A few CI fixes for `DocumentQuestionAnsweringPipeline` (#19584)

* Fixes

* update expected values

* style

* fix
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/pipelines/document_question_answering.py

@@ -235,7 +235,6 @@ class DocumentQuestionAnsweringPipeline(ChunkPipeline):
               `word_boxes`).
             - **answer** (`str`) -- The answer to the question.
             - **words** (`list[int]`) -- The index of each word/box pair that is in the answer
-            - **page** (`int`) -- The page of the answer
         """
         if isinstance(question, str):
             inputs = {"question": question, "image": image}
@@ -315,7 +314,6 @@ class DocumentQuestionAnsweringPipeline(ChunkPipeline):
                 "p_mask": None,
                 "word_ids": None,
                 "words": None,
-                "page": None,
                 "output_attentions": True,
                 "is_last": True,
             }
@@ -339,6 +337,7 @@ class DocumentQuestionAnsweringPipeline(ChunkPipeline):
                 return_overflowing_tokens=True,
                 **tokenizer_kwargs,
             )
+            encoding.pop("overflow_to_sample_mapping")  # We do not use this
 
             num_spans = len(encoding["input_ids"])
 
@@ -395,9 +394,6 @@ class DocumentQuestionAnsweringPipeline(ChunkPipeline):
         words = model_inputs.pop("words", None)
         is_last = model_inputs.pop("is_last", False)
 
-        if "overflow_to_sample_mapping" in model_inputs:
-            model_inputs.pop("overflow_to_sample_mapping")
-
         if self.model_type == ModelType.VisionEncoderDecoder:
             model_outputs = self.model.generate(**model_inputs)
         else:
@@ -421,7 +417,7 @@ class DocumentQuestionAnsweringPipeline(ChunkPipeline):
         return answers
 
     def postprocess_encoder_decoder_single(self, model_outputs, **kwargs):
-        sequence = self.tokenizer.batch_decode(model_outputs.sequences)[0]
+        sequence = self.tokenizer.batch_decode(model_outputs["sequences"])[0]
 
         # TODO: A lot of this logic is specific to Donut and should probably be handled in the tokenizer
         # (see https://github.com/huggingface/transformers/pull/18414/files#r961747408 for more context).

tests/pipelines/test_pipelines_document_question_answering.py

@@ -209,8 +209,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
         self.assertEqual(
             nested_simplify(outputs, decimals=4),
             [
-                {"score": 0.9967, "answer": "1102/2019", "start": 22, "end": 22},
-                {"score": 0.996, "answer": "us-001", "start": 15, "end": 15},
+                {"score": 0.9974, "answer": "1110212019", "start": 23, "end": 23},
+                {"score": 0.9948, "answer": "us-001", "start": 16, "end": 16},
             ],
         )
 
@@ -218,8 +218,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
         self.assertEqual(
             nested_simplify(outputs, decimals=4),
             [
-                {"score": 0.9967, "answer": "1102/2019", "start": 22, "end": 22},
-                {"score": 0.996, "answer": "us-001", "start": 15, "end": 15},
+                {"score": 0.9974, "answer": "1110212019", "start": 23, "end": 23},
+                {"score": 0.9948, "answer": "us-001", "start": 16, "end": 16},
             ],
         )
 
@@ -230,8 +230,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
             nested_simplify(outputs, decimals=4),
             [
                 [
-                    {"score": 0.9967, "answer": "1102/2019", "start": 22, "end": 22},
-                    {"score": 0.996, "answer": "us-001", "start": 15, "end": 15},
+                    {"score": 0.9974, "answer": "1110212019", "start": 23, "end": 23},
+                    {"score": 0.9948, "answer": "us-001", "start": 16, "end": 16},
                 ]
             ]
             * 2,
@@ -320,8 +320,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
         self.assertEqual(
             nested_simplify(outputs, decimals=4),
             [
-                {"score": 0.9999, "answer": "us-001", "start": 15, "end": 15},
-                {"score": 0.9924, "answer": "us-001", "start": 15, "end": 15},
+                {"score": 0.9999, "answer": "us-001", "start": 16, "end": 16},
+                {"score": 0.9998, "answer": "us-001", "start": 16, "end": 16},
             ],
         )
 
@@ -332,8 +332,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
             nested_simplify(outputs, decimals=4),
             [
                 [
-                    {"score": 0.9999, "answer": "us-001", "start": 15, "end": 15},
-                    {"score": 0.9924, "answer": "us-001", "start": 15, "end": 15},
+                    {"score": 0.9999, "answer": "us-001", "start": 16, "end": 16},
+                    {"score": 0.9998, "answer": "us-001", "start": 16, "end": 16},
                 ]
             ]
             * 2,
@@ -346,8 +346,8 @@ class DocumentQuestionAnsweringPipelineTests(unittest.TestCase, metaclass=Pipeli
         self.assertEqual(
             nested_simplify(outputs, decimals=4),
             [
-                {"score": 0.9999, "answer": "us-001", "start": 15, "end": 15},
-                {"score": 0.9924, "answer": "us-001", "start": 15, "end": 15},
+                {"score": 0.9999, "answer": "us-001", "start": 16, "end": 16},
+                {"score": 0.9998, "answer": "us-001", "start": 16, "end": 16},
             ],
         )