Add deepseek-ai/DeepSeek-OCR OCR results (50 samples) [deepseek-ocr]

README.md

@@ -6,6 +6,32 @@ tags:
 - markdown
 - uv-script
 - generated
+configs:
+- config_name: deepseek-ocr
+  data_files:
+  - split: train
+    path: deepseek-ocr/train-*
+dataset_info:
+  config_name: deepseek-ocr
+  features:
+  - name: image
+    dtype: image
+  - name: b_number
+    dtype: string
+  - name: page_index
+    dtype: int64
+  - name: source_row
+    dtype: int64
+  - name: markdown
+    dtype: string
+  - name: inference_info
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 20476090
+    num_examples: 50
+  download_size: 20340094
+  dataset_size: 20476090
 ---
 
 # Document OCR using LightOnOCR-2-1B

deepseek-ocr/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cc55dfd576682f4d2dfaa862255bf82a9bec66adb45d81390be42d8f3103773
+size 20340094