Hebrew document-to-markdown OCR dataset v1.0

README.md

@@ -0,0 +1,41 @@
+---
+dataset_info:
+  features:
+  - name: image
+    dtype: image
+  - name: markdown
+    dtype: string
+  - name: doc_id
+    dtype: string
+  - name: page_num
+    dtype: int64
+  - name: doc_type
+    dtype: string
+  - name: orientation
+    dtype: string
+  - name: style
+    dtype: string
+  - name: title
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 1107433
+    num_examples: 46
+  - name: validation
+    num_bytes: 445459
+    num_examples: 15
+  - name: test
+    num_bytes: 558846
+    num_examples: 17
+  download_size: 2092818
+  dataset_size: 2111738
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: validation
+    path: data/validation-*
+  - split: test
+    path: data/test-*
+---

data/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f66b8784b005b6718bea7949a51b6e93b09752796af922dafa2c05b64e3adb8a
+size 555047

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fba71af9189293fee039d094af7b2856d8fd136587026755112988b856e028c
+size 1097418

data/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d73d6cb89913165eeaabbbc995ceb56efa19133a108ba1f35d97696b86c7ace
+size 440353