Update README.md

README.md

@@ -29,47 +29,64 @@ This dataset is used for the training of the [LiveCC-7B-Base](https://huggingfac
 - **Paper**: https://arxiv.org/abs/xxxx.xxxxx
 - **Training Code**: https://www.github.com/showlab/livecc
 
-### Data Sources
+### Live-CC-5M Dataset
+  
+  - Statistics: 5,047,208 YouTube Video-CC 30~240s instances.
+    ![image/png](https://cdn-uploads.huggingface.co/production/uploads/642435a1a3adbc7142c3b0a6/-RR-sI7F1a1XpxuQad2DH.png)
 
-- **Live-CC-5M**: This repository includes
   - Annotation JSONL (YouTube CC): https://huggingface.co/datasets/chenjoya/Live-CC-5M/blob/main/live_cc_5m_with_seeks.jsonl
+
+    Each line of the JSONL file is organized in a common user/assistant conversation format with a special "text_stream" key. Example:
+    ```
+    [
+      {"role": "user", "content": [{"type": "video", "video": "video/youtube/-4dnPeRv1ns.mp4", "video_start": 16.8, "video_end": 158.8}, {"type": "text", "text": "", "previous": "", "title": "Airsoft G&G Combat Machine M4 Review"}]},
+      {"role": "assistant", "content": [{"type": "text_stream", "text_stream": [[16.8, 16.9, "all"], [16.9, 17.0, "right"], [17.0, 17.1, "you"], [17.1, 17.3, "guys"], [17.3, 17.4, "so"], [17.4, 17.5, "this"], ...]}]}
+    ]
+    ```
+    - "title" denotes the YouTube title.
+    - "previous" denotes previous ASR content before "video_start".
+    - Each item in "text_stream" indicates start timestamp, end timestamp, and the word.
+
+    During pre-training, we use "title" and "previous" as context. Please refer to our dataloader (https://github.com/showlab/livecc/data/lmm_dataset.py) to learn how to make it compatible with popular LMMs (e.g. QwenVL series).
   
-  Due to 5M videos are too large, we are sorry that we cannot find way to share them. But,
-  - You can find all YouTube IDs in the annotation JSONL
-  - We have released video files in SFT set: https://huggingface.co/datasets/chenjoya/Live-WhisperX-528K
-
-  This dataset contains 5,047,208 YouTube Video-CC instances, with YouTube categories:
-  ```
-  {'People & Blogs': 1130380, 'Howto & Style': 889359, 'Entertainment': 585838, 'Education': 453305, 'Sports': 449166, 'Autos & Vehicles': 410228, 'Science & Technology': 345308, 'Film & Animation': 197529, 'Travel & Events': 151357, 'Pets & Animals': 127535, 'Gaming': 117149, 'News & Politics': 83560, 'Nonprofits & Activism': 59140, 'Comedy': 47354}
-  ```
-  Each line of the JSONL file is organized in a common user/assistant conversation format with a special "text_stream" key. Example:
-  ```
-  [
-    {"role": "user", "content": [{"type": "video", "video": "video/youtube/-4dnPeRv1ns.mp4", "video_start": 16.8, "video_end": 158.8}, {"type": "text", "text": "", "previous": "", "title": "Airsoft G&G Combat Machine M4 Review"}]},
-    {"role": "assistant", "content": [{"type": "text_stream", "text_stream": [[16.8, 16.9, "all"], [16.9, 17.0, "right"], [17.0, 17.1, "you"], [17.1, 17.3, "guys"], [17.3, 17.4, "so"], [17.4, 17.5, "this"], ...]}]}
-  ]
-  ```
-  - "title" denotes the YouTube title.
-  - "previous" denotes previous ASR content before "video_start".
-  - Each item in "text_stream" indicates start timestamp, end timestamp, and the word.
-  
-  During pre-training, we use "title" and "previous" as context. Please refer to our dataloader (https://github.com/showlab/livecc/data/lmm_dataset.py) to learn how to make it compatible with popular LMMs (e.g. QwenVL series).
-  
-  The last line of JSONL contains the file handle seek indices:
-  ```
-  [0, 3149, 7796, 10436, 18949, 22917, 41985, 65721, 73045, 76797, 82262, ...]
-  ```
-  This allows for easy streaming access using:
+    The last line of JSONL contains the file handle seek indices:
+    ```
+    [0, 3149, 7796, 10436, 18949, 22917, 41985, 65721, 73045, 76797, 82262, ...]
+    ```
+    This allows for easy streaming access using:
+    
+    ```python
+    import json
+
+    # read the last line of jsonl
+    def readlastline(path: str):
+      with open(path, "rb") as f:
+          f.seek(-2, 2) # avoid last \n
+          while f.read(1) != b"\n":  
+              f.seek(-2, 1)
+          return f.readline()
+    
+    # parse to seek indices list
+    seeks = json.loads(readlastline('live_cc_5m_with_seeks.jsonl'))
+    
+    # during data loader
+    with open('live_cc_5m_with_seeks.jsonl') as f:
+      f.seek(seeks[index])
+      datum = json.loads(f.readline())
+    ```
   
-  ```
-  with open('live_whisperx_528k_with_seeks.jsonl') as f:
-    f.seek(seek_index)
-    datum = json.loads(f.readline())
-  ```
+  - Videos: Due to 5M videos are too large, we are sorry that we cannot find way to share them. But,
+    - You can find all YouTube IDs in the annotation JSONL
+    - We have released video files in SFT dataset https://huggingface.co/datasets/chenjoya/Live-WhisperX-526K
+
 
 ### Data Pipeline
 
-Please refer to Section 3 of our [paper](https://arxiv.org/abs/xxxx.xxxxx). It has been fully open-sourced at: https://github.com/showlab/livecc/data/production/pretrain
+Please refer to Section 3 of our [paper](https://arxiv.org/abs/xxxx.xxxxx). 
+
+
+
+It has been fully open-sourced at: https://github.com/showlab/livecc/data/production/pretrain
 
 ## Citation