Demos/13b. 轻松开始你的第一次 AI 视频总结（API 版）- 精简版.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6bdb36f6-b229-4e41-b32b-fb8fc00cbf97",
   "metadata": {},
   "source": [
    "# b. 使用大模型 API 对视频进行快速摘要（音频处理）- 精简版\n",
    "\n",
    "> 指导文章：[15. 用 API 实现 AI 视频摘要：动手制作属于你的 AI 视频助手](https://github.com/Hoper-J/AI-Guide-and-Demos-zh_CN/blob/master/Guide/15.%20用%20API%20实现%20AI%20视频摘要：动手制作属于你的%20AI%20视频助手.md)\n",
    "> \n",
    "\n",
    "这是精简的核心代码，非交互版，直接对应于文章内容，变量名相比于交互版会稍微修改，以更清晰的反映作用。\n",
    "\n",
    "![Pipeline](../Guide/assets/image-20240926154540268.png)\n",
    "\n",
    "你可以一键执行直接看到结果。\n",
    "\n",
    "如果你想在本地试试参数交互，访问：[a. 交互完整版](https://github.com/Hoper-J/AI-Guide-and-Demos-zh_CN/blob/master/Demos/13.%20轻松开始你的第一次%20AI%20视频总结（API%20版）.ipynb)\n",
    "\n",
    "在线链接（精简版）：[Kaggle](https://www.kaggle.com/code/aidemos/13b-ai-api) | [Colab](https://colab.research.google.com/drive/1yh2J4_Cy45fqvydH34LLtDpw10kuWutO?usp=sharing)\n",
    "\n",
    "这里还有一个简单的 [🎡 AI Summarizer 脚本](https://github.com/Hoper-J/AI-Guide-and-Demos-zh_CN/blob/master/CodePlayground/summarizer.py)供你尝试。命令行执行：\n",
    "```bash\n",
    "python summarizer.py ./example_video.mp4\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "418eee53-2d5b-407a-a9f4-c9604d51d561",
   "metadata": {},
   "source": [
    "## 第 1 部分 - 准备工作\n",
    "\n",
    "\n",
    "### 安装和导入"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7e8a3db1-98c2-492e-a07a-90484445bea1",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install srt\n",
    "!pip install datasets\n",
    "!pip install DateTime\n",
    "!pip install OpenCC==1.1.6\n",
    "!pip install opencv-contrib-python\n",
    "!pip install opencv-python\n",
    "!pip install opencv-python-headless\n",
    "!pip install openpyxl\n",
    "!pip install openai\n",
    "!pip install git+https://github.com/openai/whisper.git@ba3f3cd54b0e5b8ce1ab3de13e32122d0d5f98ab\n",
    "!pip install numpy\n",
    "!pip install soundfile\n",
    "!pip install ipywidgets\n",
    "!pip install librosa\n",
    "#!pip install 'httpx<0.28.0' # 降级 httpx 以解决关键字 'proxies' 被移除的问题，最新的 openai 库不会引发该问题，故默认注释"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "04567cf9-0147-43af-ae37-5ee5df721b37",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 标准库\n",
    "import os\n",
    "import time\n",
    "import re\n",
    "import pathlib\n",
    "import textwrap\n",
    "import datetime\n",
    "\n",
    "# 第三方库\n",
    "import numpy as np\n",
    "import srt\n",
    "import soundfile as sf\n",
    "from tqdm import tqdm\n",
    "\n",
    "# 项目相关库\n",
    "import whisper\n",
    "from datasets import load_dataset\n",
    "from openai import OpenAI"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "212aab92-ac84-4d28-b56d-37440efbe094",
   "metadata": {},
   "source": [
    "### 加载数据（二选一）\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7c073e30-3421-48d5-bd4e-f1aae59298b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 加载数据集\n",
    "# dataset_name = \"kuanhuggingface/NTU-GenAI-2024-HW9\"\n",
    "# dataset = load_dataset(dataset_name)\n",
    "\n",
    "# 加载本地 Parquet 格式的数据集\n",
    "dataset = load_dataset('parquet', data_files={'test': './data/13/test-00000-of-00001.parquet'})\n",
    "\n",
    "# 准备音频\n",
    "input_audio = dataset[\"test\"][\"audio\"][0]\n",
    "input_audio_name = input_audio[\"path\"]\n",
    "input_audio_array = input_audio[\"array\"].astype(np.float32)\n",
    "sampling_rate = input_audio[\"sampling_rate\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a0fa7c4-f197-4641-bdea-54ef2a934e1f",
   "metadata": {},
   "source": [
    "#### 直接加载 .mp3 文件\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5af8de20-6746-494a-ba6e-8087da1fe6ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "import librosa\n",
    "\n",
    "# 指定音频文件路径\n",
    "mp3_file_path = './data/13/audio.mp3'\n",
    "\n",
    "input_audio_name = os.path.basename(mp3_file_path)\n",
    "\n",
    "# 加载音频文件，指定采样率为 16000\n",
    "input_audio_array, sampling_rate = librosa.load(mp3_file_path, sr=16000)\n",
    "\n",
    "# 打印音频数据的采样率和形状，确保加载成功\n",
    "print(f\"采样率: {sampling_rate}\")\n",
    "print(f\"音频数据形状: {input_audio_array.shape}\")\n",
    "\n",
    "print(f\"现在我们将转录音频: ({input_audio_name})。\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "693fc7e4-316d-48a1-8bdc-8e8a0e55b5b4",
   "metadata": {},
   "source": [
    "## 第 2 部分 - 自动语音识别 (ASR)\n",
    "\n",
    "我们将使用 OpenAI 的 Whisper 模型将音频转换为字幕。\n",
    "\n",
    "下图是处理的样例过程，对应于之后所使用的音频：\n",
    "\n",
    "![image-20240926155512340](../Guide/assets/image-20240926155512340.png)\n",
    "\n",
    "### 定义语音识别函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "522d3863-ce0d-4304-aa89-307b2c46d556",
   "metadata": {},
   "outputs": [],
   "source": [
    "def speech_recognition(model_name, input_audio, output_subtitle_path, decode_options, cache_dir=\"./\"):\n",
    "    # 加载模型\n",
    "    model = whisper.load_model(name=model_name, download_root=cache_dir)\n",
    "\n",
    "    # 转录音频\n",
    "    transcription = model.transcribe(\n",
    "        audio=input_audio,\n",
    "        language=decode_options[\"language\"],\n",
    "        verbose=False,\n",
    "        initial_prompt=decode_options[\"initial_prompt\"],\n",
    "        temperature=decode_options[\"temperature\"]\n",
    "    )\n",
    "\n",
    "    # 处理转录结果，生成字幕文件\n",
    "    subtitles = []\n",
    "    for i, segment in enumerate(transcription[\"segments\"]):\n",
    "        start_time = datetime.timedelta(seconds=segment[\"start\"])\n",
    "        end_time = datetime.timedelta(seconds=segment[\"end\"])\n",
    "        text = segment[\"text\"]\n",
    "        subtitles.append(srt.Subtitle(index=i, start=start_time, end=end_time, content=text))\n",
    "\n",
    "    srt_content = srt.compose(subtitles)\n",
    "\n",
    "    # 保存字幕文件\n",
    "    with open(output_subtitle_path, \"w\", encoding=\"utf-8\") as file:\n",
    "        file.write(srt_content)\n",
    "\n",
    "    print(f\"字幕已保存到 {output_subtitle_path}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0353e6dd-e80c-45d9-b162-80a53fbc9a59",
   "metadata": {},
   "source": [
    "### 设置参数\n",
    "\n",
    "注意，这里设置的参数都是 Whisper 相关的，与后续的 AI 摘要不同。\n",
    "\n",
    "你可以通过 model_name 设置不同的模型，使用**标识**指定。通过[官方仓库](https://github.com/openai/whisper)所提供的数据，我们可以看到不同模型需要的显存大小：\n",
    "\n",
    "| Size   | Parameters | English-only model | Multilingual model | Required VRAM | Relative speed |\n",
    "| ------ | ---------- | ------------------ | ------------------ | ------------- | -------------- |\n",
    "| tiny   | 39 M       | `tiny.en`          | `tiny`             | ~1 GB         | ~32x           |\n",
    "| base   | 74 M       | `base.en`          | `base`             | ~1 GB         | ~16x           |\n",
    "| small  | 244 M      | `small.en`         | `small`            | ~2 GB         | ~6x            |\n",
    "| medium | 769 M      | `medium.en`        | `medium`           | ~5 GB         | ~2x            |\n",
    "| large  | 1550 M     | N/A                | `large`            | ~10 GB        | 1x             |\n",
    "\n",
    "**解释：**\n",
    "\n",
    "- **Size (大小)**：表示模型的尺寸，不同大小的模型训练时使用的数据量不同，因此性能和精度也不同。较大的模型通常会有更高的精度。`Medium` 是个不错的选择，`tiny` 和 `base` 效果一般，用于学习的话也可以。\n",
    "- **Parameters (参数量)**：模型的参数数量，表示模型的复杂度。参数越多，模型的性能通常越好，但也会占用更多的计算资源。\n",
    "- **English-only model (仅限英文模型)**：模型的**标识**名称，只用于处理英文音频转录，适用于仅需要处理英文语音的场景。\n",
    "- **Multilingual model (多语言模型)**：模型的**标识**名称，用于在代码中加载相应的模型，对应于接下来的 `model_name` 参数。\n",
    "- **Required VRAM (所需显存)**：指运行该模型时所需的显存大小。如果你对参数和显存的对应关系感兴趣，可以阅读之前的文章：[《07. 探究模型参数与显存的关系以及不同精度造成的影响.md》](https://github.com/Hoper-J/AI-Guide-and-Demos-zh_CN/blob/master/Guide/07.%20探究模型参数与显存的关系以及不同精度造成的影响.md)。\n",
    "- **Relative speed (相对速度)**：相对速度表示模型处理语音转录任务的效率。数字越大，模型处理速度越快，与模型的参数量成反比。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "aac5b7aa-1f0a-4a53-90a3-250b13ea51d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 模型名称，可选 'tiny', 'base', 'small', 'medium', 'large-v3'\n",
    "model_name = 'medium'\n",
    "\n",
    "# 语言\n",
    "language = 'zh'  # 选择语音识别的目标语言，如 'zh' 表示中文\n",
    "\n",
    "# 初始 prompt，可选\n",
    "initial_prompt = '请用中文'  # 如果需要，可以为 Whisper 模型设置初始 prompt 语句\n",
    "\n",
    "# 采样温度，控制模型的输出多样性\n",
    "temperature = 0.0  # 0 表示最确定性的输出，范围为 0-1\n",
    "\n",
    "# 输出文件后缀\n",
    "suffix = '信号与人生'\n",
    "\n",
    "# 字幕文件路径\n",
    "output_subtitle_path = f\"./output-{suffix}.srt\"\n",
    "\n",
    "# 模型缓存目录\n",
    "cache_dir = './'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f04e8038-ba29-4818-b6ed-213bd8cc5b49",
   "metadata": {},
   "source": [
    "### 运行语音识别\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "769eea63-59a1-4f7d-a9dd-e01e876543ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████| 104500/104500 [01:10<00:00, 1479.91frames/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "字幕已保存到 ./output-信号与人生.srt\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# 构建解码选项\n",
    "decode_options = {\n",
    "    \"language\": language,\n",
    "    \"initial_prompt\": initial_prompt,\n",
    "    \"temperature\": temperature\n",
    "}\n",
    "\n",
    "# 运行 ASR\n",
    "speech_recognition(\n",
    "    model_name=model_name,\n",
    "    input_audio=input_audio_array,\n",
    "    output_subtitle_path=output_subtitle_path,\n",
    "    decode_options=decode_options,\n",
    "    cache_dir=cache_dir\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1fc7ce55-820e-4b63-a6dd-a8413f1dd55e",
   "metadata": {},
   "source": [
    "### 检查结果\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3d884844-13a7-4e70-acda-6bd242e571bc",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "00:00:00,000 --> 00:00:04,000\n",
      "每次说 学问是做出来的\n",
      "\n",
      "2\n",
      "00:00:06,000 --> 00:00:08,000\n",
      "什么意思\n",
      "\n",
      "3\n",
      "00:00:08,000 --> 00:00:12,000\n",
      "要做 才会获得学问\n",
      "\n",
      "4\n",
      "00:00:13,000 --> 00:00:16,000\n",
      "你如果每天光是坐在那里听\n",
      "\n",
      "5\n",
      "00:00:17,000 --> 00:00:20,000\n",
      "学问很可能是左耳近右耳出的\n",
      "\n",
      "6\n",
      "00:00:21,000 --> 00:00:23,000\n",
      "你光是坐在那儿读\n",
      "\n",
      "7\n",
      "00:00:23,000 --> 00:00:26,000\n",
      "学问可能从眼睛进入脑海之后就忘掉了\n",
      "\n",
      "8\n",
      "00:00:26,000 --> 00:00:29,000\n",
      "如何能够学问在脑海里面\n",
      "\n",
      "9\n",
      "00:00:31,000 --> 00:00:33,000\n",
      "真的变成你自己学问\n",
      "\n",
      "10\n",
      "00:00:33,000 --> 00:00:35,000\n",
      "就是要做\n",
      "\n",
      "11\n",
      "00:00:36,000 --> 00:00:39,000\n",
      "可能有很多同学有这个经验\n",
      "\n",
      "12\n",
      "00:00:39,000 --> 00:00:41,000\n",
      "你如果去修某一门课\n",
      "\n",
      "13\n",
      "00:00:41,000 --> 00:00:44,000\n",
      "或者做某一个实验\n",
      "\n",
      "14\n",
      "00:00:44,000 --> 00:00:47,000\n",
      "在期末就是要教一个final project\n",
      "\n",
      "15\n",
      "00:00:48,000 --> 00:00:50,000\n",
      "那个final project就是要你把\n",
      "\n",
      "16\n",
      "00:00:51,000 --> 00:00:53,000\n",
      "学到的很多东西\n",
      "\n",
      "17\n",
      "00:00:53,000 --> 00:00:56,000\n",
      "最后整合在你的final project里面\n",
      "\n",
      "18\n",
      "00:00:56,000 --> 00:00:58,000\n",
      "最后做出来的时候\n",
      "\n",
      "19\n",
      "00:00:58,000 --> 00:01:00,000\n",
      "就是把它们都整合了\n",
      "\n",
      "20\n",
      "00:01:00,000 --> 00:01:02,000\n",
      "当你学期结束\n",
      "\n",
      "21\n",
      "00:01:02,000 --> 00:01:04,000\n",
      "真的把final project做完的时候\n",
      "\n",
      "22\n",
      "00:01:04,000 --> 00:01:05,000\n",
      "你会忽然发现\n",
      "\n",
      "23\n",
      "00:01:05,000 --> 00:01:07,000\n",
      "我真的学到很多东西\n",
      "\n",
      "24\n",
      "00:01:07,000 --> 00:01:10,000\n",
      "那就是做出来的学问\n",
      "\n",
      "25\n",
      "00:01:11,000 --> 00:01:13,000\n",
      "也许可以举另外一个例子\n",
      "\n",
      "26\n",
      "00:01:13,000 --> 00:01:15,000\n",
      "就是你如果学了某一些\n",
      "\n",
      "27\n",
      "00:01:15,000 --> 00:01:17,000\n",
      "很复杂的演算法或者什么\n",
      "\n",
      "28\n",
      "00:01:17,000 --> 00:01:21,000\n",
      "好像觉得那些不见得在你的脑海里\n",
      "\n",
      "29\n",
      "00:01:21,000 --> 00:01:24,000\n",
      "可是后来老师出了个习题\n",
      "\n",
      "30\n",
      "00:01:24,000 --> 00:01:26,000\n",
      "那个习题叫你写一个很大的程式\n",
      "\n",
      "31\n",
      "00:01:26,000 --> 00:01:28,000\n",
      "要把所有东西都包进去\n",
      "\n",
      "32\n",
      "00:01:29,000 --> 00:01:31,000\n",
      "当你把这个程式写完的时候\n",
      "\n",
      "33\n",
      "00:01:31,000 --> 00:01:32,000\n",
      "你会发现\n",
      "\n",
      "34\n",
      "00:01:32,000 --> 00:01:34,000\n",
      "你忽然把演算法\n",
      "\n",
      "35\n",
      "00:01:34,000 --> 00:01:35,000\n",
      "所有东西都弄通了\n",
      "\n",
      "36\n",
      "00:01:35,000 --> 00:01:38,000\n",
      "那就是学问是做出来的\n",
      "\n",
      "37\n",
      "00:01:38,000 --> 00:01:40,000\n",
      "所以我们永远要记得\n",
      "\n",
      "38\n",
      "00:01:40,000 --> 00:01:44,000\n",
      "尽量多动手多做\n",
      "\n",
      "39\n",
      "00:01:44,000 --> 00:01:46,000\n",
      "在动手跟做的过程之中\n",
      "\n",
      "40\n",
      "00:01:46,000 --> 00:01:49,000\n",
      "学问才可以变成是自己的\n",
      "\n",
      "41\n",
      "00:01:49,000 --> 00:01:51,000\n",
      "同样的情形就是说\n",
      "\n",
      "42\n",
      "00:01:51,000 --> 00:01:54,000\n",
      "很多时候这样动手或者做的\n",
      "\n",
      "43\n",
      "00:01:56,000 --> 00:01:57,000\n",
      "表现或者成绩\n",
      "\n",
      "44\n",
      "00:01:57,000 --> 00:02:00,000\n",
      "没有一个成绩单上的数字\n",
      "\n",
      "45\n",
      "00:02:00,000 --> 00:02:03,000\n",
      "使得很多人觉得那不重要\n",
      "\n",
      "46\n",
      "00:02:03,000 --> 00:02:05,000\n",
      "很多人甚至觉得\n",
      "\n",
      "47\n",
      "00:02:05,000 --> 00:02:07,000\n",
      "这门课要做final project\n",
      "\n",
      "48\n",
      "00:02:07,000 --> 00:02:09,000\n",
      "我就不修了太累了\n",
      "\n",
      "49\n",
      "00:02:09,000 --> 00:02:11,000\n",
      "或者说那门课需要\n",
      "\n",
      "50\n",
      "00:02:11,000 --> 00:02:13,000\n",
      "怎么样太累我就不要做了\n",
      "\n",
      "51\n",
      "00:02:13,000 --> 00:02:14,000\n",
      "而不知道\n",
      "\n",
      "52\n",
      "00:02:14,000 --> 00:02:17,000\n",
      "其实那个才是让你做的机会\n",
      "\n",
      "53\n",
      "00:02:17,000 --> 00:02:19,000\n",
      "然后可以学到最多\n",
      "\n",
      "54\n",
      "00:02:19,000 --> 00:02:22,000\n",
      "也就是说虽然很可能\n",
      "\n",
      "55\n",
      "00:02:22,000 --> 00:02:24,000\n",
      "那么辛苦的做很多事\n",
      "\n",
      "56\n",
      "00:02:24,000 --> 00:02:28,000\n",
      "没有让你获得什么具体成绩\n",
      "\n",
      "57\n",
      "00:02:28,000 --> 00:02:30,000\n",
      "对你的overfitting可能没有帮助\n",
      "\n",
      "58\n",
      "00:02:30,000 --> 00:02:32,000\n",
      "可是对你的全面学习\n",
      "\n",
      "59\n",
      "00:02:32,000 --> 00:02:35,000\n",
      "是很有帮助是该学的\n",
      "\n",
      "60\n",
      "00:02:35,000 --> 00:02:38,000\n",
      "不要漏掉这些事\n",
      "\n",
      "61\n",
      "00:02:38,000 --> 00:02:41,000\n",
      "这是我所说的\n",
      "\n",
      "62\n",
      "00:02:41,000 --> 00:02:46,000\n",
      "这个课业内可以做的这些事\n",
      "\n",
      "63\n",
      "00:02:46,000 --> 00:02:50,000\n",
      "刚才我们讲到思考的时候\n",
      "\n",
      "64\n",
      "00:02:50,000 --> 00:02:52,000\n",
      "我觉得我漏掉一点\n",
      "\n",
      "65\n",
      "00:02:52,000 --> 00:02:55,000\n",
      "你如果修我的信号课\n",
      "\n",
      "66\n",
      "00:02:55,000 --> 00:02:56,000\n",
      "你可能会发现\n",
      "\n",
      "67\n",
      "00:02:56,000 --> 00:03:00,000\n",
      "我上课没讲到一个数学试字的时候\n",
      "\n",
      "68\n",
      "00:03:00,000 --> 00:03:02,000\n",
      "我通常都不推它的\n",
      "\n",
      "69\n",
      "00:03:02,000 --> 00:03:06,000\n",
      "我是在解释那个数学试字在说什么话\n",
      "\n",
      "70\n",
      "00:03:06,000 --> 00:03:10,000\n",
      "同样的没讲到一个什么事情的时候\n",
      "\n",
      "71\n",
      "00:03:10,000 --> 00:03:13,000\n",
      "我通常就在解释他在说什么话\n",
      "\n",
      "72\n",
      "00:03:13,000 --> 00:03:17,000\n",
      "也就是说我在讲的就是\n",
      "\n",
      "73\n",
      "00:03:17,000 --> 00:03:20,000\n",
      "我读到特本那里的时候\n",
      "\n",
      "74\n",
      "00:03:20,000 --> 00:03:21,000\n",
      "我心里怎么想的\n",
      "\n",
      "75\n",
      "00:03:21,000 --> 00:03:24,000\n",
      "也就是我在告诉同学\n",
      "\n",
      "76\n",
      "00:03:24,000 --> 00:03:28,000\n",
      "如何读书的时候\n",
      "\n",
      "77\n",
      "00:03:28,000 --> 00:03:31,000\n",
      "如何一面读一面练习思考\n",
      "\n",
      "78\n",
      "00:03:31,000 --> 00:03:35,000\n",
      "这个才是最重要的一件事\n",
      "\n",
      "79\n",
      "00:03:35,000 --> 00:03:40,000\n",
      "如何培养自己思考的能力\n",
      "\n",
      "80\n",
      "00:03:40,000 --> 00:03:42,000\n",
      "跟培养思考的习惯\n",
      "\n",
      "81\n",
      "00:03:42,000 --> 00:03:45,000\n",
      "我觉得最好的办法就是\n",
      "\n",
      "82\n",
      "00:03:45,000 --> 00:03:46,000\n",
      "读书的时候\n",
      "\n",
      "83\n",
      "00:03:46,000 --> 00:03:49,000\n",
      "凡是读到一个数学试字\n",
      "\n",
      "84\n",
      "00:03:49,000 --> 00:03:50,000\n",
      "都去想一想\n",
      "\n",
      "85\n",
      "00:03:50,000 --> 00:03:53,000\n",
      "那个数学试字到底在说什么\n",
      "\n",
      "86\n",
      "00:03:53,000 --> 00:03:56,000\n",
      "凡是读到特本上讲什么\n",
      "\n",
      "87\n",
      "00:03:56,000 --> 00:03:57,000\n",
      "就去想一想\n",
      "\n",
      "88\n",
      "00:03:57,000 --> 00:04:00,000\n",
      "那个到底在说什么\n",
      "\n",
      "89\n",
      "00:04:00,000 --> 00:04:03,000\n",
      "你要真的了解他在说什么的时候\n",
      "\n",
      "90\n",
      "00:04:03,000 --> 00:04:06,000\n",
      "你就用了很多思考的功夫\n",
      "\n",
      "91\n",
      "00:04:06,000 --> 00:04:10,000\n",
      "你就在练习自己思考的能力了\n",
      "\n",
      "92\n",
      "00:04:10,000 --> 00:04:15,000\n",
      "好 以上说的是课业内的部分\n",
      "\n",
      "93\n",
      "00:04:15,000 --> 00:04:18,000\n",
      "那当然除了课业内之外\n",
      "\n",
      "94\n",
      "00:04:18,000 --> 00:04:21,000\n",
      "还有一大堆是不在课业内的\n",
      "\n",
      "95\n",
      "00:04:21,000 --> 00:04:25,000\n",
      "那就是课业外的\n",
      "\n",
      "96\n",
      "00:04:25,000 --> 00:04:27,000\n",
      "课业外也有很多式的\n",
      "\n",
      "97\n",
      "00:04:27,000 --> 00:04:33,000\n",
      "那我们可以举例来说\n",
      "\n",
      "98\n",
      "00:04:39,000 --> 00:04:41,000\n",
      "课业外有什么可以学习的\n",
      "\n",
      "99\n",
      "00:04:41,000 --> 00:04:45,000\n",
      "那我通常把学习定义成为\n",
      "\n",
      "100\n",
      "00:04:45,000 --> 00:04:47,000\n",
      "什么是学习\n",
      "\n",
      "101\n",
      "00:04:47,000 --> 00:04:50,000\n",
      "学习就是一种增长\n",
      "\n",
      "102\n",
      "00:04:53,000 --> 00:04:56,000\n",
      "一种进步\n",
      "\n",
      "103\n",
      "00:04:57,000 --> 00:05:01,000\n",
      "然后获得快乐\n",
      "\n",
      "104\n",
      "00:05:01,000 --> 00:05:03,000\n",
      "这就是学习\n",
      "\n",
      "105\n",
      "00:05:03,000 --> 00:05:07,000\n",
      "所以即使是课业外的任何事情\n",
      "\n",
      "106\n",
      "00:05:07,000 --> 00:05:09,000\n",
      "只要你觉得是有增长的\n",
      "\n",
      "107\n",
      "00:05:09,000 --> 00:05:11,000\n",
      "是有进步的\n",
      "\n",
      "108\n",
      "00:05:11,000 --> 00:05:12,000\n",
      "让你觉得快乐的\n",
      "\n",
      "109\n",
      "00:05:12,000 --> 00:05:16,000\n",
      "那应该就是值得学习的地方\n",
      "\n",
      "110\n",
      "00:05:16,000 --> 00:05:19,000\n",
      "那我们可以举很多例子\n",
      "\n",
      "111\n",
      "00:05:19,000 --> 00:05:21,000\n",
      "那譬如说\n",
      "\n",
      "112\n",
      "00:05:21,000 --> 00:05:23,000\n",
      "很多同学喜欢打球\n",
      "\n",
      "113\n",
      "00:05:23,000 --> 00:05:26,000\n",
      "打球是不是学习\n",
      "\n",
      "114\n",
      "00:05:26,000 --> 00:05:28,000\n",
      "当然是\n",
      "\n",
      "115\n",
      "00:05:28,000 --> 00:05:30,000\n",
      "打球中间有没有增长\n",
      "\n",
      "116\n",
      "00:05:30,000 --> 00:05:32,000\n",
      "当然有增长\n",
      "\n",
      "117\n",
      "00:05:32,000 --> 00:05:35,000\n",
      "打球不只是对健康有增长\n",
      "\n",
      "118\n",
      "00:05:35,000 --> 00:05:37,000\n",
      "而且可能对于\n",
      "\n",
      "119\n",
      "00:05:37,000 --> 00:05:39,000\n",
      "比如说手脑协调\n",
      "\n",
      "120\n",
      "00:05:39,000 --> 00:05:41,000\n",
      "譬如说团队精神\n",
      "\n",
      "121\n",
      "00:05:41,000 --> 00:05:44,000\n",
      "譬如说个人之间的互动\n",
      "\n",
      "122\n",
      "00:05:44,000 --> 00:05:46,000\n",
      "可能都有帮助\n",
      "\n",
      "123\n",
      "00:05:46,000 --> 00:05:48,000\n",
      "所以打球当然是有增长的\n",
      "\n",
      "124\n",
      "00:05:48,000 --> 00:05:50,000\n",
      "那当然是很好的学习的机会\n",
      "\n",
      "125\n",
      "00:05:50,000 --> 00:05:52,000\n",
      "有人喜欢爬山\n",
      "\n",
      "126\n",
      "00:05:52,000 --> 00:05:54,000\n",
      "爬山是不是好的学习机会\n",
      "\n",
      "127\n",
      "00:05:54,000 --> 00:05:55,000\n",
      "当然是\n",
      "\n",
      "128\n",
      "00:05:55,000 --> 00:05:57,000\n",
      "这个我以前两年前就讲过很多\n",
      "\n",
      "129\n",
      "00:05:57,000 --> 00:05:59,000\n",
      "爬山可以学到很多的\n",
      "\n",
      "130\n",
      "00:05:59,000 --> 00:06:02,000\n",
      "那爬山当然是一种学习\n",
      "\n",
      "131\n",
      "00:06:02,000 --> 00:06:04,000\n",
      "有人说我不喜欢爬山\n",
      "\n",
      "132\n",
      "00:06:04,000 --> 00:06:05,000\n",
      "我去旅行好不好\n",
      "\n",
      "133\n",
      "00:06:05,000 --> 00:06:07,000\n",
      "旅行当然好\n",
      "\n",
      "134\n",
      "00:06:07,000 --> 00:06:09,000\n",
      "旅行可以增长建设\n",
      "\n",
      "135\n",
      "00:06:09,000 --> 00:06:11,000\n",
      "可以扩增事业\n",
      "\n",
      "136\n",
      "00:06:11,000 --> 00:06:13,000\n",
      "可以增加很多很多\n",
      "\n",
      "137\n",
      "00:06:13,000 --> 00:06:14,000\n",
      "当然是有进步的\n",
      "\n",
      "138\n",
      "00:06:14,000 --> 00:06:16,000\n",
      "所以当然是很好的学习\n",
      "\n",
      "139\n",
      "00:06:16,000 --> 00:06:18,000\n",
      "你凡是获得快乐\n",
      "\n",
      "140\n",
      "00:06:18,000 --> 00:06:20,000\n",
      "都是很好的事\n",
      "\n",
      "141\n",
      "00:06:20,000 --> 00:06:23,000\n",
      "那这些都值得下功夫去\n",
      "\n",
      "142\n",
      "00:06:23,000 --> 00:06:25,000\n",
      "把它看成是学习\n",
      "\n",
      "143\n",
      "00:06:25,000 --> 00:06:27,000\n",
      "都值得下功夫去做的\n",
      "\n",
      "144\n",
      "00:06:27,000 --> 00:06:29,000\n",
      "我们再讲另外一系列\n",
      "\n",
      "145\n",
      "00:06:29,000 --> 00:06:30,000\n",
      "譬如说\n",
      "\n",
      "146\n",
      "00:06:30,000 --> 00:06:34,000\n",
      "有人说谈恋爱是不是学习\n",
      "\n",
      "147\n",
      "00:06:34,000 --> 00:06:37,000\n",
      "谈恋爱除了你在谈恋爱上\n",
      "\n",
      "148\n",
      "00:06:37,000 --> 00:06:39,000\n",
      "会有收获以外\n",
      "\n",
      "149\n",
      "00:06:39,000 --> 00:06:41,000\n",
      "本身也是有收获的\n",
      "\n",
      "150\n",
      "00:06:41,000 --> 00:06:44,000\n",
      "因为让你体验到人跟人之间\n",
      "\n",
      "151\n",
      "00:06:44,000 --> 00:06:45,000\n",
      "的各种感觉\n",
      "\n",
      "152\n",
      "00:06:45,000 --> 00:06:48,000\n",
      "人跟人之间的各种期待等等\n",
      "\n",
      "153\n",
      "00:06:48,000 --> 00:06:50,000\n",
      "有没有帮助\n",
      "\n",
      "154\n",
      "00:06:50,000 --> 00:06:52,000\n",
      "当然有帮助\n",
      "\n",
      "155\n",
      "00:06:52,000 --> 00:06:54,000\n",
      "对每一个人都是很好的学习\n",
      "\n",
      "156\n",
      "00:06:54,000 --> 00:06:57,000\n",
      "所以谈恋爱当然是一件很好的事\n",
      "\n",
      "157\n",
      "00:06:57,000 --> 00:07:00,000\n",
      "有人会说那要靠缘分\n",
      "\n",
      "158\n",
      "00:07:00,000 --> 00:07:02,000\n",
      "没有缘分没有办法\n",
      "\n",
      "159\n",
      "00:07:02,000 --> 00:07:03,000\n",
      "对不对\n",
      "\n",
      "160\n",
      "00:07:03,000 --> 00:07:06,000\n",
      "但是你不是一定要谈恋爱吗\n",
      "\n",
      "161\n",
      "00:07:06,000 --> 00:07:07,000\n",
      "你可以交朋友\n",
      "\n",
      "162\n",
      "00:07:07,000 --> 00:07:10,000\n",
      "交朋友是不是学习\n",
      "\n",
      "163\n",
      "00:07:10,000 --> 00:07:11,000\n",
      "当然是\n",
      "\n",
      "164\n",
      "00:07:11,000 --> 00:07:13,000\n",
      "交朋友也一样\n",
      "\n",
      "165\n",
      "00:07:13,000 --> 00:07:16,000\n",
      "让我们学到很多人际的互动\n",
      "\n",
      "166\n",
      "00:07:16,000 --> 00:07:20,000\n",
      "学到很多人跟人之间的沟通\n",
      "\n",
      "167\n",
      "00:07:20,000 --> 00:07:22,000\n",
      "人跟人之间的期待\n",
      "\n",
      "168\n",
      "00:07:22,000 --> 00:07:24,000\n",
      "人跟人之间的感觉\n",
      "\n",
      "169\n",
      "00:07:24,000 --> 00:07:26,000\n",
      "这都是交朋友之后学到的\n",
      "\n",
      "170\n",
      "00:07:26,000 --> 00:07:28,000\n",
      "对我们电机系的同学而言\n",
      "\n",
      "171\n",
      "00:07:28,000 --> 00:07:31,000\n",
      "你四周有一大群好同学\n",
      "\n",
      "172\n",
      "00:07:31,000 --> 00:07:34,000\n",
      "都是很好的交朋友的对象\n",
      "\n",
      "173\n",
      "00:07:34,000 --> 00:07:36,000\n",
      "你下一番功夫交朋友好不好\n",
      "\n",
      "174\n",
      "00:07:36,000 --> 00:07:37,000\n",
      "好\n",
      "\n",
      "175\n",
      "00:07:37,000 --> 00:07:40,000\n",
      "当然是有帮助的\n",
      "\n",
      "176\n",
      "00:07:40,000 --> 00:07:42,000\n",
      "另外当然我们可以举很多\n",
      "\n",
      "177\n",
      "00:07:42,000 --> 00:07:44,000\n",
      "我们最现成的例子\n",
      "\n",
      "178\n",
      "00:07:44,000 --> 00:07:47,000\n",
      "譬如说我们的系学会办各种活动\n",
      "\n",
      "179\n",
      "00:07:47,000 --> 00:07:49,000\n",
      "那些活动有没有帮助\n",
      "\n",
      "180\n",
      "00:07:49,000 --> 00:07:50,000\n",
      "当然有\n",
      "\n",
      "181\n",
      "00:07:50,000 --> 00:07:54,000\n",
      "我们举例来讲电业\n",
      "\n",
      "182\n",
      "00:07:54,000 --> 00:07:57,000\n",
      "你如果去参加某一个舞跳个舞\n",
      "\n",
      "183\n",
      "00:07:57,000 --> 00:08:00,000\n",
      "或者参加某个剧演个剧\n",
      "\n",
      "184\n",
      "00:08:00,000 --> 00:08:01,000\n",
      "有没有帮助\n",
      "\n",
      "185\n",
      "00:08:01,000 --> 00:08:02,000\n",
      "当然有帮助\n",
      "\n",
      "186\n",
      "00:08:02,000 --> 00:08:05,000\n",
      "你在这中间一定发现有所增长\n",
      "\n",
      "187\n",
      "00:08:05,000 --> 00:08:06,000\n",
      "有所进步\n",
      "\n",
      "188\n",
      "00:08:06,000 --> 00:08:08,000\n",
      "那是为什么有那么多同学\n",
      "\n",
      "189\n",
      "00:08:08,000 --> 00:08:09,000\n",
      "要去参加\n",
      "\n",
      "190\n",
      "00:08:09,000 --> 00:08:11,000\n",
      "就是因为发现那个确实是有增长\n",
      "\n",
      "191\n",
      "00:08:11,000 --> 00:08:13,000\n",
      "有进步\n",
      "\n",
      "192\n",
      "00:08:13,000 --> 00:08:15,000\n",
      "有的人说\n",
      "\n",
      "193\n",
      "00:08:15,000 --> 00:08:17,000\n",
      "我不去跳那个舞\n",
      "\n",
      "194\n",
      "00:08:17,000 --> 00:08:20,000\n",
      "或者演那个剧\n",
      "\n",
      "195\n",
      "00:08:20,000 --> 00:08:22,000\n",
      "我做幕后的\n",
      "\n",
      "196\n",
      "00:08:22,000 --> 00:08:25,000\n",
      "譬如说是幕后的什么什么规划\n",
      "\n",
      "197\n",
      "00:08:25,000 --> 00:08:29,000\n",
      "或者说是什么这个光舞的什么软体组\n",
      "\n",
      "198\n",
      "00:08:29,000 --> 00:08:32,000\n",
      "还是什么服装道具组\n",
      "\n",
      "199\n",
      "00:08:32,000 --> 00:08:33,000\n",
      "一样\n",
      "\n",
      "200\n",
      "00:08:33,000 --> 00:08:36,000\n",
      "那个都是可以获得很多的增长\n",
      "\n",
      "201\n",
      "00:08:36,000 --> 00:08:37,000\n",
      "很多进步的\n",
      "\n",
      "202\n",
      "00:08:37,000 --> 00:08:39,000\n",
      "当然都是很有用的\n",
      "\n",
      "203\n",
      "00:08:39,000 --> 00:08:42,000\n",
      "都是很好的学习\n",
      "\n",
      "204\n",
      "00:08:42,000 --> 00:08:47,000\n",
      "那当然也包括电业以外的戏学会\n",
      "\n",
      "205\n",
      "00:08:47,000 --> 00:08:50,000\n",
      "其他的各种活动都一样\n",
      "\n",
      "206\n",
      "00:08:50,000 --> 00:08:55,000\n",
      "也包括电机系以外的其他的校内\n",
      "\n",
      "207\n",
      "00:08:55,000 --> 00:08:59,000\n",
      "或者校外的各种活动几乎都一样\n",
      "\n",
      "208\n",
      "00:08:59,000 --> 00:09:01,000\n",
      "都可以让人有所增长\n",
      "\n",
      "209\n",
      "00:09:01,000 --> 00:09:02,000\n",
      "有所进步\n",
      "\n",
      "210\n",
      "00:09:02,000 --> 00:09:04,000\n",
      "都是很好的学习的机会\n",
      "\n",
      "211\n",
      "00:09:04,000 --> 00:09:06,000\n",
      "都是很好的学习\n",
      "\n",
      "212\n",
      "00:09:06,000 --> 00:09:08,000\n",
      "同样的问题是\n",
      "\n",
      "213\n",
      "00:09:08,000 --> 00:09:11,000\n",
      "这些东西都没有考试\n",
      "\n",
      "214\n",
      "00:09:11,000 --> 00:09:12,000\n",
      "没有成绩\n",
      "\n",
      "215\n",
      "00:09:12,000 --> 00:09:14,000\n",
      "不能显示在成绩单上\n",
      "\n",
      "216\n",
      "00:09:14,000 --> 00:09:18,000\n",
      "因此对有一些同学会认为那个浪费时间\n",
      "\n",
      "217\n",
      "00:09:18,000 --> 00:09:20,000\n",
      "我不需要花时间去做那个\n",
      "\n",
      "218\n",
      "00:09:20,000 --> 00:09:22,000\n",
      "因为不影响我的\n",
      "\n",
      "219\n",
      "00:09:22,000 --> 00:09:24,000\n",
      "overfitting的目标\n",
      "\n",
      "220\n",
      "00:09:24,000 --> 00:09:26,000\n",
      "里面没有这个嘛\n",
      "\n",
      "221\n",
      "00:09:26,000 --> 00:09:29,000\n",
      "具体成绩没有这些嘛\n",
      "\n",
      "222\n",
      "00:09:29,000 --> 00:09:30,000\n",
      "那不要这样想\n",
      "\n",
      "223\n",
      "00:09:30,000 --> 00:09:33,000\n",
      "因为那些都非常的重要\n",
      "\n",
      "224\n",
      "00:09:33,000 --> 00:09:36,000\n",
      "都对你发展非常的重要\n",
      "\n",
      "225\n",
      "00:09:36,000 --> 00:09:39,000\n",
      "那我们说电机工程\n",
      "\n",
      "226\n",
      "00:09:39,000 --> 00:09:41,000\n",
      "今天的电机工程\n",
      "\n",
      "227\n",
      "00:09:41,000 --> 00:09:44,000\n",
      "很少什么事情自己一个人可以做成功的\n",
      "\n",
      "228\n",
      "00:09:44,000 --> 00:09:47,000\n",
      "你必须跟很多人一起\n",
      "\n",
      "229\n",
      "00:09:47,000 --> 00:09:50,000\n",
      "才可能做成功一个非常重要的\n",
      "\n",
      "230\n",
      "00:09:50,000 --> 00:09:52,000\n",
      "有意义的工作\n",
      "\n",
      "231\n",
      "00:09:52,000 --> 00:09:55,000\n",
      "那当你跟一群人在一起做的时候\n",
      "\n",
      "232\n",
      "00:09:55,000 --> 00:09:59,000\n",
      "你必须学会如何进入一个团队\n",
      "\n",
      "233\n",
      "00:09:59,000 --> 00:10:03,000\n",
      "从边缘开始慢慢进入核心\n",
      "\n",
      "234\n",
      "00:10:03,000 --> 00:10:06,000\n",
      "从底层开始慢慢变成leader\n",
      "\n",
      "235\n",
      "00:10:06,000 --> 00:10:10,000\n",
      "然后如何可以推动你想做的事\n",
      "\n",
      "236\n",
      "00:10:10,000 --> 00:10:14,000\n",
      "如何变成可以做到你想做的事等等\n",
      "\n",
      "237\n",
      "00:10:14,000 --> 00:10:16,000\n",
      "这些都是很重要的\n",
      "\n",
      "238\n",
      "00:10:16,000 --> 00:10:18,000\n",
      "那我们通常称这些东西\n",
      "\n",
      "239\n",
      "00:10:18,000 --> 00:10:21,000\n",
      "是所谓的soft skills\n",
      "\n",
      "240\n",
      "00:10:21,000 --> 00:10:25,000\n",
      "也就是软实力\n",
      "\n",
      "241\n",
      "00:10:25,000 --> 00:10:31,000\n",
      "所谓软实力就是硬实力以外的软实力\n",
      "\n",
      "242\n",
      "00:10:31,000 --> 00:10:34,000\n",
      "那硬实力是说你电子学的功力\n",
      "\n",
      "243\n",
      "00:10:34,000 --> 00:10:36,000\n",
      "数学的功力\n",
      "\n",
      "244\n",
      "00:10:36,000 --> 00:10:40,000\n",
      "这个城市能力这种是硬实力\n",
      "\n",
      "245\n",
      "00:10:40,000 --> 00:10:45,000\n",
      "软实力我们主要就是讲各种人际之间的\n",
      "\n",
      "246\n",
      "00:10:45,000 --> 00:10:49,000\n",
      "在人跟人之间的各种能力\n",
      "\n",
      "247\n",
      "00:10:49,000 --> 00:10:52,000\n",
      "包括沟通能力协调能力\n",
      "\n",
      "248\n",
      "00:10:52,000 --> 00:10:53,000\n",
      "交朋友的能力\n",
      "\n",
      "249\n",
      "00:10:53,000 --> 00:10:55,000\n",
      "这个说服人的能力\n",
      "\n",
      "250\n",
      "00:10:55,000 --> 00:10:59,000\n",
      "这个团队精神领导能力等等\n",
      "\n",
      "251\n",
      "00:10:59,000 --> 00:11:02,000\n",
      "那那些就是所谓的soft skills\n",
      "\n",
      "252\n",
      "00:11:02,000 --> 00:11:03,000\n",
      "重要不重要\n",
      "\n",
      "253\n",
      "00:11:03,000 --> 00:11:04,000\n",
      "重要\n",
      "\n",
      "254\n",
      "00:11:04,000 --> 00:11:07,000\n",
      "你看到任何一个成功的电机工程师\n",
      "\n",
      "255\n",
      "00:11:07,000 --> 00:11:09,000\n",
      "他都有一堆这种\n",
      "\n",
      "256\n",
      "00:11:09,000 --> 00:11:12,000\n",
      "这个才是他成功的一个非常重要的关键\n",
      "\n",
      "257\n",
      "00:11:12,000 --> 00:11:15,000\n",
      "这种东西怎么来\n",
      "\n",
      "258\n",
      "00:11:15,000 --> 00:11:20,000\n",
      "我们刚才讲的各种课业外的各种学习增长机会\n",
      "\n",
      "259\n",
      "00:11:20,000 --> 00:11:26,000\n",
      "都可以帮助一个人塑造他的soft skills\n",
      "\n",
      "260\n",
      "00:11:26,000 --> 00:11:30,000\n",
      "是有少数人的这些soft skills是天生的\n",
      "\n",
      "261\n",
      "00:11:30,000 --> 00:11:31,000\n",
      "他天生就厉害\n",
      "\n",
      "262\n",
      "00:11:31,000 --> 00:11:32,000\n",
      "有没有\n",
      "\n",
      "263\n",
      "00:11:32,000 --> 00:11:33,000\n",
      "有\n",
      "\n",
      "264\n",
      "00:11:33,000 --> 00:11:35,000\n",
      "但这种人毕竟没那么多\n",
      "\n",
      "265\n",
      "00:11:35,000 --> 00:11:37,000\n",
      "对很多人而言\n",
      "\n",
      "266\n",
      "00:11:37,000 --> 00:11:42,000\n",
      "他的soft skills是自己努力慢慢培养起来的\n",
      "\n",
      "267\n",
      "00:11:42,000 --> 00:11:45,000\n",
      "我刚才一开始前面讲的那一段\n",
      "\n",
      "268\n",
      "00:11:45,000 --> 00:11:49,000\n",
      "我说我在这个进台大电机系以前\n",
      "\n",
      "269\n",
      "00:11:49,000 --> 00:11:51,000\n",
      "我几乎不会交朋友\n",
      "\n",
      "270\n",
      "00:11:51,000 --> 00:11:53,000\n",
      "我不太会说话\n",
      "\n",
      "271\n",
      "00:11:53,000 --> 00:11:57,000\n",
      "我在读大学的四年里面改变我自己\n",
      "\n",
      "272\n",
      "00:11:57,000 --> 00:12:02,000\n",
      "让我变成有很多这方面的能力的人\n",
      "\n",
      "273\n",
      "00:12:02,000 --> 00:12:06,000\n",
      "其实最重要就是我的很多soft skills\n",
      "\n",
      "274\n",
      "00:12:06,000 --> 00:12:07,000\n",
      "都是我自己培养\n",
      "\n",
      "275\n",
      "00:12:07,000 --> 00:12:10,000\n",
      "在读台大电机系的四年里面\n",
      "\n",
      "276\n",
      "00:12:10,000 --> 00:12:15,000\n",
      "获得的非常多这方面的收获的\n",
      "\n",
      "277\n",
      "00:12:15,000 --> 00:12:21,000\n",
      "那是为什么我每次都要强调这个东西有多么重要\n",
      "\n",
      "278\n",
      "00:12:21,000 --> 00:12:29,000\n",
      "那我之前曾经在几年前的这个信号与人生里面\n",
      "\n",
      "279\n",
      "00:12:29,000 --> 00:12:31,000\n",
      "有说到这一件事\n",
      "\n",
      "280\n",
      "00:12:31,000 --> 00:12:33,000\n",
      "我现在不要重复\n",
      "\n",
      "281\n",
      "00:12:33,000 --> 00:12:36,000\n",
      "但是我简单的summarize\n",
      "\n",
      "282\n",
      "00:12:36,000 --> 00:12:39,000\n",
      "那我说我们电机系的\n",
      "\n",
      "283\n",
      "00:12:39,000 --> 00:12:45,000\n",
      "电机工程师的一生career的发展\n",
      "\n",
      "284\n",
      "00:12:45,000 --> 00:12:48,000\n",
      "那黄金实在是在什么时候\n",
      "\n",
      "285\n",
      "00:12:48,000 --> 00:12:52,000\n",
      "我认为是在35岁到55岁\n",
      "\n",
      "286\n",
      "00:12:52,000 --> 00:12:56,000\n",
      "这20年是我们的黄金时代\n",
      "\n",
      "287\n",
      "00:12:56,000 --> 00:12:58,000\n",
      "在这以前当然更好\n",
      "\n",
      "288\n",
      "00:12:58,000 --> 00:13:01,000\n",
      "只是说可能各方面尚未具备\n",
      "\n",
      "289\n",
      "00:13:01,000 --> 00:13:03,000\n",
      "还没有完全训练的好\n",
      "\n",
      "290\n",
      "00:13:03,000 --> 00:13:05,000\n",
      "在这以后是最好的\n",
      "\n",
      "291\n",
      "00:13:05,000 --> 00:13:10,000\n",
      "这以后年纪大了难免有一些要打个折扣等等\n",
      "\n",
      "292\n",
      "00:13:10,000 --> 00:13:15,000\n",
      "就这里面我们看到我们的电机系的毕业的同学\n",
      "\n",
      "293\n",
      "00:13:15,000 --> 00:13:18,000\n",
      "过去有几千人毕业我都看到\n",
      "\n",
      "294\n",
      "00:13:18,000 --> 00:13:22,000\n",
      "那我觉得有的人的发展是像这样\n",
      "\n",
      "295\n",
      "00:13:22,000 --> 00:13:24,000\n",
      "有一定的斜率\n",
      "\n",
      "296\n",
      "00:13:24,000 --> 00:13:28,000\n",
      "但到某一个阶段它会慢慢saturate\n",
      "\n",
      "297\n",
      "00:13:28,000 --> 00:13:31,000\n",
      "有的人也许开始向上比较晚\n",
      "\n",
      "298\n",
      "00:13:31,000 --> 00:13:33,000\n",
      "但它斜率比较高\n",
      "\n",
      "299\n",
      "00:13:33,000 --> 00:13:38,000\n",
      "它最后会saturate在比较高的地方\n",
      "\n",
      "300\n",
      "00:13:38,000 --> 00:13:41,000\n",
      "也有的人也许开始的比较快\n",
      "\n",
      "301\n",
      "00:13:41,000 --> 00:13:46,000\n",
      "但是后来会overshoot之后收敛在比较低的地方等等\n",
      "\n",
      "302\n",
      "00:13:46,000 --> 00:13:47,000\n",
      "这个每一个人都不一样\n",
      "\n",
      "303\n",
      "00:13:47,000 --> 00:13:49,000\n",
      "但是当然也有一种人\n",
      "\n",
      "304\n",
      "00:13:49,000 --> 00:13:53,000\n",
      "你会看到它一直向上走\n",
      "\n",
      "305\n",
      "00:13:53,000 --> 00:13:57,000\n",
      "完全没有saturate\n",
      "\n",
      "306\n",
      "00:13:57,000 --> 00:14:00,000\n",
      "这些人这些差别在哪里\n",
      "\n",
      "307\n",
      "00:14:00,000 --> 00:14:03,000\n",
      "这些东西差别在哪里\n",
      "\n",
      "308\n",
      "00:14:03,000 --> 00:14:05,000\n",
      "我以前已经说过这件事\n",
      "\n",
      "309\n",
      "00:14:05,000 --> 00:14:07,000\n",
      "我不要多重复\n",
      "\n",
      "310\n",
      "00:14:07,000 --> 00:14:10,000\n",
      "我说最主要因素有四个\n",
      "\n",
      "311\n",
      "00:14:10,000 --> 00:14:12,000\n",
      "就是实力\n",
      "\n",
      "312\n",
      "00:14:12,000 --> 00:14:16,000\n",
      "努力\n",
      "\n",
      "313\n",
      "00:14:16,000 --> 00:14:20,000\n",
      "大致\n",
      "\n",
      "314\n",
      "00:14:20,000 --> 00:14:23,000\n",
      "跟self skill这四件事情\n",
      "\n",
      "315\n",
      "00:14:23,000 --> 00:14:26,000\n",
      "那我认为\n",
      "\n",
      "316\n",
      "00:14:26,000 --> 00:14:29,000\n",
      "真正影响这个的\n",
      "\n",
      "317\n",
      "00:14:29,000 --> 00:14:32,000\n",
      "不是因为电子学考得好不好\n",
      "\n",
      "318\n",
      "00:14:32,000 --> 00:14:35,000\n",
      "不是因为信号与系统念得好不好\n",
      "\n",
      "319\n",
      "00:14:35,000 --> 00:14:37,000\n",
      "也就我刚才讲\n",
      "\n",
      "320\n",
      "00:14:37,000 --> 00:14:41,000\n",
      "你把每一门必修课当成是单一跑道\n",
      "\n",
      "321\n",
      "00:14:41,000 --> 00:14:43,000\n",
      "跑到第一名并不表示怎样\n",
      "\n",
      "322\n",
      "00:14:43,000 --> 00:14:45,000\n",
      "我们最后不看那个的\n",
      "\n",
      "323\n",
      "00:14:45,000 --> 00:14:47,000\n",
      "最后看的是这个\n",
      "\n",
      "324\n",
      "00:14:47,000 --> 00:14:49,000\n",
      "那这个是怎么样影响\n",
      "\n",
      "325\n",
      "00:14:49,000 --> 00:14:51,000\n",
      "我认为是这四件事\n",
      "\n",
      "326\n",
      "00:14:51,000 --> 00:14:53,000\n",
      "就是实力努力\n",
      "\n",
      "327\n",
      "00:14:53,000 --> 00:14:56,000\n",
      "大致跟self skills\n",
      "\n",
      "328\n",
      "00:14:56,000 --> 00:14:59,000\n",
      "这四件事里面我们现在可以summarize\n",
      "\n",
      "329\n",
      "00:14:59,000 --> 00:15:01,000\n",
      "我刚才讲的\n",
      "\n",
      "330\n",
      "00:15:01,000 --> 00:15:03,000\n",
      "什么是实力\n",
      "\n",
      "331\n",
      "00:15:03,000 --> 00:15:06,000\n",
      "实力就是所有的这些我们电机工程的\n",
      "\n",
      "332\n",
      "00:15:06,000 --> 00:15:09,000\n",
      "专业领域里面的各种东西的实力\n",
      "\n",
      "333\n",
      "00:15:09,000 --> 00:15:11,000\n",
      "实力怎么厉害法\n",
      "\n",
      "334\n",
      "00:15:11,000 --> 00:15:13,000\n",
      "就是我刚才讲的\n",
      "\n",
      "335\n",
      "00:15:13,000 --> 00:15:16,000\n",
      "你如果都是在做全面的学习的话\n",
      "\n",
      "336\n",
      "00:15:16,000 --> 00:15:18,000\n",
      "你就会学到各种该学到的\n",
      "\n",
      "337\n",
      "00:15:18,000 --> 00:15:20,000\n",
      "最后你的实力就是很强的\n",
      "\n",
      "338\n",
      "00:15:20,000 --> 00:15:25,000\n",
      "所以实力最主要就是不要overfitting\n",
      "\n",
      "339\n",
      "00:15:25,000 --> 00:15:31,000\n",
      "要尽量都做学到该学的全面的学习\n",
      "\n",
      "340\n",
      "00:15:31,000 --> 00:15:33,000\n",
      "努力是没有疑问\n",
      "\n",
      "341\n",
      "00:15:33,000 --> 00:15:34,000\n",
      "每个人都了解\n",
      "\n",
      "342\n",
      "00:15:34,000 --> 00:15:38,000\n",
      "确实我们可以看到一个人在未来的几十年里面\n",
      "\n",
      "343\n",
      "00:15:38,000 --> 00:15:40,000\n",
      "有的人他一直努力\n",
      "\n",
      "344\n",
      "00:15:40,000 --> 00:15:42,000\n",
      "有的人慢慢不太努力等等\n",
      "\n",
      "345\n",
      "00:15:42,000 --> 00:15:44,000\n",
      "这个是有明显差别的\n",
      "\n",
      "346\n",
      "00:15:44,000 --> 00:15:46,000\n",
      "那self skills我刚才已经讲了\n",
      "\n",
      "347\n",
      "00:15:46,000 --> 00:15:50,000\n",
      "就是很多我们平常没有算成绩\n",
      "\n",
      "348\n",
      "00:15:50,000 --> 00:15:52,000\n",
      "觉得大家不重视的事情\n",
      "\n",
      "349\n",
      "00:15:52,000 --> 00:15:54,000\n",
      "那其他常常是很重要的\n",
      "\n",
      "350\n",
      "00:15:54,000 --> 00:15:56,000\n",
      "你如果好好的\n",
      "\n",
      "351\n",
      "00:15:56,000 --> 00:15:59,000\n",
      "多在各种课业外的事情上\n",
      "\n",
      "352\n",
      "00:15:59,000 --> 00:16:01,000\n",
      "增长进步的话\n",
      "\n",
      "353\n",
      "00:16:01,000 --> 00:16:03,000\n",
      "你这些东西会很强\n",
      "\n",
      "354\n",
      "00:16:03,000 --> 00:16:05,000\n",
      "这个东西会很厉害的\n",
      "\n",
      "355\n",
      "00:16:05,000 --> 00:16:07,000\n",
      "当然对少数人而言\n",
      "\n",
      "356\n",
      "00:16:07,000 --> 00:16:08,000\n",
      "他天生就有\n",
      "\n",
      "357\n",
      "00:16:08,000 --> 00:16:09,000\n",
      "他可能不需要\n",
      "\n",
      "358\n",
      "00:16:09,000 --> 00:16:11,000\n",
      "就是每一个人不一样的\n",
      "\n",
      "359\n",
      "00:16:11,000 --> 00:16:13,000\n",
      "这三个我都提过了\n",
      "\n",
      "360\n",
      "00:16:13,000 --> 00:16:15,000\n",
      "那么大致我还没有提\n",
      "\n",
      "361\n",
      "00:16:15,000 --> 00:16:18,000\n",
      "其实大致没有什么要特别说的\n",
      "\n",
      "362\n",
      "00:16:18,000 --> 00:16:21,000\n",
      "那应该就是我刚才前面有讲过\n",
      "\n",
      "363\n",
      "00:16:21,000 --> 00:16:26,000\n",
      "就是每一个人可以有你自己的长程目标\n",
      "\n",
      "364\n",
      "00:16:26,000 --> 00:16:29,000\n",
      "那有的人本来就有了\n",
      "\n",
      "365\n",
      "00:16:29,000 --> 00:16:32,000\n",
      "有的人也许我平常没有想过\n",
      "\n",
      "366\n",
      "00:16:32,000 --> 00:16:37,000\n",
      "那你可以在适当时机开始想\n",
      "\n",
      "367\n",
      "00:16:37,000 --> 00:16:40,000\n",
      "我有没有想要做什么事情\n",
      "\n",
      "368\n",
      "00:16:40,000 --> 00:16:43,000\n",
      "哪些事情可能是我的长程目标\n",
      "\n",
      "369\n",
      "00:16:43,000 --> 00:16:48,000\n",
      "我希望最后让我花个5年10年\n",
      "\n",
      "370\n",
      "00:16:48,000 --> 00:16:50,000\n",
      "15年或者更长\n",
      "\n",
      "371\n",
      "00:16:50,000 --> 00:16:53,000\n",
      "我把我的很多的努力都来\n",
      "\n",
      "372\n",
      "00:16:53,000 --> 00:16:55,000\n",
      "把某一些事情做得非常漂亮\n",
      "\n",
      "373\n",
      "00:16:55,000 --> 00:16:57,000\n",
      "那是我很想做的事\n",
      "\n",
      "374\n",
      "00:16:57,000 --> 00:17:00,000\n",
      "那就是长程目标\n",
      "\n",
      "375\n",
      "00:17:00,000 --> 00:17:02,000\n",
      "如果我觉得做那些事情\n",
      "\n",
      "376\n",
      "00:17:02,000 --> 00:17:04,000\n",
      "会让我非常的\n",
      "\n",
      "377\n",
      "00:17:04,000 --> 00:17:06,000\n",
      "觉得有意义\n",
      "\n",
      "378\n",
      "00:17:06,000 --> 00:17:08,000\n",
      "愿意花功夫下去做的\n",
      "\n",
      "379\n",
      "00:17:08,000 --> 00:17:10,000\n",
      "那就是我的长程目标\n",
      "\n",
      "380\n",
      "00:17:10,000 --> 00:17:13,000\n",
      "有的人如果可以想出这个来的话\n",
      "\n",
      "381\n",
      "00:17:13,000 --> 00:17:16,000\n",
      "那就是他的大致\n",
      "\n",
      "382\n",
      "00:17:16,000 --> 00:17:18,000\n",
      "越是有这种大致的人\n",
      "\n",
      "383\n",
      "00:17:18,000 --> 00:17:21,000\n",
      "也比较容易向上冲\n",
      "\n",
      "384\n",
      "00:17:21,000 --> 00:17:22,000\n",
      "那我感觉起来\n",
      "\n",
      "385\n",
      "00:17:22,000 --> 00:17:25,000\n",
      "真正影响的就是这四件事\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# 读取并打印字幕内容\n",
    "with open(output_subtitle_path, 'r', encoding='utf-8') as file:\n",
    "    content = file.read()\n",
    "print(content)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5a7ea77-e0e1-458b-9f92-9c37bfb0bf8e",
   "metadata": {},
   "source": [
    "## 第 3 部分 - 处理自动语音识别的结果\n",
    "\n",
    "### 提取字幕文本"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "285c7d95-991b-4ba4-95e4-d085cd1e6fea",
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_and_save_text(srt_filename, output_filename):\n",
    "    # 读取 SRT 文件\n",
    "    with open(srt_filename, 'r', encoding='utf-8') as file:\n",
    "        content = file.read()\n",
    "\n",
    "    # 去除时间戳和索引\n",
    "    pure_text = re.sub(r'\\d+\\n\\d{2}:\\d{2}:\\d{2},\\d{3} --> \\d{2}:\\d{2}:\\d{2},\\d{3}\\n', '', content)\n",
    "    pure_text = re.sub(r'\\n\\n+', '\\n', pure_text)\n",
    "\n",
    "    # 保存纯文本\n",
    "    with open(output_filename, 'w', encoding='utf-8') as output_file:\n",
    "        output_file.write(pure_text)\n",
    "\n",
    "    print(f'提取的文本已保存到 {output_filename}')\n",
    "\n",
    "    return pure_text"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd4b801f-e421-4aa6-806b-e0497185ac39",
   "metadata": {},
   "source": [
    "### 拆分文本\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1e2e83e5-a538-4ce6-8187-9379c0db43bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "def chunk_text(text, max_length):\n",
    "    return textwrap.wrap(text, max_length)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63aaaa7f-f343-45d8-aa01-bfd621011e46",
   "metadata": {},
   "source": [
    "### 执行文本处理\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "25b27081-ca3a-4534-a777-3ad7e76b3fa9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "提取的文本已保存到 ./output-信号与人生.txt\n"
     ]
    }
   ],
   "source": [
    "# 文本块长度\n",
    "chunk_length = 512\n",
    "\n",
    "# 是否转换为繁体中文\n",
    "convert_to_traditional = False\n",
    "\n",
    "# 提取文本并拆分\n",
    "pure_text = extract_and_save_text(\n",
    "    srt_filename=output_subtitle_path,\n",
    "    output_filename=f\"./output-{suffix}.txt\",\n",
    ")\n",
    "\n",
    "chunks = chunk_text(text=pure_text, max_length=chunk_length)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfdefb4d-0d21-4ea5-ab05-4d78810d0ff2",
   "metadata": {},
   "source": [
    "## 第 4 部分 - 文本摘要\n",
    "\n",
    "### 设置 OpenAI API\n",
    "\n",
    "首先，需要设置 OpenAI API 密钥，这里使用的是阿里云的大模型，你可以通过[《00. 大模型 API 获取步骤》](https://github.com/Hoper-J/AI-Guide-and-Demos-zh_CN/blob/master/Guide/00.%20大模型%20API%20获取步骤.md)获取 API 密钥。\n",
    "\n",
    "如果需要使用其他平台，请参考对应的开发文档后对应修改 base_url。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "30f8e9ac-9dfe-436c-af82-b0a49b1889f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# TODO: 设置你的 OPENAI API 密钥，这里以阿里云 DashScope API 为例进行演示\n",
    "OPENAI_API_KEY = \"\"\n",
    "\n",
    "# 不设置则默认使用环境变量\n",
    "if not OPENAI_API_KEY:\n",
    "    OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')\n",
    "\n",
    "# 构建 OpenAI 客户端\n",
    "client = OpenAI(\n",
    "    api_key=openai_api_key,\n",
    "    base_url=\"https://dashscope.aliyuncs.com/compatible-mode/v1\", # 这里使用的是阿里云的大模型，如果需要使用其他平台，请参考对应的开发文档后对应修改。如果使用 GPT 的 API，删除这行就可以直接运行。\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1dbb20a-afb0-448c-992d-6b3ac2a26eb1",
   "metadata": {},
   "source": [
    "### 设置参数\n",
    "\n",
    "默认使用 qwen-turbo，其他模型可以参阅[模型广场 -- 阿里云百炼](https://bailian.console.aliyun.com/?spm=5176.29619931.J__Z58Z6CX7MY__Ll8p1ZOR.1.4d1d59fcWwSqvr#/model-market)，点击对应模型的`查看详情`。\n",
    "\n",
    "![image-20240924091151684](../Guide/assets/image-20240924091151684.png)\n",
    "\n",
    "在界面可以左上角看到对应的英文名称，复制它，然后替换 `model_name`。\n",
    "\n",
    "![image-20240924091414350](../Guide/assets/image-20240924091414350.png)\n",
    "\n",
    "你可以随意更换为你想要的模型，不过可能要先申请使用（通过大概要几个小时，会有短信prompt）。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "65b3f633-9548-4849-b212-f34aa61421da",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 模型名称\n",
    "model_name = 'qwen-turbo'\n",
    "\n",
    "# 控制响应的随机性\n",
    "temperature = 0.0\n",
    "\n",
    "# 控制多样性\n",
    "top_p = 1.0\n",
    "\n",
    "# 最大生成tokens\n",
    "max_tokens = 512"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22064928-8064-4ac7-b506-a2a358a61dc7",
   "metadata": {},
   "source": [
    "### 定义摘要函数\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e891f9d7-2b77-4935-bbe6-6ded3d7b6c1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def summarization(client, summarization_prompt, model_name=\"qwen-turbo\", temperature=0.0, top_p=1.0, max_tokens=512):\n",
    "    response = client.chat.completions.create(\n",
    "        messages=[{\"role\": \"user\", \"content\": summarization_prompt}],\n",
    "        model=model_name,\n",
    "        temperature=temperature,\n",
    "        top_p=top_p,\n",
    "        max_tokens=max_tokens\n",
    "    )\n",
    "    return response.choices[0].message.content"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb17103c-10e5-4184-8d0c-dbb84703ff1c",
   "metadata": {},
   "source": [
    "### 这里演示两种摘要方式\n",
    "\n",
    "分别对应于 `MapReduce` 和 `Refine`，你可以通过接下来的代码来感受二者的区别。\n",
    "\n",
    "#### 方法一：拆分为多段进行摘要（Multi-Stage Summarization）- MapReduce\n",
    "\n",
    "![image.png](../Guide/assets/image-20240924092040340.png)\n",
    "\n",
    "1. 将长文本分成多个较小的部分，并分别获取每个小段落的摘要"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a3a31190-0096-4d82-b2e5-63a07e54c287",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "========== 正在生成第 1 段摘要 ==========\n",
      "\n",
      "原始文本 (第 1 段):\n",
      "每次说 学问是做出来的 什么意思 要做 才会获得学问 你如果每天光是坐在那里听 学问很可能是左耳近右耳出的 你光是坐在那儿读 学问可能从眼睛进入脑海之后就忘掉了 如何能够学问在脑海里面 真的变成你自己学问 就是要做 可能有很多同学有这个经验 你如果去修某一门课 或者做某一个实验 在期末就是要教一个final project 那个final project就是要你把 学到的很多东西 最后整合在你的final project里面 最后做出来的时候 就是把它们都整合了 当你学期结束 真的把final project做完的时候 你会忽然发现 我真的学到很多东西 那就是做出来的学问 也许可以举另外一个例子 就是你如果学了某一些 很复杂的演算法或者什么 好像觉得那些不见得在你的脑海里 可是后来老师出了个习题 那个习题叫你写一个很大的程式 要把所有东西都包进去 当你把这个程式写完的时候 你会发现 你忽然把演算法 所有东西都弄通了 那就是学问是做出来的 所以我们永远要记得 尽量多动手多做 在动手跟做的过程之中 学问才可以变成是自己的 同样的情形就是说 很多时候这样动手或者做的 表现或者成绩 没有一个成绩单上的数字\n",
      "\n",
      "生成的摘要 (第 1 段):\n",
      "这段视频强调了“学问是做出来的”这一观念。其核心信息如下：\n",
      "\n",
      "1. **行动的重要性**：只听和阅读知识无法确保真正理解或内化知识成为个人学问。通过实际操作和实践，才能深入理解并掌握知识。\n",
      "\n",
      "2. **学习与实践的结合**：通过参加课程、实验或完成最终项目等活动，将所学知识与技能应用于实际问题解决中，能有效提升学习效果。\n",
      "\n",
      "3. **整合与应用**：在完成最终项目或大型编程任务时，学生需综合运用所学知识，这一过程有助于知识的深度理解和长期记忆。\n",
      "\n",
      "4. **实例阐述**：以学习复杂算法为例，通过解决实际问题而非仅停留在理论层面，学生能够达到对算法的深入理解和灵活运用。\n",
      "\n",
      "5. **个人化学习体验**：通过动手实践，学生不仅获得知识，还能形成独特的学习体验和技能，远非标准化测试成绩所能完全评价。\n",
      "\n",
      "6. **持续行动的重要性**：强调持续进行实践和创造活动，让学习过程变得生动且有意义，知识因此成为个人的财富而非被动的积累。\n",
      "\n",
      "\n",
      "========== 正在生成第 2 段摘要 ==========\n",
      "\n",
      "原始文本 (第 2 段):\n",
      "使得很多人觉得那不重要 很多人甚至觉得 这门课要做final project 我就不修了太累了 或者说那门课需要 怎么样太累我就不要做了 而不知道 其实那个才是让你做的机会 然后可以学到最多 也就是说虽然很可能 那么辛苦的做很多事 没有让你获得什么具体成绩 对你的overfitting可能没有帮助 可是对你的全面学习 是很有帮助是该学的 不要漏掉这些事 这是我所说的 这个课业内可以做的这些事 刚才我们讲到思考的时候 我觉得我漏掉一点 你如果修我的信号课 你可能会发现 我上课没讲到一个数学试字的时候 我通常都不推它的 我是在解释那个数学试字在说什么话 同样的没讲到一个什么事情的时候 我通常就在解释他在说什么话 也就是说我在讲的就是 我读到特本那里的时候 我心里怎么想的 也就是我在告诉同学 如何读书的时候 如何一面读一面练习思考 这个才是最重要的一件事 如何培养自己思考的能力 跟培养思考的习惯 我觉得最好的办法就是 读书的时候 凡是读到一个数学试字 都去想一想 那个数学试字到底在说什么 凡是读到特本上讲什么 就去想一想 那个到底在说什么 你要真的了解他在说什么的时候 你就用了很多思考的功夫\n",
      "\n",
      "生成的摘要 (第 2 段):\n",
      "这段视频的摘要着重强调了参与一些看起来可能困难或没有立即回报的学习活动的重要性。许多人倾向于避免这些任务，认为它们过于繁重或无助于提高具体的成绩。然而，主讲人指出，正是这样的挑战性任务，如期末项目，提供了最为宝贵的学习机会，它们有助于提升全面的学术素养。\n",
      "\n",
      "视频中提到，在教授信号课程时，讲师通常不会详细推导数学公式，而是解释其含义以及如何应用于实际问题中，以帮助学生理解学习材料背后的思想。这种方法鼓励学生在阅读材料时主动思考，通过提出问题、探索概念，以及理解复杂的理论与实践应用之间的联系，来培养自己的思考能力和习惯。\n",
      "\n",
      "讲师建议学生应主动将这种思考习惯融入日常学习中，无论是遇到数学公式还是其他理论知识，都尝试去思考其含义，从而深入理解所学内容。这种持续的思考不仅有助于个人学术能力的提升，还能促进批判性思维、解决问题技巧的发展，对个人成长具有长远的价值。\n",
      "\n",
      "\n",
      "========== 正在生成第 3 段摘要 ==========\n",
      "\n",
      "原始文本 (第 3 段):\n",
      "你就在练习自己思考的能力了 好 以上说的是课业内的部分 那当然除了课业内之外 还有一大堆是不在课业内的 那就是课业外的 课业外也有很多式的 那我们可以举例来说 课业外有什么可以学习的 那我通常把学习定义成为 什么是学习 学习就是一种增长 一种进步 然后获得快乐 这就是学习 所以即使是课业外的任何事情 只要你觉得是有增长的 是有进步的 让你觉得快乐的 那应该就是值得学习的地方 那我们可以举很多例子 那譬如说 很多同学喜欢打球 打球是不是学习 当然是 打球中间有没有增长 当然有增长 打球不只是对健康有增长 而且可能对于 比如说手脑协调 譬如说团队精神 譬如说个人之间的互动 可能都有帮助 所以打球当然是有增长的 那当然是很好的学习的机会 有人喜欢爬山 爬山是不是好的学习机会 当然是 这个我以前两年前就讲过很多 爬山可以学到很多的 那爬山当然是一种学习 有人说我不喜欢爬山 我去旅行好不好 旅行当然好 旅行可以增长建设 可以扩增事业 可以增加很多很多 当然是有进步的 所以当然是很好的学习 你凡是获得快乐 都是很好的事 那这些都值得下功夫去 把它看成是学习 都值得下功夫去做的 我们再讲另外一系列 譬如说\n",
      "\n",
      "生成的摘要 (第 3 段):\n",
      "该段视频主要讲述了学习不仅仅局限于课业之内，同样在日常生活中的各种活动都能成为学习的机会。学习的本质是增长、进步并从中获得快乐。无论是打球、爬山、旅行还是其他任何活动，只要能够促进个人成长、提高技能、增进人际关系或带来乐趣，都可以视为有价值的学习体验。\n",
      "\n",
      "- **课业内与课业外学习**: 提出学习不应仅限定于学校教育，日常生活中任何能够促进个人成长的经历都是学习。\n",
      "- **学习的定义**: 学习定义为通过经历增长、进步，并从中获得快乐的过程。\n",
      "- **例证**:\n",
      "  - 打球：促进身体健康、提升手脑协调能力、培养团队合作与人际交往。\n",
      "  - 爬山：锻炼身体、增强毅力、学习自然知识和团队协作。\n",
      "  - 旅行：增长见闻、拓宽视野、提升适应不同环境的能力。\n",
      "- **普遍性**：强调各种形式的活动，只要能带来积极影响和乐趣，都值得投入时间和精力进行深入探索和学习。\n",
      "\n",
      "总之，视频鼓励观众将生活视为一个持续学习和成长的过程，无论是在学校还是日常生活中的每一个瞬间，都应该被视为学习的宝贵机会。\n",
      "\n",
      "\n",
      "========== 正在生成第 4 段摘要 ==========\n",
      "\n",
      "原始文本 (第 4 段):\n",
      "有人说谈恋爱是不是学习 谈恋爱除了你在谈恋爱上 会有收获以外 本身也是有收获的 因为让你体验到人跟人之间 的各种感觉 人跟人之间的各种期待等等 有没有帮助 当然有帮助 对每一个人都是很好的学习 所以谈恋爱当然是一件很好的事 有人会说那要靠缘分 没有缘分没有办法 对不对 但是你不是一定要谈恋爱吗 你可以交朋友 交朋友是不是学习 当然是 交朋友也一样 让我们学到很多人际的互动 学到很多人跟人之间的沟通 人跟人之间的期待 人跟人之间的感觉 这都是交朋友之后学到的 对我们电机系的同学而言 你四周有一大群好同学 都是很好的交朋友的对象 你下一番功夫交朋友好不好 好 当然是有帮助的 另外当然我们可以举很多 我们最现成的例子 譬如说我们的系学会办各种活动 那些活动有没有帮助 当然有 我们举例来讲电业 你如果去参加某一个舞跳个舞 或者参加某个剧演个剧 有没有帮助 当然有帮助 你在这中间一定发现有所增长 有所进步 那是为什么有那么多同学 要去参加 就是因为发现那个确实是有增长 有进步 有的人说 我不去跳那个舞 或者演那个剧 我做幕后的 譬如说是幕后的什么什么规划 或者说是什么这个光舞的什么软体组 还是什么服装道具组 一样\n",
      "\n",
      "生成的摘要 (第 4 段):\n",
      "这段视频讨论了恋爱、交友以及参与学生社团活动对个人成长的重要性。主要观点包括：\n",
      "\n",
      "1. **恋爱与学习**：恋爱不仅能让个体获得情感体验和成长，还能教会如何处理人际关系、了解人与人之间的期待和感受，对个人有着宝贵的学习机会。\n",
      "\n",
      "2. **交友的价值**：与朋友交往同样能提供人际互动、沟通技巧、理解他人期望和感受等方面的教育。交朋友对每个人都是很好的学习过程，有助于个人成长。\n",
      "\n",
      "3. **同学间的交流**：对电机系的学生来说，周围的同学是进行友好交往的绝佳对象。花时间培养友谊，能促进个人发展。\n",
      "\n",
      "4. **参与学生社团活动**：参加学生活动，如舞蹈、戏剧等，不仅能增强个人能力（如领导力、团队合作），还能带来成就感和新技能的获取。参与幕后工作（如策划、技术、服装道具等）也同样有价值，可以培养专业技能和团队协作精神。\n",
      "\n",
      "总结来说，视频强调了恋爱、交友及参加社团活动对于个人成长的积极影响，鼓励大学生充分利用这些机会，通过实践学习和探索，促进自我提升和发展。\n",
      "\n",
      "\n",
      "========== 正在生成第 5 段摘要 ==========\n",
      "\n",
      "原始文本 (第 5 段):\n",
      "那个都是可以获得很多的增长 很多进步的 当然都是很有用的 都是很好的学习 那当然也包括电业以外的戏学会 其他的各种活动都一样 也包括电机系以外的其他的校内 或者校外的各种活动几乎都一样 都可以让人有所增长 有所进步 都是很好的学习的机会 都是很好的学习 同样的问题是 这些东西都没有考试 没有成绩 不能显示在成绩单上 因此对有一些同学会认为那个浪费时间 我不需要花时间去做那个 因为不影响我的 overfitting的目标 里面没有这个嘛 具体成绩没有这些嘛 那不要这样想 因为那些都非常的重要 都对你发展非常的重要 那我们说电机工程 今天的电机工程 很少什么事情自己一个人可以做成功的 你必须跟很多人一起 才可能做成功一个非常重要的 有意义的工作 那当你跟一群人在一起做的时候 你必须学会如何进入一个团队 从边缘开始慢慢进入核心 从底层开始慢慢变成leader 然后如何可以推动你想做的事 如何变成可以做到你想做的事等等 这些都是很重要的 那我们通常称这些东西 是所谓的soft skills 也就是软实力 所谓软实力就是硬实力以外的软实力 那硬实力是说你电子学的功力 数学的功力 这个城市能力这种是硬实力\n",
      "\n",
      "生成的摘要 (第 5 段):\n",
      "这段视频文本主要强调了个人成长、学习以及团队协作的重要性，并针对一些学生可能存在的观念误区进行了澄清。以下是视频内容的摘要：\n",
      "\n",
      "1. **个人成长与学习机会**：视频指出，无论是在专业领域（如电机工程）或其他非专业活动中都能获得成长和进步，这些都是宝贵的学习机会。\n",
      "\n",
      "2. **价值与评价机制**：强调了某些活动的价值并不体现在考试成绩或成绩单上，这可能导致部分学生认为这些活动浪费时间。然而，这些经历对于个人发展至关重要。\n",
      "\n",
      "3. **团队协作与软技能**：提到电机工程等项目很少能由个人单独完成，团队合作极为重要。通过团队合作，个体能够学会如何融入团队、担任领导角色、推动目标实现，这些都是软技能的一部分。\n",
      "\n",
      "4. **软实力与硬实力的区别**：强调了除了专业知识（硬实力）之外，如沟通、合作、领导力等软实力同样重要。软实力帮助个体在团队中发挥更大作用，促进项目成功。\n",
      "\n",
      "5. **误区澄清**：纠正了认为只有与专业课程相关的活动才影响学业成绩的看法，强调了所有学习活动对于个人成长的重要性，特别是那些培养软技能的活动。\n",
      "\n",
      "总结，这段视频强调了全面发展的必要性，不仅限于专业技能的学习，也包括人际交往、团队合作等软技能的培养，这些都是现代教育体系中不可或缺的部分。\n",
      "\n",
      "\n",
      "========== 正在生成第 6 段摘要 ==========\n",
      "\n",
      "原始文本 (第 6 段):\n",
      "软实力我们主要就是讲各种人际之间的 在人跟人之间的各种能力 包括沟通能力协调能力 交朋友的能力 这个说服人的能力 这个团队精神领导能力等等 那那些就是所谓的soft skills 重要不重要 重要 你看到任何一个成功的电机工程师 他都有一堆这种 这个才是他成功的一个非常重要的关键 这种东西怎么来 我们刚才讲的各种课业外的各种学习增长机会 都可以帮助一个人塑造他的soft skills 是有少数人的这些soft skills是天生的 他天生就厉害 有没有 有 但这种人毕竟没那么多 对很多人而言 他的soft skills是自己努力慢慢培养起来的 我刚才一开始前面讲的那一段 我说我在这个进台大电机系以前 我几乎不会交朋友 我不太会说话 我在读大学的四年里面改变我自己 让我变成有很多这方面的能力的人 其实最重要就是我的很多soft skills 都是我自己培养 在读台大电机系的四年里面 获得的非常多这方面的收获的 那是为什么我每次都要强调这个东西有多么重要 那我之前曾经在几年前的这个信号与人生里面 有说到这一件事 我现在不要重复 但是我简单的summarize 那我说我们电机系的 电机工程师的一生career的发展\n",
      "\n",
      "生成的摘要 (第 6 段):\n",
      "这段视频文本强调了软技能在人际互动中的重要性。软技能包括沟通能力、协调能力、交友能力、说服能力及团队精神与领导力等，这些都是成功电机工程师的关键因素。视频指出，软技能的价值不容忽视，任何成功人士往往拥有这类技能，并将其作为成功的关键。软技能并非天生就具备，大部分可以通过课程外的学习、培训和个人努力获得。\n",
      "\n",
      "作者以个人经历为例，讲述自己从台大电机系学生到具备各种软技能的成长过程。在大学期间，通过自我提升，作者学会了如何交朋友、说话、表达观点，并在团队中发挥作用，这些软技能成为其职业生涯发展的重要支撑。\n",
      "\n",
      "视频总结了电机工程师一生的职业生涯发展中，软技能的重要性。作者在过去的作品《信号与人生》中曾详细讨论过这一话题，并通过简短回顾，强调软技能在个人成长和职业成功中的不可替代作用。\n",
      "\n",
      "\n",
      "========== 正在生成第 7 段摘要 ==========\n",
      "\n",
      "原始文本 (第 7 段):\n",
      "那黄金实在是在什么时候 我认为是在35岁到55岁 这20年是我们的黄金时代 在这以前当然更好 只是说可能各方面尚未具备 还没有完全训练的好 在这以后是最好的 这以后年纪大了难免有一些要打个折扣等等 就这里面我们看到我们的电机系的毕业的同学 过去有几千人毕业我都看到 那我觉得有的人的发展是像这样 有一定的斜率 但到某一个阶段它会慢慢saturate 有的人也许开始向上比较晚 但它斜率比较高 它最后会saturate在比较高的地方 也有的人也许开始的比较快 但是后来会overshoot之后收敛在比较低的地方等等 这个每一个人都不一样 但是当然也有一种人 你会看到它一直向上走 完全没有saturate 这些人这些差别在哪里 这些东西差别在哪里 我以前已经说过这件事 我不要多重复 我说最主要因素有四个 就是实力 努力 大致 跟self skill这四件事情 那我认为 真正影响这个的 不是因为电子学考得好不好 不是因为信号与系统念得好不好 也就我刚才讲 你把每一门必修课当成是单一跑道 跑到第一名并不表示怎样 我们最后不看那个的 最后看的是这个 那这个是怎么样影响 我认为是这四件事 就是实力努力 大致跟self\n",
      "\n",
      "生成的摘要 (第 7 段):\n",
      "视频内容主要围绕个人发展与黄金年龄时期的探讨，提出了35至55岁这一阶段是多数人的“黄金时代”，前后的阶段各有优劣。视频中强调，个人发展的轨迹如同斜率曲线，从某点开始逐渐饱和（saturate），并指出不同人在开始时间、斜率及最终饱和点上存在差异。\n",
      "\n",
      "视频重点关注电机系毕业生的发展路径，通过观察大量案例，发现个人发展的模式多样且独特。有人起步较晚却能达到较高的饱和点，而另一些人则可能早期发展迅速但最终回归到较低水平。强调每个人的情况不同，但成功的关键因素可以归纳为四大方面：实力、努力、天赋和自我技能。视频指出，这些因素比单纯的考试成绩更能决定个人长期发展的潜力和方向。\n",
      "\n",
      "此外，视频还提到了一个重要的观点，即真正的成功评价标准并非单一科目或考试成绩，而是个人在整个生涯中的综合成就与成长。这意味着，在追求个人发展时，不仅要注重知识和技能的积累，更要关注其在实践中的应用以及对社会的贡献。\n",
      "\n",
      "\n",
      "========== 正在生成第 8 段摘要 ==========\n",
      "\n",
      "原始文本 (第 8 段):\n",
      "skills 这四件事里面我们现在可以summarize 我刚才讲的 什么是实力 实力就是所有的这些我们电机工程的 专业领域里面的各种东西的实力 实力怎么厉害法 就是我刚才讲的 你如果都是在做全面的学习的话 你就会学到各种该学到的 最后你的实力就是很强的 所以实力最主要就是不要overfitting 要尽量都做学到该学的全面的学习 努力是没有疑问 每个人都了解 确实我们可以看到一个人在未来的几十年里面 有的人他一直努力 有的人慢慢不太努力等等 这个是有明显差别的 那self skills我刚才已经讲了 就是很多我们平常没有算成绩 觉得大家不重视的事情 那其他常常是很重要的 你如果好好的 多在各种课业外的事情上 增长进步的话 你这些东西会很强 这个东西会很厉害的 当然对少数人而言 他天生就有 他可能不需要 就是每一个人不一样的 这三个我都提过了 那么大致我还没有提 其实大致没有什么要特别说的 那应该就是我刚才前面有讲过 就是每一个人可以有你自己的长程目标 那有的人本来就有了 有的人也许我平常没有想过 那你可以在适当时机开始想 我有没有想要做什么事情 哪些事情可能是我的长程目标 我希望最后让我花个5年10年\n",
      "\n",
      "生成的摘要 (第 8 段):\n",
      "这段视频内容主要围绕“实力”、“自我技能”、“长期目标”三个关键点展开讨论。以下是摘要：\n",
      "\n",
      "实力指的是在电机工程等专业领域的全面学习与精通能力。要具备实力，应进行广泛而深入的学习，避免过度专注于某一方面，从而确保掌握所有应知应会的内容。实力的提升关键在于全面的学习和持续的努力，而不是仅靠偶尔的努力。\n",
      "\n",
      "自我技能则涵盖了那些日常学习中可能被忽视的、非学术性技能，如团队合作、沟通、领导力、时间管理等。这些技能同样能够显著增强个人竞争力，并且对于职业发展至关重要。通过不断实践和提升这些技能，可以在职场和社会生活中获得更大的优势。\n",
      "\n",
      "长期目标方面，视频鼓励听众思考并设定适合自己的长远规划。这包括识别自己真正感兴趣或擅长的领域，并为实现这些目标制定计划。无论是职业发展还是个人兴趣的发展，设定并追求长期目标都能使个人生活更加充实和有意义。重要的是，目标应该是可实现的，并且随着时间和经验的积累可以适时调整和优化。\n",
      "\n",
      "\n",
      "========== 正在生成第 9 段摘要 ==========\n",
      "\n",
      "原始文本 (第 9 段):\n",
      "15年或者更长 我把我的很多的努力都来 把某一些事情做得非常漂亮 那是我很想做的事 那就是长程目标 如果我觉得做那些事情 会让我非常的 觉得有意义 愿意花功夫下去做的 那就是我的长程目标 有的人如果可以想出这个来的话 那就是他的大致 越是有这种大致的人 也比较容易向上冲 那我感觉起来 真正影响的就是这四件事\n",
      "\n",
      "生成的摘要 (第 9 段):\n",
      "这段视频内容是关于设定及追求长程目标的重要性。主要强调了以下几个关键点：\n",
      "\n",
      "1. **长期投入与努力**：长期目标需要持续的关注和付出，意味着将大量时间、精力投入到某一件事情上，以期望达到出色的结果。\n",
      "\n",
      "2. **主观意义感**：只有当个人认为某个行动或目标本身具有内在意义时，才愿意进行长期的努力。这意味着目标必须与个人的价值观和兴趣相匹配，才能激发持久的动力。\n",
      "\n",
      "3. **目标设定的价值**：能够明确自己长期想要达成的事情不仅能够提升自我效能感，还能作为动力源泉驱动个人不断进步。设定这样的目标对于个人的成长和发展有着至关重要的作用。\n",
      "\n",
      "4. **实现目标的内在驱动力**：对于拥有清晰长期目标的人来说，他们更能把握自己的方向，更有可能克服挑战，实现目标。这种内在驱动力是区分一般人和卓越者的分水岭。\n",
      "\n",
      "通过这些要点，视频传达了设定并追求长程目标对于个人生活和职业发展的重要性，鼓励听众思考并明确自己的长远目标。\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# 定义摘要prompt模板\n",
    "summarization_prompt_template = \"用 300 个字以内写出这段视频文本的摘要，其中包括要点和所有重要细节：<text>\"\n",
    "\n",
    "# 对每个文本块生成摘要\n",
    "paragraph_summaries = []\n",
    "for index, chunk in enumerate(chunks):\n",
    "    print(f\"\\n========== 正在生成第 {index + 1} 段摘要 ==========\\n\")\n",
    "    print(f\"原始文本 (第 {index + 1} 段):\\n{chunk}\\n\")\n",
    "    \n",
    "    # 构建摘要prompt\n",
    "    summarization_prompt = summarization_prompt_template.replace(\"<text>\", chunk)\n",
    "    \n",
    "    # 调用摘要函数\n",
    "    summary = summarization(\n",
    "        client=client,\n",
    "        summarization_prompt=summarization_prompt,\n",
    "        model_name=model_name,\n",
    "        temperature=temperature,\n",
    "        top_p=top_p,\n",
    "        max_tokens=max_tokens\n",
    "    )\n",
    "    \n",
    "    # 打印生成的摘要\n",
    "    print(f\"生成的摘要 (第 {index + 1} 段):\\n{summary}\\n\")\n",
    "    \n",
    "    # 将生成的摘要保存到列表\n",
    "    paragraph_summaries.append(summary)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d70eef3a-c6ec-45c2-9d0d-71890c397a8a",
   "metadata": {},
   "source": [
    "2. 在分别获取每个小段落的摘要后，处理这些摘要以生成最终的摘要。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "caea2fe3-d128-4484-a347-df18be5b011c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "这段视频强调“学问是做出来的”，通过五个关键点阐述了这一理念：\n",
      "\n",
      "1. **行动的重要性**：理论知识的掌握需通过实践来深化理解与内化。\n",
      "2. **学习与实践结合**：通过具体活动学习，如课程、实验、项目，将知识应用于解决实际问题。\n",
      "3. **整合与应用**：完成项目或大型任务时综合运用知识，促进深入理解及长期记忆。\n",
      "4. **实例教学**：以复杂算法为例，通过解决实际问题而非单纯理论讨论，提升理解与运用能力。\n",
      "5. **个性化学习体验**：动手实践不仅获取知识，还能形成独特经验，超越考试成绩的评估范围。\n",
      "6. **持续行动的重要性**：强调实践与创造力的持续性，使学习过程生动有趣，知识转化为个人财富。\n",
      "\n",
      "视频进一步指出，学习不应局限在学校教育，生活中各种活动均可成为学习机会，强调全面成长的价值。它鼓励将生活视作持续学习的过程，无论活动是否直接关联学业，都能促进个人成长和进步。\n",
      "\n",
      "最后，视频聚焦于个人发展与黄金年龄段（35-55岁），强调这一时期的重要性，提出实力、自我技能、长期目标三大关键要素，并强调设定并追求个人目标的重要性。实力与自我技能的培养，以及明确的长期目标设定，共同构成个人发展的强大动力和方向。\n"
     ]
    }
   ],
   "source": [
    "# 合并段落摘要\n",
    "collected_summaries = \"\\n\".join(paragraph_summaries)\n",
    "\n",
    "# 定义最终摘要prompt模板\n",
    "final_summarization_prompt = \"在 500 字以内写出以下文字的简洁摘要：<text>\"\n",
    "final_summarization_prompt = final_summarization_prompt.replace(\"<text>\", collected_summaries)\n",
    "\n",
    "# 生成最终摘要\n",
    "final_summary = summarization(\n",
    "    client=client,\n",
    "    summarization_prompt=final_summarization_prompt,\n",
    "    model_name=model_name,\n",
    "    temperature=temperature,\n",
    "    top_p=top_p,\n",
    "    max_tokens=max_tokens\n",
    ")\n",
    "\n",
    "print(final_summary)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64f55c6e-9a41-409d-b130-679f122fd228",
   "metadata": {},
   "source": [
    "#### 方法二：精炼方法（the method of Refinement) - Refine\n",
    "\n",
    "Refinement 就是把每次的文本和之前的摘要结合起来丢给大模型，类似于迭代：\n",
    "\n",
    "![Refinement](../Guide/assets/image-20240924092753352.png)\n",
    "\n",
    "步骤（Pipeline）如下：\n",
    "- 第1步：从一小部分数据开始，运行prompt生成初始输出。\n",
    "- 第2步：对后续每个文档，将前一个输出与新文档结合输入。\n",
    "- 第3步：LLM 根据新文档中的信息精炼输出。\n",
    "- 第4步：此过程持续迭代，直到处理完所有文档。\n",
    "\n",
    "对应的核心代码："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d71e7f38-3eaa-4173-97aa-2d2052d1f729",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "========== 正在生成第 1 段的初始摘要 ==========\n",
      "\n",
      "原始文本 (第 1 段):\n",
      "每次说 学问是做出来的 什么意思 要做 才会获得学问 你如果每天光是坐在那里听 学问很可能是左耳近右耳出的 你光是坐在那儿读 学问可能从眼睛进入脑海之后就忘掉了 如何能够学问在脑海里面 真的变成你自己学问 就是要做 可能有很多同学有这个经验 你如果去修某一门课 或者做某一个实验 在期末就是要教一个final project 那个final project就是要你把 学到的很多东西 最后整合在你的final project里面 最后做出来的时候 就是把它们都整合了 当你学期结束 真的把final project做完的时候 你会忽然发现 我真的学到很多东西 那就是做出来的学问 也许可以举另外一个例子 就是你如果学了某一些 很复杂的演算法或者什么 好像觉得那些不见得在你的脑海里 可是后来老师出了个习题 那个习题叫你写一个很大的程式 要把所有东西都包进去 当你把这个程式写完的时候 你会发现 你忽然把演算法 所有东西都弄通了 那就是学问是做出来的 所以我们永远要记得 尽量多动手多做 在动手跟做的过程之中 学问才可以变成是自己的 同样的情形就是说 很多时候这样动手或者做的 表现或者成绩 没有一个成绩单上的数字\n",
      "\n",
      "生成的摘要 (第 1 段):\n",
      "这段视频强调\"学问是做出来的\"的核心理念。学习不仅仅是听或读，更要通过实践应用来深化理解并真正掌握知识。视频中引用了两种实例来解释这一观点：\n",
      "\n",
      "1. **课程与实验**：在学习某一门课程或进行实验项目时，学生通过执行具体任务（如完成最终项目）来整合并应用所学知识。这种主动参与的过程使学生在学期末时能够显著感受到自己已掌握了大量信息，即\"做出来的学问\"。\n",
      "\n",
      "2. **复杂算法应用**：学习复杂算法或技术后，学生可能会认为这些知识并未完全内化。但当他们面临需要将这些知识应用于实际编程任务时，通过实践完成大项目，就能将所学知识真正理解和消化。\n",
      "\n",
      "视频反复强调动手实践的重要性，指出只有通过实践才能将知识转化为个人的学问，而这种学问往往无法通过成绩单上的数字直接衡量。因此，鼓励观众把握机会主动实践，从而在实际操作中深化学习效果，让学问真正成为自身的一部分。\n",
      "\n",
      "\n",
      "========== 正在生成第 2 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 2 段):\n",
      "使得很多人觉得那不重要 很多人甚至觉得 这门课要做final project 我就不修了太累了 或者说那门课需要 怎么样太累我就不要做了 而不知道 其实那个才是让你做的机会 然后可以学到最多 也就是说虽然很可能 那么辛苦的做很多事 没有让你获得什么具体成绩 对你的overfitting可能没有帮助 可是对你的全面学习 是很有帮助是该学的 不要漏掉这些事 这是我所说的 这个课业内可以做的这些事 刚才我们讲到思考的时候 我觉得我漏掉一点 你如果修我的信号课 你可能会发现 我上课没讲到一个数学试字的时候 我通常都不推它的 我是在解释那个数学试字在说什么话 同样的没讲到一个什么事情的时候 我通常就在解释他在说什么话 也就是说我在讲的就是 我读到特本那里的时候 我心里怎么想的 也就是我在告诉同学 如何读书的时候 如何一面读一面练习思考 这个才是最重要的一件事 如何培养自己思考的能力 跟培养思考的习惯 我觉得最好的办法就是 读书的时候 凡是读到一个数学试字 都去想一想 那个数学试字到底在说什么 凡是读到特本上讲什么 就去想一想 那个到底在说什么 你要真的了解他在说什么的时候 你就用了很多思考的功夫\n",
      "\n",
      "生成的摘要 (第 2 段):\n",
      "这段视频强调了“学问是做出来的”这一核心理念，意味着真正的学习不仅仅是理论知识的吸收，更在于通过实践应用加深理解。视频通过两种实例阐述了这个观点：学生在完成课程项目和实验时，通过实践整合并运用所学知识；学习复杂算法后，实际操作项目能促进知识的真正内化。视频鼓励观众把握实践机会，通过动手操作深化学习，即使过程艰苦且不易量化为具体成绩，也对个人全面发展至关重要。此外，视频还提到，有效的学习方法不仅包括理解课程内容，更重要的是培养独立思考能力，这可以通过在阅读过程中主动提问、深入思考实现。因此，视频强调了实践和思考在学习过程中的重要性，并倡导在学习中积极运用这两种方法以提升自我。\n",
      "\n",
      "\n",
      "========== 正在生成第 3 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 3 段):\n",
      "你就在练习自己思考的能力了 好 以上说的是课业内的部分 那当然除了课业内之外 还有一大堆是不在课业内的 那就是课业外的 课业外也有很多式的 那我们可以举例来说 课业外有什么可以学习的 那我通常把学习定义成为 什么是学习 学习就是一种增长 一种进步 然后获得快乐 这就是学习 所以即使是课业外的任何事情 只要你觉得是有增长的 是有进步的 让你觉得快乐的 那应该就是值得学习的地方 那我们可以举很多例子 那譬如说 很多同学喜欢打球 打球是不是学习 当然是 打球中间有没有增长 当然有增长 打球不只是对健康有增长 而且可能对于 比如说手脑协调 譬如说团队精神 譬如说个人之间的互动 可能都有帮助 所以打球当然是有增长的 那当然是很好的学习的机会 有人喜欢爬山 爬山是不是好的学习机会 当然是 这个我以前两年前就讲过很多 爬山可以学到很多的 那爬山当然是一种学习 有人说我不喜欢爬山 我去旅行好不好 旅行当然好 旅行可以增长建设 可以扩增事业 可以增加很多很多 当然是有进步的 所以当然是很好的学习 你凡是获得快乐 都是很好的事 那这些都值得下功夫去 把它看成是学习 都值得下功夫去做的 我们再讲另外一系列 譬如说\n",
      "\n",
      "生成的摘要 (第 3 段):\n",
      "这段视频聚焦于“学问是做出来的”的核心理念，强调理论知识与实践应用相结合的重要性。通过学生在完成课程项目及实验时的实际操作，以及学习复杂算法后的项目实践，视频展现了理论知识如何通过实践而被内化和深化理解。它鼓励观众把握实践机会，通过亲自动手提高学习效率，即便这些体验难以量化为具体成绩，但对个人全面发展至关重要。同时，有效学习应包括独立思考能力的培养，这可通过主动提问和深入思考的方式实现。\n",
      "\n",
      "视频进一步扩展了学习的概念，指出学习不仅仅局限于课堂之内，也体现在课外活动上。它提出，任何形式的学习都旨在促进个体成长、进步并带来快乐。无论是体育锻炼、户外探险还是旅行，只要能带来增长、进步和快乐，都是值得投入时间和精力进行的学习活动。视频最后提到，探索更多学习途径，如技能训练、社交互动等，都是促进个人全面发展的有效方式。\n",
      "\n",
      "\n",
      "========== 正在生成第 4 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 4 段):\n",
      "有人说谈恋爱是不是学习 谈恋爱除了你在谈恋爱上 会有收获以外 本身也是有收获的 因为让你体验到人跟人之间 的各种感觉 人跟人之间的各种期待等等 有没有帮助 当然有帮助 对每一个人都是很好的学习 所以谈恋爱当然是一件很好的事 有人会说那要靠缘分 没有缘分没有办法 对不对 但是你不是一定要谈恋爱吗 你可以交朋友 交朋友是不是学习 当然是 交朋友也一样 让我们学到很多人际的互动 学到很多人跟人之间的沟通 人跟人之间的期待 人跟人之间的感觉 这都是交朋友之后学到的 对我们电机系的同学而言 你四周有一大群好同学 都是很好的交朋友的对象 你下一番功夫交朋友好不好 好 当然是有帮助的 另外当然我们可以举很多 我们最现成的例子 譬如说我们的系学会办各种活动 那些活动有没有帮助 当然有 我们举例来讲电业 你如果去参加某一个舞跳个舞 或者参加某个剧演个剧 有没有帮助 当然有帮助 你在这中间一定发现有所增长 有所进步 那是为什么有那么多同学 要去参加 就是因为发现那个确实是有增长 有进步 有的人说 我不去跳那个舞 或者演那个剧 我做幕后的 譬如说是幕后的什么什么规划 或者说是什么这个光舞的什么软体组 还是什么服装道具组 一样\n",
      "\n",
      "生成的摘要 (第 4 段):\n",
      "这段视频中心围绕“学问是做出来的”的理念，探讨理论与实践结合的重要性，通过学生实际操作来深化理解。视频鼓励观众在实践中提高学习效率，即使体验不易量化的成绩，也强调对个人全方面的关键性价值。它指出学习应拓展至课堂之外，包括体育、户外、旅行等活动，任何能够促进成长、带来快乐的活动都值得投入时间。视频还特别提及恋爱和交友作为人际互动的学习途径，强调它们对个人成长的积极影响。此外，参与系会举办的各类活动，如舞蹈、戏剧表演，也被视作促进个人成长的有效方式，因为它们能带来增长和进步。参与幕后工作，如策划、软件开发、服装道具等，同样被视为有价值的活动，因为它们提供了不同的视角和技能提升的机会。总而言之，这段视频强调了多样化的学习形式对于个人发展的重要性。\n",
      "\n",
      "\n",
      "========== 正在生成第 5 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 5 段):\n",
      "那个都是可以获得很多的增长 很多进步的 当然都是很有用的 都是很好的学习 那当然也包括电业以外的戏学会 其他的各种活动都一样 也包括电机系以外的其他的校内 或者校外的各种活动几乎都一样 都可以让人有所增长 有所进步 都是很好的学习的机会 都是很好的学习 同样的问题是 这些东西都没有考试 没有成绩 不能显示在成绩单上 因此对有一些同学会认为那个浪费时间 我不需要花时间去做那个 因为不影响我的 overfitting的目标 里面没有这个嘛 具体成绩没有这些嘛 那不要这样想 因为那些都非常的重要 都对你发展非常的重要 那我们说电机工程 今天的电机工程 很少什么事情自己一个人可以做成功的 你必须跟很多人一起 才可能做成功一个非常重要的 有意义的工作 那当你跟一群人在一起做的时候 你必须学会如何进入一个团队 从边缘开始慢慢进入核心 从底层开始慢慢变成leader 然后如何可以推动你想做的事 如何变成可以做到你想做的事等等 这些都是很重要的 那我们通常称这些东西 是所谓的soft skills 也就是软实力 所谓软实力就是硬实力以外的软实力 那硬实力是说你电子学的功力 数学的功力 这个城市能力这种是硬实力\n",
      "\n",
      "生成的摘要 (第 5 段):\n",
      "这段视频强调理论与实践结合的重要性，倡导通过实际行动加深理解和提高学习效率。它鼓励观众走出教室，参与体育、户外活动、旅行、恋爱、交友以及系会组织的舞蹈、戏剧表演等多种活动，认为这些活动对个人成长至关重要。视频还特别关注参与幕后工作的机会，比如策划、软件开发、服装道具设计等，认为这不仅能带来增长和进步，还能培养团队合作、领导力等软实力。尽管这些活动不体现在成绩单上，但它们对个人发展具有深远意义。\n",
      "\n",
      "第五段进一步阐述了软实力（即非专业技能）的重要性。在现代社会，尤其是电机工程领域，很少有项目能由个人独立完成。成功的项目往往需要团队合作，成员们需从边缘角色逐渐成长为团队核心，学会推动项目的进展。这些软实力，如团队协作、领导能力、沟通技巧等，虽然无法量化成具体的考试成绩或成果，但对于个人职业发展和生活经历来说，其价值无可估量。因此，视频提倡将时间和精力投入到这些软实力的培养上，尽管它们不在传统的学术评价体系内，却对个人的全面成长不可或缺。\n",
      "\n",
      "\n",
      "========== 正在生成第 6 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 6 段):\n",
      "软实力我们主要就是讲各种人际之间的 在人跟人之间的各种能力 包括沟通能力协调能力 交朋友的能力 这个说服人的能力 这个团队精神领导能力等等 那那些就是所谓的soft skills 重要不重要 重要 你看到任何一个成功的电机工程师 他都有一堆这种 这个才是他成功的一个非常重要的关键 这种东西怎么来 我们刚才讲的各种课业外的各种学习增长机会 都可以帮助一个人塑造他的soft skills 是有少数人的这些soft skills是天生的 他天生就厉害 有没有 有 但这种人毕竟没那么多 对很多人而言 他的soft skills是自己努力慢慢培养起来的 我刚才一开始前面讲的那一段 我说我在这个进台大电机系以前 我几乎不会交朋友 我不太会说话 我在读大学的四年里面改变我自己 让我变成有很多这方面的能力的人 其实最重要就是我的很多soft skills 都是我自己培养 在读台大电机系的四年里面 获得的非常多这方面的收获的 那是为什么我每次都要强调这个东西有多么重要 那我之前曾经在几年前的这个信号与人生里面 有说到这一件事 我现在不要重复 但是我简单的summarize 那我说我们电机系的 电机工程师的一生career的发展\n",
      "\n",
      "生成的摘要 (第 6 段):\n",
      "这段视频着重于理论与实践相结合的重要性，提倡通过参与多样化的活动，如体育、户外探险、旅行、社交活动和文艺演出等，来深化理解并提升学习效率。它指出，这些经历对个人成长和自我提升具有重要意义，特别是参与幕后工作，如活动策划、软件开发、服装设计等，不仅能促进技能增长，还能够培养团队协作、领导力等软实力，这些能力虽难以量化，但对职业发展和生活经验至关重要。\n",
      "\n",
      "视频进一步强调了软实力在现代社会，特别是在电机工程等领域的价值，尤其是在项目合作中，团队合作、领导能力、沟通技巧等软实力不可或缺。软实力并非天生具备，而是通过课外学习和实践经验逐步培养起来的。作者分享了自己在台大电机系求学期间，如何通过参与各种活动，从不擅长交际到拥有丰富人际交往能力的经历，证明了软实力对于个人成长的重要性。作者曾在此前的演讲中详细讨论过这一话题，提醒大家认识到软实力培养对于个人职业生涯发展的关键作用。视频最后总结，电机工程师的职业生涯发展中，软实力的培养至关重要，应鼓励学生积极参与课外活动，以促进全面发展。\n",
      "\n",
      "\n",
      "========== 正在生成第 7 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 7 段):\n",
      "那黄金实在是在什么时候 我认为是在35岁到55岁 这20年是我们的黄金时代 在这以前当然更好 只是说可能各方面尚未具备 还没有完全训练的好 在这以后是最好的 这以后年纪大了难免有一些要打个折扣等等 就这里面我们看到我们的电机系的毕业的同学 过去有几千人毕业我都看到 那我觉得有的人的发展是像这样 有一定的斜率 但到某一个阶段它会慢慢saturate 有的人也许开始向上比较晚 但它斜率比较高 它最后会saturate在比较高的地方 也有的人也许开始的比较快 但是后来会overshoot之后收敛在比较低的地方等等 这个每一个人都不一样 但是当然也有一种人 你会看到它一直向上走 完全没有saturate 这些人这些差别在哪里 这些东西差别在哪里 我以前已经说过这件事 我不要多重复 我说最主要因素有四个 就是实力 努力 大致 跟self skill这四件事情 那我认为 真正影响这个的 不是因为电子学考得好不好 不是因为信号与系统念得好不好 也就我刚才讲 你把每一门必修课当成是单一跑道 跑到第一名并不表示怎样 我们最后不看那个的 最后看的是这个 那这个是怎么样影响 我认为是这四件事 就是实力努力 大致跟self\n",
      "\n",
      "生成的摘要 (第 7 段):\n",
      "这段新增内容围绕电机工程专业毕业生的事业发展，提出35至55岁这一阶段为个人职业发展的黄金时期，强调在这之前或之后的发展亦有价值，但可能受限于经验和成熟度。作者通过观察数千名台大电机系毕业生的生涯轨迹，发现个人的成长曲线各不相同，有人发展迅速，有人起步较晚，但最终达到较高水平。他指出，影响个体职业生涯发展的关键因素包括实力、努力、机遇和自我技能。\n",
      "\n",
      "结合前文摘要，这段新增内容补充了关于职业发展的重要性和个人成长的关键因素的话题。原摘要强调了理论与实践结合的重要性，以及参与多样活动如何促进技能增长和个人能力的培养，尤其是软实力的提升。新增内容则聚焦于电机工程师的职业生涯，指出黄金成长期为35-55岁，并探讨了个人职业发展中的关键因素，如实力、努力、机遇和自我技能的作用。通过两者结合，文章旨在鼓励学生不仅在学业上追求卓越，同时积极参与课外活动，以全面促进个人成长和职业发展。\n",
      "\n",
      "\n",
      "========== 正在生成第 8 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 8 段):\n",
      "skills 这四件事里面我们现在可以summarize 我刚才讲的 什么是实力 实力就是所有的这些我们电机工程的 专业领域里面的各种东西的实力 实力怎么厉害法 就是我刚才讲的 你如果都是在做全面的学习的话 你就会学到各种该学到的 最后你的实力就是很强的 所以实力最主要就是不要overfitting 要尽量都做学到该学的全面的学习 努力是没有疑问 每个人都了解 确实我们可以看到一个人在未来的几十年里面 有的人他一直努力 有的人慢慢不太努力等等 这个是有明显差别的 那self skills我刚才已经讲了 就是很多我们平常没有算成绩 觉得大家不重视的事情 那其他常常是很重要的 你如果好好的 多在各种课业外的事情上 增长进步的话 你这些东西会很强 这个东西会很厉害的 当然对少数人而言 他天生就有 他可能不需要 就是每一个人不一样的 这三个我都提过了 那么大致我还没有提 其实大致没有什么要特别说的 那应该就是我刚才前面有讲过 就是每一个人可以有你自己的长程目标 那有的人本来就有了 有的人也许我平常没有想过 那你可以在适当时机开始想 我有没有想要做什么事情 哪些事情可能是我的长程目标 我希望最后让我花个5年10年\n",
      "\n",
      "生成的摘要 (第 8 段):\n",
      "摘要：\n",
      "\n",
      "本文聚焦于电机工程专业毕业生的职业生涯发展，提出35至55岁被视为个人职业黄金发展时期的观点。尽管不同人在不同阶段发展速度各异，最终都能实现较高的成就水平，但其成功取决于实力、努力、机遇和自我技能等多个关键因素。实力主要指在电机工程专业领域的广泛学习和实践，确保知识体系全面，避免过度专业化。努力是个人持续投入工作与学习的体现，不同的努力程度会对职业发展产生显著影响。自我技能，即非学术性能力的提升，如团队协作、问题解决等，在个人成长中同样至关重要。文章鼓励学生不仅要专注于专业知识的学习，还应积极参与多方面活动，全面发展自身能力，为未来的职业生涯奠定坚实基础。\n",
      "\n",
      "\n",
      "========== 正在生成第 9 段的精炼摘要 ==========\n",
      "\n",
      "原始文本 (第 9 段):\n",
      "15年或者更长 我把我的很多的努力都来 把某一些事情做得非常漂亮 那是我很想做的事 那就是长程目标 如果我觉得做那些事情 会让我非常的 觉得有意义 愿意花功夫下去做的 那就是我的长程目标 有的人如果可以想出这个来的话 那就是他的大致 越是有这种大致的人 也比较容易向上冲 那我感觉起来 真正影响的就是这四件事\n",
      "\n",
      "生成的摘要 (第 9 段):\n",
      "摘要：\n",
      "\n",
      "本文探讨电机工程专业毕业生的职业发展路径，强调35至55岁为职业黄金期，成就高度取决于实力、努力、机遇及自我技能。实力涉及电机工程领域的广泛学习与实践，确保知识体系的全面性。个人努力体现在持续投入工作与学习，显著影响职业进程。此外，提升非学术性能力，如团队协作和问题解决，对个人成长至关重要。文章提倡学生不仅应专注专业学习，还需参与多样活动，综合发展能力，为职场生涯打下坚实基础。\n",
      "\n",
      "个人职业规划中，设定长期目标尤为重要，应选择能带来意义且愿意投入大量精力的事物作为长程目标。具备明确长期目标的人往往更容易取得职业成功。影响职业发展的关键因素包括实力、努力、机遇和自我技能提升。\n",
      "\n",
      "\n",
      "========== 最终精炼摘要结果 ==========\n",
      "\n",
      "摘要：\n",
      "\n",
      "本文探讨电机工程专业毕业生的职业发展路径，强调35至55岁为职业黄金期，成就高度取决于实力、努力、机遇及自我技能。实力涉及电机工程领域的广泛学习与实践，确保知识体系的全面性。个人努力体现在持续投入工作与学习，显著影响职业进程。此外，提升非学术性能力，如团队协作和问题解决，对个人成长至关重要。文章提倡学生不仅应专注专业学习，还需参与多样活动，综合发展能力，为职场生涯打下坚实基础。\n",
      "\n",
      "个人职业规划中，设定长期目标尤为重要，应选择能带来意义且愿意投入大量精力的事物作为长程目标。具备明确长期目标的人往往更容易取得职业成功。影响职业发展的关键因素包括实力、努力、机遇和自我技能提升。\n"
     ]
    }
   ],
   "source": [
    "# 定义初始摘要prompt模板\n",
    "summarization_prompt_template = \"用 300 个字以内写出这段视频文本的摘要，其中包括要点和所有重要细节:<text>\"\n",
    "\n",
    "# 定义精炼摘要prompt模板\n",
    "summarization_prompt_refinement_template = \"请在 500 字以内，结合原先的摘要和新的内容，提供简洁的摘要:<text>\"\n",
    "\n",
    "# 初始化保存摘要的列表\n",
    "refined_summaries = []\n",
    "\n",
    "# 对文本块逐步进行精炼摘要，并打印中间过程\n",
    "for index, chunk in enumerate(chunks):\n",
    "    if index == 0:\n",
    "        # 第一步：对第一段文本生成初始摘要\n",
    "        print(f\"\\n========== 正在生成第 {index + 1} 段的初始摘要 ==========\\n\")\n",
    "        print(f\"原始文本 (第 {index + 1} 段):\\n{chunk}\\n\")\n",
    "        \n",
    "        # 构建初始摘要prompt\n",
    "        summarization_prompt = summarization_prompt_template.replace(\"<text>\", chunk)\n",
    "        \n",
    "        # 调用摘要函数生成第一个摘要\n",
    "        first_summary = summarization(\n",
    "            client=client,\n",
    "            summarization_prompt=summarization_prompt,\n",
    "            model_name=model_name,\n",
    "            temperature=temperature,\n",
    "            top_p=top_p,\n",
    "            max_tokens=max_tokens\n",
    "        )\n",
    "        \n",
    "        # 打印生成的初始摘要\n",
    "        print(f\"生成的摘要 (第 {index + 1} 段):\\n{first_summary}\\n\")\n",
    "        \n",
    "        # 保存生成的摘要\n",
    "        refined_summaries.append(first_summary)\n",
    "\n",
    "    else:\n",
    "        # 后续步骤：结合前一个摘要与当前段落进行精炼\n",
    "        print(f\"\\n========== 正在生成第 {index + 1} 段的精炼摘要 ==========\\n\")\n",
    "        print(f\"原始文本 (第 {index + 1} 段):\\n{chunk}\\n\")\n",
    "        \n",
    "        # 构建精炼摘要的输入文本，将前一个摘要与当前段落内容结合\n",
    "        chunk_with_previous_summary = f\"前 {index} 段的摘要: {refined_summaries[-1]}\\n第 {index + 1} 段的内容: {chunk}\"\n",
    "        \n",
    "        # 构建精炼摘要prompt\n",
    "        summarization_prompt = summarization_prompt_refinement_template.replace(\"<text>\", chunk_with_previous_summary)\n",
    "        \n",
    "        # 调用摘要函数生成精炼摘要\n",
    "        refined_summary = summarization(\n",
    "            client=client,\n",
    "            summarization_prompt=summarization_prompt,\n",
    "            model_name=model_name,\n",
    "            temperature=temperature,\n",
    "            top_p=top_p,\n",
    "            max_tokens=max_tokens\n",
    "        )\n",
    "        \n",
    "        # 打印生成的精炼摘要\n",
    "        print(f\"生成的摘要 (第 {index + 1} 段):\\n{refined_summary}\\n\")\n",
    "        \n",
    "        # 保存生成的精炼摘要\n",
    "        refined_summaries.append(refined_summary)\n",
    "\n",
    "# 最终的精炼摘要结果就是 refined_summaries 列表的最后一个元素\n",
    "final_refined_summary = refined_summaries[-1]\n",
    "\n",
    "print(\"\\n========== 最终精炼摘要结果 ==========\\n\")\n",
    "print(final_refined_summary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d290b55d-60e1-48ec-bdf1-7c90a904af8b",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}