dataDatasic commited on
Commit
ff31cd9
·
verified ·
1 Parent(s): b68076f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +62 -5
README.md CHANGED
@@ -1,5 +1,62 @@
1
- ---
2
- license: other
3
- license_name: zion-vintage-media-proprietary
4
- license_link: LICENSE
5
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: zion-vintage-media-proprietary
4
+ license_link: LICENSE
5
+ task_categories:
6
+ - text-generation
7
+ - image-to-text
8
+ - visual-question-answering
9
+ language:
10
+ - en
11
+ tags:
12
+ - historical
13
+ - recipes
14
+ - multimodal
15
+ - ocr
16
+ - alternative-data
17
+ - Text-Generation
18
+ - Historical
19
+ size_categories:
20
+ - n<1K
21
+ pretty_name: Zion Vintage Recipes (Sample)
22
+ ---
23
+
24
+ # Zion Vintage Media: Historical Multimodal Culinary Archive (Sample)
25
+
26
+ ## 📌 Commercial Licensing & Full Access
27
+ **This repository contains a 50-item evaluation sample.** The full, production-ready dataset contains **10,000+ multimodal pairs** (High-Resolution Scans + Structured JSON) of unpublished, offline historical recipes spanning the 1940s to the 1990s.
28
+
29
+ To purchase a commercial license for the full SQL database and high-resolution TIF/JPG scans for AI/ML training, please contact:
30
+ * **Email:** admin@zionvintagemedia.com
31
+ * **Website:** [zionvintagemedia.com]Vintage store(https://zionvintagemedia.com)
32
+ * **Showcase:** [Dataset Explorer](https://ohcc-dataset-explore.vercel.app/)
33
+
34
+ ---
35
+
36
+ ## 📊 Dataset Overview
37
+ This dataset provides rare, Alternative Data for training Large Language Models (LLMs) and Vision-Language Models (VLMs). Because these physical, regional church cookbooks have never been published on the public internet, they provide completely novel tokens, historical dialects, archaic measurements, and procedural logic free from standard web-scraping contamination.
38
+
39
+ ### Ground Truth Quality
40
+ All text has been transcribed and structured from original physical media using high-resolution CZUR archival scanners, allowing for perfect Image-to-Text grounding.
41
+
42
+ ---
43
+
44
+ ## 🗄️ Data Structure (22 Highly-Enriched Columns)
45
+ This sample `metadata.csv` demonstrates our rigorous data engineering. It includes 22 structured columns:
46
+
47
+ **Recipe Data:**
48
+ * `recipe_id`, `contributor`, `contributor_role`
49
+ * `yield_amount`, `oven_temperature`
50
+ * `ingredients` (Serialized JSON)
51
+ * `instructions` & `instruction_steps` (Serialized JSON)
52
+ * `special_notes` & `data_quirks` (Captures historical oddities and OCR edge-cases)
53
+
54
+ **Historical Cookbook Context:**
55
+ * `book_title`, `book_author`, `publication_year`
56
+ * `location`, `sales_purpose`, `historical_notes`
57
+
58
+ **File Mapping:**
59
+ * `source_filename`, `source_filenames` (For multimodal image pairing)
60
+
61
+ ---
62
+ *© 2026 Bot and Dot Retail Management LLC / Zion Vintage Media.*