Update app.py

app.py

@@ -1,43 +1,24 @@
-import gradio as gr
-from transformers import pipeline
-
-# 加载 Hugging Face 上的预训练模型（以 DeepSeek 为例）
-from huggingface_hub import login
-from transformers import pipeline
-
-# 登录 Hugging Face
-login(token="your_huggingface_token")
-
-# 加载模型
-model_name = "deepseek-ai/deepseek-7b"
-pipe = pipeline("text-generation", model=model_name, tokenizer=model_name)
-
-
-# 定义与模型对话的函数
-def chat_with_ai(prompt):
-    # 使用模型生成文本
-    response = pipe(prompt, max_length=100, do_sample=True)
-    return response[0]["generated_text"]
-
-# 创建 Gradio 界面
-with gr.Blocks() as demo:
-    gr.Markdown("# 🤖 AI Chatbot powered by DeepSeek")
-    
-    # 聊天框组件
-    chatbot = gr.Chatbot()
-    msg = gr.Textbox(label="Type your message:")
-    clear = gr.Button("Clear")
-
-    # 定义响应函数，处理用户输入并更新聊天历史
-    def respond(message, chat_history):
-        response = chat_with_ai(message)
-        chat_history.append((message, response))  # 将对话记录添加到聊天历史
-        return "", chat_history
-
-    # 提交消息并更新聊天记录
-    msg.submit(respond, [msg, chatbot], [msg, chatbot])
-    # 清空聊天记录
-    clear.click(lambda: [], None, chatbot)
-
-# 启动 Gradio 界面
-demo.launch()
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
+
+# 加载 DeepSeekMath 模型和分词器
+model_name = "deepseek-ai/deepseek-math-7b-instruct"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
+model.generation_config = GenerationConfig.from_pretrained(model_name)
+model.generation_config.pad_token_id = model.generation_config.eos_token_id
+
+# 定义带有链式推理的数学问题
+messages = [
+    {"role": "user", "content": "what is the integral of x^2 from 0 to 2?\nPlease reason step by step, and put your final answer within \\boxed{}."}
+]
+
+# 将问题转换为模型输入格式
+input_tensor = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
+
+# 生成模型的输出
+outputs = model.generate(input_tensor.to(model.device), max_new_tokens=100)
+
+# 解码输出并打印结果
+result = tokenizer.decode(outputs[0][input_tensor.shape[1]:], skip_special_tokens=True)
+print(result)