Upload README.md with huggingface_hub

README.md

@@ -1,47 +1,36 @@
 ---
-dataset_info:
-  features:
-  - name: chunk_id
-    dtype: string
-  - name: doc_id
-    dtype: string
-  - name: doc_title
-    dtype: string
-  - name: doc_type
-    dtype: string
-  - name: section_title
-    dtype: string
-  - name: section_level
-    dtype: int32
-  - name: semantic_label
-    dtype: string
-  - name: text
-    dtype: string
-  - name: page
-    dtype: int32
-  - name: word_count
-    dtype: int32
-  - name: char_count
-    dtype: int32
-  splits:
-  - name: train
-    num_bytes: 248719
-    num_examples: 635
-  - name: validation
-    num_bytes: 30943
-    num_examples: 79
-  - name: test
-    num_bytes: 31334
-    num_examples: 80
-  download_size: 138142
-  dataset_size: 310996
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: validation
-    path: data/validation-*
-  - split: test
-    path: data/test-*
+language:
+- en
+license: apache-2.0
+task_categories:
+- text-generation
+task_ids:
+- text2text-generation
+pretty_name: Amtp Chunks V1
+size_categories:
+- n<1K
 ---
+
+# Amtp Chunks V1
+
+Amtp Chunks V1 is a high-quality document dataset generated by [DocParserEngine](https://github.com/remixonwin/data-parser).
+
+## Dataset Summary
+- **Documents Processed**: 1
+- **Total Records**: 794
+- **Schema Format**: `chunks`
+- **Extraction Features**: Structural detection, image extraction, AI-powered captioning, and OCR.
+
+## Supported Tasks
+- **OCR & Text Extraction**: High-accuracy text extraction from complex document layouts.
+- **Image Captioning & Categorization**: Vision-based descriptions and classification of extracted images.
+- **RAG & Information Retrieval**: (When using `chunks` schema) Ideal for building knowledge bases.
+
+## Dataset Structure
+The dataset follows a standard `datasets.DatasetDict` structure with `train`, `validation`, and `test` splits.
+
+### Schema: `chunks`
+Detailed feature descriptions can be found in the dataset configuration on the Hugging Face Hub.
+
+## Maintenance
+Generated and maintained using **DocParserEngine**.