClient
 
+# 设置缓存目录（节省空间）
+os.environ["HF_HOME"] = "/tmp/hf_cache"
 
-def respond(
-    message,
-    history: list[dict[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    hf_token: gr.OAuthToken,
-):
-    """
-    For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-    """
-    client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")
+# 模型名称（来自 Hugging Face）
+model_name = "UnfilteredAI/DAN-L3-R1-8B"
 
-    messages = [{"role": "system", "content": system_message}]
+# 加载分词器
+tokenizer = AutoTokenizer.from_pretrained(model_name)
 
-    messages.extend(history)
+# 创建文本生成 pipeline
+pipe = pipeline(
+    "text-generation",
+    model=model_name,
+    tokenizer=tokenizer,
+    model_kwargs={"torch_dtype": "auto"},  # 自动选择精度（如 float16）
+    device_map="auto",  # 自动使用 GPU（如果有）
+    max_new_tokens=512,
+    temperature=0.7,
+    top_p=0.9,
+    repetition_penalty=1.1,
+    return_full_text=False,  # 只返回生成的回复，不包含输入
+)
 
-    messages.append({"role": "user", "content": message})
+# 定义生成函数
+def generate_response(user_input, history):
+    # 将对话历史转换为 messages 格式
+    messages = []
+    for idx, (user_msg, assistant_msg) in enumerate(history):
+        messages.append({"role": "user", "content": user_msg})
+        messages.append({"role": "assistant", "content": assistant_msg})
+    
+    # 添加当前用户输入
+    messages.append({"role": "user", "content": user_input})
 
-    response = ""
+    # 调用模型生成回复
+    try:
+        outputs = pipe(messages)
+        response = outputs[0]["generated_text"].strip()
+    except Exception as e:
+        response = f"❌ 模型生成出错：{str(e)}"
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
+    # 返回更新后的对话历史
+    history.append((user_input, response))
+    return "", history
 
-        response += token
-        yield response
+# 创建 Gradio 界面
+with gr.Blocks(title="💬 DAN-L3-R1-8B AI 助手", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 💬 DAN-L3-R1-8B AI 助手")
+    gr.Markdown("基于 [UnfilteredAI/DAN-L3-R1-8B](https://huggingface.co/UnfilteredAI/DAN-L3-R1-8B) 的对话系统")
 
+    chatbot = gr.Chatbot(height=600, show_copy_button=True)
+    with gr.Row():
+        textbox = gr.Textbox(
+            placeholder="请输入你的问题...",
+            show_label=False,
+            scale=7,
+            container=False,
+        )
+        submit_btn = gr.Button("发送", scale=1)
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-chatbot = gr.ChatInterface(
-    respond,
-    type="messages",
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+    # 示例问题
+    gr.Examples(
+        examples=[
+            "你是谁？",
+            "讲个笑话",
+            "写一首关于秋天的诗",
+            "解释量子力学"
+        ],
+        inputs=textbox
+    )
 
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
-    chatbot.render()
+    # 清空按钮
+    clear_btn = gr.Button("🗑️ 清空聊天")
 
+    # 事件绑定
+    textbox.submit(fn=generate_response, inputs=[textbox, chatbot], outputs=[textbox, chatbot])
+    submit_btn.click(fn=generate_response, inputs=[textbox, chatbot], outputs=[textbox, chatbot])
+    clear_btn.click(fn=lambda: ("", None), outputs=[textbox, chatbot])
 
+# 启动应用
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()

requirements.txt.txt

@@ -0,0 +1,7 @@
+transformers>=4.34.0
+torch>=2.1.0
+accelerate>=0.25.0
+gradio>=4.0.0
+sentencepiece
+safetensors
+huggingface_hub