Upload README.md with huggingface_hub

README.md

@@ -13,7 +13,7 @@ size_categories:
 - 1M<n<10M
 ---
 
-# Multimodal-Textbook
+# Multimodal-Textbook-6.5M
 <img src="./src/logo.png" alt="Image" style="width: 900px;">  
 
 [![arXiv](https://img.shields.io/badge/arXiv-Paper-<COLOR>.svg)](https://arxiv.org/abs/2306.07209) [![Project](https://img.shields.io/badge/Project-Website-blue.svg)](https://multi-modal-self-instruct.github.io) [![GitHub](https://img.shields.io/badge/GitHub-Code-181717?logo=github)](https://github.com/DAMO-NLP-SG/multimodal_textbook/tree/master)
@@ -23,7 +23,8 @@ size_categories:
 
 ## Overview
 
-This dataset is for ["2.5 Years in Class: A Multimodal Textbook for Vision-Language Pretraining"](https://arxiv.org/pdf/2306.07209). 
+This dataset is for ["2.5 Years in Class: A Multimodal Textbook for Vision-Language Pretraining"](https://arxiv.org/pdf/2306.07209), containing 6.5M images interleaving with 0.8B text from instructional videos.
+
 - It contains **pre-training corpus using interleaved image-text format**. Specifically, our multimodal-textbook includes **6.5M keyframes** extracted from instructional videos, interleaving with 0.8B **ASR texts**.
 - All the images and text are extracted from online instructional videos (22,000 class hours), covering multiple fundamental subjects, e.g., mathematics, physics, and chemistry. 
 - Our textbook corpus providing a more coherent context and richer knowledge for image-text aligning. 
@@ -71,12 +72,16 @@ show_interleaved_dataset.ipynb
 In the notebook, you can see keyframes interleaving with text.
 
 
+## Dataset Statistics
+
 
 
 ## Using Our Dataset
+### Dataset
 We provide the json file and corresponding images folder for textbook: 
-- json file: `multimodal_textbook.json` (610k samples ~ 11GB)
-- image_folder: `dataset_images_interval_7.tar.gz` (6.5M image ~ 700GB)
+- Dataset json-file: `./multimodal_textbook.json` (610k samples ~ 11GB)
+- Dataset image_folder: `./dataset_images_interval_7.tar.gz` (6.5M image ~ 700GB)
+- videometa_data: `video_meta_data/video_meta_data1.json` and `video_meta_data/video_meta_data2.json` represent the meta information of crawled videos, including video vid, title, description, duration, language, and searched knowledge points. `multimodal_textbook_meta_data.json.zip` records the textbook in its original format, not in the OBELICS format.
 
 Each sample has approximately 10.7 images and 1927 text tokens. After you download and unzip the folder, you need to replace the each image path in json file (`/mnt/workspace/zwq_data/interleaved_dataset/`) with your personal image folder path.
 
@@ -98,6 +103,7 @@ Each sample has approximately 10.7 images and 1927 text tokens. After you downlo
 ```
 
 
+
 ### Naming Format for keyframe  
 
 For each keyframe, its naming format rule is:   
@@ -110,6 +116,73 @@ This means that this image is extracted from the video (`-1uixJ1V-As`), more spe
 
 
 
+### MetaData of Instructional Video
+The format of the `video_meta_data/video_meta_data1.json`: 
+```
+    {
+        "file_path": xxx,
+        "file_size (MB)": 85.54160022735596,
+        "file_name": "-r7-s1z3lFY.mp4",
+        "video_duration": 0,
+        "unique": true,
+        "asr_path": xxxx,
+        "asr_len": 2990,
+        "caption_path": xxx,
+        "caption_len": 0,
+        "search_keyword": "1.3B parameter size models comparison",
+        "title": "DeepSeek Coder LLM | A Revolutionary Coder Model",
+        "desc": "In this video, we are going to test out Deepseek Coder, a coding LLM.....,
+        "llm_response": " The video appears to be a detailed and technical analysis of DeepSeek Coder LLM..... ###Score: 10###",
+        "language": "en",
+        "asr is repetive": false,
+        "deepseek_score": 10,
+        "llama_score": 2,
+        "deepseek_score long context": 10
+    },
+```
 
+In addition, the `multimodal_textbook_meta_data.json.zip` records the textbook in its original format as follows.It is stored with "video clip" as a dict. Each sample includes multiple video clips: 
 
-
+```
+{'token_num': 1657,
+ 'conversations': [
+    {
+        'vid': video id-1,
+        'clip_path': the path of video clip,
+        'asr': ASR transcribed from audio,
+        'extracted_frames': Extract keyframe sequences according to time intervals.,
+        'image_tokens': xxx,
+        'token_num': xxx,
+        'refined_asr': Refine the original ASR,
+        'ocr_internvl_8b': OCR obtained using internvl_8b,
+        'ocr_image': the image does OCR come from,
+        'ocr_internvl_8b_deduplicates': xxx,
+        'keyframe_ssim': Keyframe sequence extracted according to SSIM algorithm.,
+        'asr_token_num': xxx,
+        'ocr_qwen2_vl_72b': 'OCR obtained using qwen2_vl_72b'
+   },
+   {
+        'vid': 'End of a Video',
+        'clip_path': xxxx,
+        'image_tokens': 0,
+        'token_num': 0
+   },
+   {
+        'vid': video id-2,
+        'clip_path': the path of video clip,
+        'asr': ASR transcribed from audio,
+        'extracted_frames': Extract keyframe sequences according to time intervals.,
+        'image_tokens': xxx,
+        'token_num': xxx,
+        'refined_asr': Refine the original ASR,
+        'ocr_internvl_8b': OCR obtained using internvl_8b,
+        'ocr_image': the image does OCR come from,
+        'ocr_internvl_8b_deduplicates': xxx,
+        'keyframe_ssim': Keyframe sequence extracted according to SSIM algorithm.,
+        'asr_token_num': xxx,
+        'ocr_qwen2_vl_72b': 'OCR obtained using qwen2_vl_72b'
+   },
+    ....
+   ]
+}
+```