Cong123779 commited on
Commit
06dc987
·
verified ·
1 Parent(s): 1884b82

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +33 -29
README.md CHANGED
@@ -1,45 +1,49 @@
1
  ---
2
- title: Bigdata Project Assets
3
- emoji: 📂
4
  colorFrom: blue
5
  colorTo: indigo
6
  sdk: static
7
  pinned: false
8
  ---
9
 
10
- # Bigdata Project Assets 🚀
11
 
12
- This repository contains the essential data snapshots for the **Bigdata AI Document Processing System**. These assets facilitate the seamless migration and setup of the project across different machines.
13
 
14
- ## 📦 What's Included?
 
 
 
15
 
16
- Currently, this repository houses the **Core Database Snapshots** (the "Brain" of the RAG system):
17
 
18
- - **`infra_milvus_data.tar.gz`**: Vector database snapshots containing all indexed document embeddings (BGE-M3).
19
- - **`infra_postgres_data.tar.gz`**: Metadata and relational data including document references and system configurations.
20
- - **`infra_minio_data.tar.gz`**: Object storage metadata and small-scale processed assets.
21
- - **`infra_etcd_data.tar.gz`**: Configuration state for the Milvus cluster.
 
22
 
23
- ### Note on Data Size
24
- The raw source datasets (like **RVL-CDIP**, which is **48GB**) are **NOT** included in these primary snapshots to keep the migration process fast and efficient. These snapshots allow the RAG (Retrieval-Augmented Generation) system to remain fully functional for querying and chat, even without the massive raw source files.
25
 
26
- ## 🛠️ How to Restore
 
 
 
 
 
 
 
 
 
 
27
 
28
- These files are intended to be used with the `restore_from_hf.py` script found in the main project repository.
29
 
30
- 1. **Clone the Repo**:
31
- ```bash
32
- git clone https://github.com/congkx123789/bigdata_project.git
33
- ```
34
- 2. **Set Hugging Face Token**:
35
- ```bash
36
- export HF_TOKEN="your_token_here"
37
- ```
38
- 3. **Run Setup**:
39
- ```bash
40
- ./setup.sh
41
- # Select 'y' when prompted to restore data from Hugging Face.
42
- ```
43
 
44
- ## 👨‍💻 Maintainer
45
- Maintained by **Cong123779**.
 
1
  ---
2
+ title: Bigdata Project Assets (The Brain)
3
+ emoji: 🧠
4
  colorFrom: blue
5
  colorTo: indigo
6
  sdk: static
7
  pinned: false
8
  ---
9
 
10
+ # Bigdata Project Assets (The "Brain") 🧠
11
 
12
+ Đây repository lưu trữ các bản sao lưu (snapshots) sở dữ liệu cho hệ thống **Bigdata AI Document Processing**.
13
 
14
+ ## 🎯 Mục đích & Ý nghĩa
15
+ Trong các hệ thống AI xử lý dữ liệu lớn, việc tách biệt **Dữ liệu thô (Raw Data)** và **Dữ liệu đã xử lý (Processed/Brain Data)** là cực kỳ quan trọng:
16
+ - **Raw Data (~48GB)**: Là các file ảnh/PDF ban đầu. Việc tải và xử lý lại từ đầu tốn rất nhiều thời gian.
17
+ - **Brain Data (Vài MB - GB)**: Là kết quả của quá trình index (Vector embeddings, Metadata). Đây là những gì được lưu trữ tại repository này.
18
 
19
+ **=> Lợi ích:** Bạn thể setup máy mới và Chat được ngay lập tức chỉ bằng cách tải "Bộ não" này về, không cần tải 48GB dữ liệu thô ngay lúc đó.
20
 
21
+ ## 📦 Danh sách Assets
22
+ - **`infra_milvus_data.tar.gz`**: Chứa toàn bộ Vector Embeddings của tài liệu (BGE-M3).
23
+ - **`infra_postgres_data.tar.gz`**: Chứa thông tin Metadata, cấu hình hệ thống và lịch sử.
24
+ - **`infra_minio_data.tar.gz`**: Chứa metadata của Object Storage và các file đã xử lý quy mô nhỏ.
25
+ - **`infra_etcd_data.tar.gz`**: Trạng thái cấu hình của cluster Milvus.
26
 
27
+ ## 🛠Quy trình khôi phục nhanh (Quick Restore)
28
+ Các assets này được thiết kế để hoạt động cùng với script `setup.sh` trong nguồn chính.
29
 
30
+ 1. **Clone nguồn từ GitHub**:
31
+ ```bash
32
+ git clone https://github.com/congkx123789/bigdata_project.git
33
+ cd bigdata_project
34
+ ```
35
+ 2. **Chạy script Setup thông minh**:
36
+ ```bash
37
+ export HF_TOKEN="your_huggingface_token_here"
38
+ ./setup.sh
39
+ ```
40
+ 3. **Xác nhận Restore**: Khi script hỏi `Do you want to restore data from Hugging Face?`, hãy chọn `y`.
41
 
42
+ ---
43
 
44
+ ## 🔗 Liên kết quan trọng
45
+ - **Mã nguồn chính (Github)**: [bigdata_project](https://github.com/congkx123789/bigdata_project)
46
+ - **Tác giả**: Cong123779
 
 
 
 
 
 
 
 
 
 
47
 
48
+ ---
49
+ *Lưu ý: Nếu bạn muốn tải dữ liệu thô (48GB RVL-CDIP), hãy sử dụng công cụ `download_datasets.sh` có sẵn trong repository GitHub.*