06_chatgpt_and_application.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ChatGPT来了，更快的速度更低的价格"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import openai\n",
    "import os\n",
    "\n",
    "openai.api_key = os.environ.get(\"OPENAI_API_KEY\")\n",
    "\n",
    "class Conversation:\n",
    "    def __init__(self, prompt, num_of_round):\n",
    "        self.prompt = prompt\n",
    "        self.num_of_round = num_of_round\n",
    "        self.messages = []\n",
    "        self.messages.append({\"role\": \"system\", \"content\": self.prompt})\n",
    "\n",
    "    def ask(self, question):\n",
    "        try:\n",
    "            self.messages.append( {\"role\": \"user\", \"content\": question})\n",
    "            response = openai.ChatCompletion.create(\n",
    "                model=\"gpt-3.5-turbo\",\n",
    "                messages=self.messages,\n",
    "                temperature=0.5,\n",
    "                max_tokens=2048,\n",
    "                top_p=1,\n",
    "            )\n",
    "        except Exception as e:\n",
    "            print(e)\n",
    "            return e\n",
    "\n",
    "        message = response[\"choices\"][0][\"message\"][\"content\"]\n",
    "        self.messages.append({\"role\": \"assistant\", \"content\": message})\n",
    "        \n",
    "        if len(self.messages) > self.num_of_round*2 + 1:\n",
    "            del self.messages[1:3]\n",
    "        return message\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "User : 你是谁？\n",
      "Assistant : 我是一个人工智能语言模型，专门为您提供中文语言服务。如果您有任何关于做菜的问题需要帮助，我很乐意为您提供帮助。\n",
      "\n",
      "User : 请问鱼香肉丝怎么做？\n",
      "Assistant : 鱼香肉丝的做法如下：\n",
      "1. 瘦肉切成丝，加入盐、淀粉、料酒腌制10分钟；\n",
      "2. 蒜、姜、葱切末备用；\n",
      "3. 青红椒切丝备用；\n",
      "4. 热锅凉油，放入肉丝煸炒至变色，盛出备用；\n",
      "5. 锅中留底油，放入蒜、姜、葱末煸炒出香味，加入青红椒丝煸炒；\n",
      "6. 加入肉丝翻炒，加入酱油、糖、醋、盐、味精、水淀粉烧开汁后关火，淋上香油即可。\n",
      "\n",
      "User : 那蚝油牛肉呢？\n",
      "Assistant : 蚝油牛肉的做法如下：\n",
      "1. 牛肉切成薄片，加入盐、淀粉、料酒腌制10分钟；\n",
      "2. 青红椒、洋葱切成丝备用；\n",
      "3. 热锅凉油，放入牛肉片煸炒至变色，盛出备用；\n",
      "4. 锅中留底油，放入青红椒、洋葱丝煸炒；\n",
      "5. 加入蚝油、酱油、糖、盐、味精、水淀粉烧开汁后放入牛肉片翻炒均匀；\n",
      "6. 最后淋上少许香油即可上桌。\n",
      "\n"
     ]
    }
   ],
   "source": [
    "prompt = \"\"\"你是一个中国厨师，用中文回答做菜的问题。你的回答需要满足以下要求:\n",
    "1. 你的回答必须是中文\n",
    "2. 回答限制在100个字以内\"\"\"\n",
    "conv1 = Conversation(prompt, 3)\n",
    "question1 = \"你是谁？\"\n",
    "print(\"User : %s\" % question1)\n",
    "print(\"Assistant : %s\\n\" % conv1.ask(question1))\n",
    "\n",
    "question2 = \"请问鱼香肉丝怎么做？\"\n",
    "print(\"User : %s\" % question2)\n",
    "print(\"Assistant : %s\\n\" % conv1.ask(question2))\n",
    "\n",
    "question3 = \"那蚝油牛肉呢？\"\n",
    "print(\"User : %s\" % question3)\n",
    "print(\"Assistant : %s\\n\" % conv1.ask(question3))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "User : 我问你的第一个问题是什么？\n",
      "Assistant : 你的第一个问题是“你是谁？”。\n",
      "\n"
     ]
    }
   ],
   "source": [
    "question4 = \"我问你的第一个问题是什么？\"\n",
    "print(\"User : %s\" % question4)\n",
    "print(\"Assistant : %s\\n\" % conv1.ask(question4))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "User : 我问你的第一个问题是什么？\n",
      "Assistant : 你的第一个问题是“请问鱼香肉丝怎么做？”。\n",
      "\n"
     ]
    }
   ],
   "source": [
    "question5 = \"我问你的第一个问题是什么？\"\n",
    "print(\"User : %s\" % question5)\n",
    "print(\"Assistant : %s\\n\" % conv1.ask(question5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 通过API计算Token数量"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Conversation2:\n",
    "    def __init__(self, prompt, num_of_round):\n",
    "        self.prompt = prompt\n",
    "        self.num_of_round = num_of_round\n",
    "        self.messages = []\n",
    "        self.messages.append({\"role\": \"system\", \"content\": self.prompt})\n",
    "\n",
    "    def ask(self, question):\n",
    "        try:\n",
    "            self.messages.append( {\"role\": \"user\", \"content\": question})\n",
    "            response = openai.ChatCompletion.create(\n",
    "                model=\"gpt-3.5-turbo\",\n",
    "                messages=self.messages,\n",
    "                temperature=0.5,\n",
    "                max_tokens=2048,\n",
    "                top_p=1,\n",
    "            )\n",
    "        except Exception as e:\n",
    "            print(e)\n",
    "            return e\n",
    "\n",
    "        message = response[\"choices\"][0][\"message\"][\"content\"]\n",
    "        num_of_tokens = response['usage']['total_tokens']\n",
    "        self.messages.append({\"role\": \"assistant\", \"content\": message})\n",
    "        \n",
    "        if len(self.messages) > self.num_of_round*2 + 1:\n",
    "            del self.messages[1:3]\n",
    "        return message, num_of_tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "询问 {你是谁？} 消耗的token数量是 : 118\n",
      "询问 {请问鱼香肉丝怎么做？} 消耗的token数量是 : 415\n",
      "询问 {那蚝油牛肉呢？} 消耗的token数量是 : 666\n",
      "询问 {我问你的第一个问题是什么？} 消耗的token数量是 : 697\n",
      "询问 {我问你的第一个问题是什么？} 消耗的token数量是 : 690\n"
     ]
    }
   ],
   "source": [
    "conv2 = Conversation2(prompt, 3)\n",
    "questions = [question1, question2, question3, question4, question5]\n",
    "for question in questions:\n",
    "    answer, num_of_tokens = conv2.ask(question)\n",
    "    print(\"询问 {%s} 消耗的token数量是 : %d\" % (question, num_of_tokens))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 通过Tiktoken库计算Token数量"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "总共消耗的token数量是 : 114\n",
      "Prompt消耗 65 Token, 问题消耗 5 Token，回答消耗 31 Token，总共消耗 101 Token\n"
     ]
    }
   ],
   "source": [
    "import tiktoken\n",
    "encoding = tiktoken.get_encoding(\"cl100k_base\")\n",
    "\n",
    "conv2 = Conversation2(prompt, 3)\n",
    "question1 = \"你是谁？\"\n",
    "answer1, num_of_tokens = conv2.ask(question1)\n",
    "print(\"总共消耗的token数量是 : %d\" % (num_of_tokens))\n",
    "\n",
    "prompt_count = len(encoding.encode(prompt))\n",
    "question1_count = len(encoding.encode(question1))\n",
    "answer1_count = len(encoding.encode(answer1))\n",
    "total_count = prompt_count + question1_count + answer1_count\n",
    "print(\"Prompt消耗 %d Token, 问题消耗 %d Token，回答消耗 %d Token，总共消耗 %d Token\" % (prompt_count, question1_count, answer1_count, total_count))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[27, 91, 318, 5011, 91, 29, 9125, 198]\n",
      "[27, 91, 318, 6345, 91, 397]\n",
      "[27, 91, 318, 5011, 91, 29, 882, 198]\n",
      "[27, 91, 318, 5011, 91, 29, 78191, 198]\n",
      "系统拼接的标记消耗 36 Token\n"
     ]
    }
   ],
   "source": [
    "system_start_count = len(encoding.encode(\"<|im_start|>system\\n\"))\n",
    "print(encoding.encode(\"<|im_start|>system\\n\"))\n",
    "end_count = len(encoding.encode(\"<|im_end|>\\n\"))\n",
    "print(encoding.encode(\"<|im_end|>\\n\"))\n",
    "user_start_count = len(encoding.encode(\"<|im_start|>user\\n\"))\n",
    "print(encoding.encode(\"<|im_start|>user\\n\"))\n",
    "assistant_start_count = len(encoding.encode(\"<|im_start|>assistant\\n\"))\n",
    "print(encoding.encode(\"<|im_start|>assistant\\n\"))\n",
    "\n",
    "total_mark_count = system_start_count + user_start_count + assistant_start_count + end_count*2\n",
    "print(\"系统拼接的标记消耗 %d Token\" % total_mark_count)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradio帮你快速搭建一个聊天界面"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting package metadata (current_repodata.json): done\n",
      "Solving environment: done\n",
      "\n",
      "# All requested packages already installed.\n",
      "\n",
      "\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "%conda install gradio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on local URL:  http://127.0.0.1:7860\n",
      "\n",
      "To create a public link, set `share=True` in `launch()`.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><iframe src=\"http://127.0.0.1:7860/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import gradio as gr\n",
    "prompt = \"\"\"你是一个中国厨师，用中文回答做菜的问题。你的回答需要满足以下要求:\n",
    "1. 你的回答必须是中文\n",
    "2. 回答限制在100个字以内\"\"\"\n",
    "\n",
    "conv = Conversation(prompt, 5)\n",
    "\n",
    "def predict(input, history=[]):\n",
    "    history.append(input)\n",
    "    response = conv.ask(input)\n",
    "    history.append(response)\n",
    "    responses = [(u,b) for u,b in zip(history[::2], history[1::2])]\n",
    "    return responses, history\n",
    "\n",
    "with gr.Blocks(css=\"#chatbot{height:350px} .overflow-y-auto{height:500px}\") as demo:\n",
    "    chatbot = gr.Chatbot(elem_id=\"chatbot\")\n",
    "    state = gr.State([])\n",
    "\n",
    "    with gr.Row():\n",
    "        txt = gr.Textbox(show_label=False, placeholder=\"Enter text and press enter\").style(container=False)\n",
    "\n",
    "    txt.submit(predict, [txt, state], [chatbot, state])\n",
    "\n",
    "demo.launch()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  },
  "vscode": {
   "interpreter": {
    "hash": "8114e84f04cf14e493992e1b725447accf84073d5ec18e7063d492738bf032cb"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}