File size: 5,892 Bytes
55500d6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 |
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# ηζζζζ½ζ₯"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|ββββββββββ| 40/40 [00:00<00:00, 171.67it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"[{\"Q\": \"What color is the reporter's coat during the interview?',\n",
" 'Options': [\n",
" 'A. Red.',\n",
" 'B. Black.',\n",
" 'C. Blue.',\n",
" 'D. White.'\n",
" ],\n",
" 'Answer': 'B\"}]\n",
"[{\"Q\": \"What color is the man's jacket in the video?', 'Options': ['A. Red.', 'B. Blue.', 'C. Green.', 'D. Black.'], 'Answer': 'B\"}]\n",
"[{\"Q\": \"What is the person's hands manipulating in the video clip?', 'Options': ['A. A small electronic component.', 'B. A large mechanical part.', 'C. A piece of paper.', 'D. A tool from the workspace.'], 'Answer': 'A\"}]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n"
]
}
],
"source": [
"import json\n",
"import os\n",
"import random\n",
"import shutil\n",
"from tqdm import tqdm\n",
"import re\n",
"\n",
"def parse(generated_text):\n",
" generated_text = generated_text.strip()\n",
" if \"```json\" in generated_text:\n",
" generated_text = re.sub(r\"^```json\\s*|\\s*```$\", \"\", generated_text.strip())\n",
" try:\n",
" data = eval(generated_text)\n",
" except:\n",
" generated_text = generated_text.replace('\\'Q\\': \\'', \"\\\"Q\\\": \\\"\").replace('\\', \\'A\\': \\'', \"\\\", \\\"A\\\": \\\"\").replace('\\'}', \"\\\"}\")\n",
" print(generated_text)\n",
" data = eval(generated_text)\n",
"\n",
" return data\n",
"\n",
"def load_jsonl(path):\n",
" datas = []\n",
" with open(path, 'r') as file:\n",
" for line in file:\n",
" data = json.loads(line)\n",
" datas.append(data)\n",
" return datas\n",
"\n",
"\n",
"results_dir = '/share/minghao/VideoProjects/Sythesis2/ObjectRecognition/Results/object_0_40'\n",
"file_names = os.listdir(results_dir)\n",
"all_datas = []\n",
"for name in file_names:\n",
" file_path = os.path.join(results_dir, name)\n",
" if os.path.isdir(file_path):\n",
" continue\n",
" datas = load_jsonl(file_path)\n",
" all_datas.extend(datas)\n",
"\n",
"total_size = len(all_datas)\n",
"check_size = total_size//10\n",
"check_size = 40\n",
"check_datas = random.sample(all_datas, check_size)\n",
"observation_dir = '/share/minghao/VideoProjects/Sythesis2/ObjectRecognition/Observations'\n",
"\n",
"for data in tqdm(check_datas):\n",
" video_id = data['video_id']\n",
" generated_text = data['generated_qa']\n",
" try:\n",
" result = parse(generated_text)\n",
" except:\n",
" continue\n",
" with open(os.path.join(observation_dir, video_id+'.json'), 'w') as file:\n",
" json.dump(result, file, indent=4)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|ββββββββββ| 2700/2700 [00:00<00:00, 1360151.43it/s]\n"
]
}
],
"source": [
"import json\n",
"from tqdm import tqdm\n",
"\n",
"path = '/share/minghao/VideoProjects/VideoChat-Flash/lmms-eval_videochat/videochat-flash-7B@448_eval_log_videomme.json'\n",
"\n",
"with open(path, 'r') as f:\n",
" datas = json.load(f)\n",
"\n",
"score_task_type = {}\n",
"score_duration_task_type = {\n",
" 'short': {},\n",
" 'medium': {},\n",
" 'long': {}\n",
"}\n",
"\n",
"record_error = {\n",
" 'short': [],\n",
" 'medium': [],\n",
" 'long': []\n",
"}\n",
"\n",
"for data in tqdm(datas['logs']):\n",
" duration = data['doc']['duration']\n",
" task_type = data['doc']['task_type']\n",
" if task_type != 'Action Reasoning':\n",
" continue\n",
" pred_answer = data['videomme_percetion_score']['pred_answer']\n",
" answer = data['videomme_percetion_score']['answer']\n",
" if pred_answer == answer:\n",
" correct_flag = True\n",
" else:\n",
" correct_flag = False\n",
"\n",
" if not correct_flag:\n",
" info = {\n",
" 'url' : data['doc']['url'],\n",
" 'Q': data['doc']['question'],\n",
" 'Options': data['doc']['options'],\n",
" 'Answer': answer,\n",
" 'pred': pred_answer\n",
" }\n",
"\n",
"\n",
" record_error[duration].append(info)\n",
"\n",
"import os\n",
"\n",
"ob_dir = '/share/minghao/VideoProjects/Sythesis/ActionReasoning/ErrorObservation'\n",
"for duration, error_infos in record_error.items():\n",
" ob_path = os.path.join(ob_dir, f'{duration}.json')\n",
" with open(ob_path, 'w') as file:\n",
" json.dump(error_infos, file, indent=4)\n",
" "
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"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": "3.10.15"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|