Skip Perceiver tests

src/transformers/models/auto/feature_extraction_auto.py

@@ -43,6 +43,7 @@
         ("detr", "DetrFeatureExtractor"),
         ("layoutlmv2", "LayoutLMv2FeatureExtractor"),
         ("clip", "CLIPFeatureExtractor"),
+        ("perceiver", "PerceiverFeatureExtractor"),
     ]
 )
 

src/transformers/models/perceiver/tokenization_perceiver.py

@@ -87,7 +87,7 @@ def __init__(
         self._utf_vocab_size = 2 ** 8  # utf is 8 bits
 
         # define special tokens dict
-        self.special_tokens_encoder: Dict[int, str] = {
+        self.special_tokens_encoder: Dict[str, int] = {
             self.pad_token: 0,
             self.bos_token: 1,
             self.eos_token: 2,
@@ -96,7 +96,15 @@ def __init__(
             self.sep_token: 5,
         }
         self._num_special_tokens = len(self.special_tokens_encoder)
-        self.special_tokens_decoder: Dict[str, int] = {v: k for k, v in self.special_tokens_encoder.items()}
+        self.special_tokens_decoder: Dict[int, str] = {v: k for k, v in self.special_tokens_encoder.items()}
+
+    def get_vocab(self) -> Dict[str, int]:
+        vocab = self.special_tokens_encoder.copy()
+        vocab.update(self.added_tokens_encoder)
+        for i in range(self._utf_vocab_size):
+            token = chr(i)
+            vocab[token] = i + len(self.special_tokens_encoder)
+        return vocab
 
     @property
     def vocab_size(self):

tests/test_pipelines_common.py

@@ -169,6 +169,11 @@ def test(self):
                 else:
                     tokenizer = None
                 feature_extractor = get_tiny_feature_extractor_from_checkpoint(checkpoint, tiny_config)
+
+                if tokenizer is None and feature_extractor is None:
+                    self.skipTest(
+                        f"Ignoring {ModelClass}, cannot create a tokenizer or feature_extractor (PerceiverConfig with no FastTokenizer ?)"
+                    )
                 pipeline, examples = self.get_test_pipeline(model, tokenizer, feature_extractor)
                 if pipeline is None:
                     # The test can disable itself, but it should be very marginal
@@ -213,6 +218,7 @@ def data(n):
                         if not tokenizer_classes:
                             # We need to test even if there are no tokenizers.
                             tokenizer_classes = [None]
+
                         for tokenizer_class in tokenizer_classes:
                             if tokenizer_class is not None:
                                 tokenizer_name = tokenizer_class.__name__

tests/test_pipelines_image_classification.py

@@ -14,7 +14,12 @@
 
 import unittest
 
-from transformers import MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING, PreTrainedTokenizer, is_vision_available
+from transformers import (
+    MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING,
+    PerceiverConfig,
+    PreTrainedTokenizer,
+    is_vision_available,
+)
 from transformers.pipelines import ImageClassificationPipeline, pipeline
 from transformers.testing_utils import (
     is_pipeline_test,
@@ -45,6 +50,10 @@ class ImageClassificationPipelineTests(unittest.TestCase, metaclass=PipelineTest
     model_mapping = MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING
 
     def get_test_pipeline(self, model, tokenizer, feature_extractor):
+        if isinstance(model.config, PerceiverConfig):
+            self.skipTest(
+                "Perceiver model tester is defined with a language one, which has no feature_extractor, so the automated test cannot work here"
+            )
 
         image_classifier = ImageClassificationPipeline(model=model, feature_extractor=feature_extractor)
         examples = [