From 6ea1df5d696a33ab3c8555d759c6b626d33f6d88 Mon Sep 17 00:00:00 2001
From: Pu Fanyi <pufanyi@gmail.com>
Date: Mon, 30 Dec 2024 15:05:32 +0800
Subject: [PATCH] =?UTF-8?q?=F0=9F=97=91=EF=B8=8F=20remove=20unused=20text?=
 =?UTF-8?q?=20processing=20notebook?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 lmms_eval/models/text.ipynb | 156 ------------------------------------
 1 file changed, 156 deletions(-)
 delete mode 100644 lmms_eval/models/text.ipynb

diff --git a/lmms_eval/models/text.ipynb b/lmms_eval/models/text.ipynb
deleted file mode 100644
index de40cc2c0..000000000
--- a/lmms_eval/models/text.ipynb
+++ /dev/null
@@ -1,156 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import re\n",
-    "\n",
-    "def split_media_tags(content):\n",
-    "    # 用正则表达式匹配 <media数字>\n",
-    "    # (\\d+) 捕获组用来提取数字\n",
-    "    pattern = r'<media(\\d+)>'\n",
-    "    \n",
-    "    # 用 split 方法分割文本,同时保留匹配到的数字\n",
-    "    # re.split 会返回一个列表,包含分割后的文本和匹配到的捕获组\n",
-    "    parts = re.split(pattern, content)\n",
-    "    \n",
-    "    # 处理结果列表,将数字转换为整型\n",
-    "    result = []\n",
-    "    for i, part in enumerate(parts):\n",
-    "        if i % 2 == 0:  # 偶数索引是文本\n",
-    "            result.append(part)\n",
-    "        else:  # 奇数索引是数字\n",
-    "            result.append(int(part))\n",
-    "            \n",
-    "    return result"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "['', 1, 'world', 2, '!']"
-      ]
-     },
-     "execution_count": 6,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "split_media_tags('<media1>world<media2>!')  # ['hello', 1, 'world', 2, '!']"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Input: Select the instrument represented in images that corresponds to the audio [audio1] from [img1] [img2] [img3] [img4].\n",
-      "Output: ['Select the instrument represented in images that corresponds to the audio ', ('audio', 1), ' from ', ('img', 1), ' ', ('img', 2), ' ', ('img', 3), ' ', ('img', 4), '.']\n",
-      "\n",
-      "Input: [video1] is a great video\n",
-      "Output: [('video', 1), ' is a great video']\n",
-      "\n",
-      "Input: Compare [img1] and [img2]\n",
-      "Output: ['Compare ', ('img', 1), ' and ', ('img', 2)]\n",
-      "\n",
-      "Input: Listen to [audio1] and watch [video1]\n",
-      "Output: ['Listen to ', ('audio', 1), ' and watch ', ('video', 1)]\n",
-      "\n",
-      "Input: [img1] at the beginning and [img2] at the end\n",
-      "Output: [('img', 1), ' at the beginning and ', ('img', 2), ' at the end']\n",
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "import re\n",
-    "\n",
-    "def split_media_tags(content):\n",
-    "    # 匹配 [类型数字] 格式的标签\n",
-    "    # 捕获组 1 捕获类型 (audio|video|img)\n",
-    "    # 捕获组 2 捕获数字\n",
-    "    pattern = r'\\[(audio|video|img)(\\d+)\\]'\n",
-    "    \n",
-    "    # 用 finditer 找到所有匹配\n",
-    "    matches = list(re.finditer(pattern, content))\n",
-    "    if not matches:\n",
-    "        return [content]\n",
-    "    \n",
-    "    result = []\n",
-    "    last_end = 0\n",
-    "    \n",
-    "    for match in matches:\n",
-    "        # 添加标签之前的文本（如果有）\n",
-    "        if match.start() > last_end:\n",
-    "            result.append(content[last_end:match.start()])\n",
-    "            \n",
-    "        # 添加标签信息为元组 (类型, 数字)\n",
-    "        media_type = match.group(1)  # audio, video 或 img\n",
-    "        media_num = int(match.group(2))  # 数字\n",
-    "        result.append((media_type, media_num))\n",
-    "        \n",
-    "        last_end = match.end()\n",
-    "    \n",
-    "    # 添加最后一个标签之后的文本（如果有）\n",
-    "    if last_end < len(content):\n",
-    "        result.append(content[last_end:])\n",
-    "    \n",
-    "    return result\n",
-    "\n",
-    "# 测试\n",
-    "test_cases = [\n",
-    "    \"Select the instrument represented in images that corresponds to the audio [audio1] from [img1] [img2] [img3] [img4].\",\n",
-    "    \"[video1] is a great video\",\n",
-    "    \"Compare [img1] and [img2]\",\n",
-    "    \"Listen to [audio1] and watch [video1]\",\n",
-    "    \"[img1] at the beginning and [img2] at the end\",\n",
-    "]\n",
-    "\n",
-    "for test in test_cases:\n",
-    "    print(f\"Input: {test}\")\n",
-    "    print(f\"Output: {split_media_tags(test)}\\n\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "av-odyssey",
-   "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": "3.11.11"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}