Remove fixtures DocVQA in favor of dataset on the hub

huggingface · Aug 30, 2021 · 2391ca5 · 2391ca5
1 parent a72f080
commit 2391ca5
Show file tree

Hide file tree

Showing 4 changed files with 11 additions and 3 deletions.
diff --git a/tests/fixtures/tests_samples/DocVQA/document.png b/tests/fixtures/tests_samples/DocVQA/document.png
diff --git a/tests/fixtures/tests_samples/DocVQA/document_2.png b/tests/fixtures/tests_samples/DocVQA/document_2.png
diff --git a/tests/test_feature_extraction_layoutlmv2.py b/tests/test_feature_extraction_layoutlmv2.py
@@ -185,7 +185,11 @@ def test_layoutlmv2_integration_test(self):
         # with apply_OCR = True
         feature_extractor = LayoutLMv2FeatureExtractor()
 
-        image = Image.open("tests/fixtures/tests_samples/DocVQA/document.png").convert("RGB")
+        from datasets import load_dataset
+
+        ds = load_dataset("hf-internal-testing/fixtures_docvqa", split="test")
+
+        image = Image.open(ds[0]["file"]).convert("RGB")
 
         encoding = feature_extractor(image, return_tensors="pt")
 

diff --git a/tests/test_processor_layoutlmv2.py b/tests/test_processor_layoutlmv2.py
@@ -141,8 +141,12 @@ class LayoutLMv2ProcessorIntegrationTests(unittest.TestCase):
     @cached_property
     def get_images(self):
         # we verify our implementation on 2 document images from the DocVQA dataset
-        image_1 = Image.open("tests/fixtures/tests_samples/DocVQA/document.png").convert("RGB")
-        image_2 = Image.open("tests/fixtures/tests_samples/DocVQA/document_2.png").convert("RGB")
+        from datasets import load_dataset
+
+        ds = load_dataset("hf-internal-testing/fixtures_docvqa", split="test")
+
+        image_1 = Image.open(ds[0]["file"]).convert("RGB")
+        image_2 = Image.open(ds[1]["file"]).convert("RGB")
 
         return image_1, image_2