Add README with stats

README.md

@@ -1,131 +1,81 @@
 ---
-dataset_info:
-  features:
-  - name: image
-    dtype: image
-  - name: text
-    dtype: string
-  - name: image_name
-    dtype: string
-  splits:
-  - name: Assamese
-    num_bytes: 233530
-    num_examples: 10
-  - name: Bengali
-    num_bytes: 284932
-    num_examples: 10
-  - name: Bodo
-    num_bytes: 78893
-    num_examples: 10
-  - name: Dogri
-    num_bytes: 214391
-    num_examples: 10
-  - name: English
-    num_bytes: 524712
-    num_examples: 10
-  - name: Gujarati
-    num_bytes: 661372
-    num_examples: 10
-  - name: Hindi
-    num_bytes: 1260712
-    num_examples: 10
-  - name: Kannada
-    num_bytes: 3649720
-    num_examples: 10
-  - name: Kashmiri
-    num_bytes: 164857
-    num_examples: 10
-  - name: Konkani
-    num_bytes: 227290
-    num_examples: 10
-  - name: Maithili
-    num_bytes: 379119
-    num_examples: 10
-  - name: Malayalam
-    num_bytes: 408148
-    num_examples: 10
-  - name: Manipuri
-    num_bytes: 348078
-    num_examples: 10
-  - name: Marathi
-    num_bytes: 993890
-    num_examples: 10
-  - name: Nepali
-    num_bytes: 2437528
-    num_examples: 10
-  - name: Odia
-    num_bytes: 1998933
-    num_examples: 10
-  - name: Punjabi
-    num_bytes: 873083
-    num_examples: 10
-  - name: Sanskrit
-    num_bytes: 83400
-    num_examples: 10
-  - name: Santhali
-    num_bytes: 851665
-    num_examples: 10
-  - name: Sindhi
-    num_bytes: 499798
-    num_examples: 10
-  - name: Tamil
-    num_bytes: 1900952
-    num_examples: 10
-  - name: Telugu
-    num_bytes: 827986
-    num_examples: 10
-  - name: Urdu
-    num_bytes: 476467
-    num_examples: 10
-  download_size: 19369992
-  dataset_size: 19379456
-configs:
-- config_name: default
-  data_files:
-  - split: Assamese
-    path: data/Assamese-*
-  - split: Bengali
-    path: data/Bengali-*
-  - split: Bodo
-    path: data/Bodo-*
-  - split: Dogri
-    path: data/Dogri-*
-  - split: English
-    path: data/English-*
-  - split: Gujarati
-    path: data/Gujarati-*
-  - split: Hindi
-    path: data/Hindi-*
-  - split: Kannada
-    path: data/Kannada-*
-  - split: Kashmiri
-    path: data/Kashmiri-*
-  - split: Konkani
-    path: data/Konkani-*
-  - split: Maithili
-    path: data/Maithili-*
-  - split: Malayalam
-    path: data/Malayalam-*
-  - split: Manipuri
-    path: data/Manipuri-*
-  - split: Marathi
-    path: data/Marathi-*
-  - split: Nepali
-    path: data/Nepali-*
-  - split: Odia
-    path: data/Odia-*
-  - split: Punjabi
-    path: data/Punjabi-*
-  - split: Sanskrit
-    path: data/Sanskrit-*
-  - split: Santhali
-    path: data/Santhali-*
-  - split: Sindhi
-    path: data/Sindhi-*
-  - split: Tamil
-    path: data/Tamil-*
-  - split: Telugu
-    path: data/Telugu-*
-  - split: Urdu
-    path: data/Urdu-*
+license: cc-by-4.0
+task_categories:
+- image-to-text
+language:
+- as
+- bn
+- brx
+- doi
+- en
+- gu
+- hi
+- kn
+- ks
+- gom
+- mai
+- ml
+- mni
+- mr
+- ne
+- or
+- pa
+- sa
+- sat
+- sd
+- ta
+- te
+- ur
+pretty_name: IndicDLP Block-Level OCR Eval Set (Sample)
+size_categories:
+- n<1K
 ---
+
+# IndicDLP Block-Level OCR Eval Set — Sample
+
+This is a **public sample** (10 entries per language) of the full private dataset [`onath/indicdlp-eval-set-22`](https://huggingface.co/datasets/onath/indicdlp-eval-set-22).
+
+Each example contains a cropped text-block image and its ground-truth transcription, covering **23 Indian languages**.
+
+## Dataset Stats
+
+| Language | Sample Rows | Full Dataset Rows |
+|----------|-------------|-------------------|
+| Assamese | 10 | 772 |
+| Bengali | 10 | 751 |
+| Bodo | 10 | 887 |
+| Dogri | 10 | 960 |
+| English | 10 | 915 |
+| Gujarati | 10 | 540 |
+| Hindi | 10 | 581 |
+| Kannada | 10 | 549 |
+| Kashmiri | 10 | 839 |
+| Konkani | 10 | 767 |
+| Maithili | 10 | 900 |
+| Malayalam | 10 | 954 |
+| Manipuri | 10 | 935 |
+| Marathi | 10 | 612 |
+| Nepali | 10 | 755 |
+| Odia | 10 | 480 |
+| Punjabi | 10 | 558 |
+| Sanskrit | 10 | 696 |
+| Santhali | 10 | 854 |
+| Sindhi | 10 | 637 |
+| Tamil | 10 | 573 |
+| Telugu | 10 | 582 |
+| Urdu | 10 | 876 |
+
+**Total sample rows:** 230
+**Total full dataset rows:** 16973
+
+## Columns
+
+| Column | Type | Description |
+|--------|------|-------------|
+| `image` | Image | Cropped block-level image |
+| `text` | string | Ground-truth transcription |
+| `image_name` | string | Original relative file path |
+
+## Splits
+
+One split per language: `Assamese, Bengali, Bodo, Dogri, English, Gujarati, Hindi, Kannada, Kashmiri, Konkani, Maithili, Malayalam, Manipuri, Marathi, Nepali, Odia, Punjabi, Sanskrit, Santhali, Sindhi, Tamil, Telugu, Urdu`.