all tests are passing locally

tests/models/vision_encoder_decoder/test_modeling_tf_vision_encoder_decoder.py

@@ -466,11 +466,15 @@ def check_pt_tf_equivalence(self, tf_model, pt_model, tf_inputs_dict):
         self.check_pt_tf_models(tf_model, pt_model, tf_inputs_dict)
 
     def check_pt_to_tf_equivalence(self, config, decoder_config, tf_inputs_dict):
+        if _run_slow_tests:
+            config._attn_implementation = "eager"
+            decoder_config._attn_implementation = "eager"
+
         encoder_decoder_config = VisionEncoderDecoderConfig.from_encoder_decoder_configs(config, decoder_config)
         # Output all for aggressive testing
         encoder_decoder_config.output_hidden_states = True
         # All models tested in this file have attentions
-        encoder_decoder_config.output_attentions = True
+        encoder_decoder_config.output_attentions = _run_slow_tests
 
         pt_model = VisionEncoderDecoderModel(encoder_decoder_config)
 
@@ -481,11 +485,17 @@ def check_pt_to_tf_equivalence(self, config, decoder_config, tf_inputs_dict):
         self.check_pt_tf_equivalence(tf_model, pt_model, tf_inputs_dict)
 
     def check_tf_to_pt_equivalence(self, config, decoder_config, tf_inputs_dict):
+        # When taking a model from tf we are using the default attention
+        # mode (sdpa) so we are not expecting attention
+        config_output_attention = config.output_attentions
+        config.output_attentions = False
+
         encoder_decoder_config = VisionEncoderDecoderConfig.from_encoder_decoder_configs(config, decoder_config)
+
         # Output all for aggressive testing
         encoder_decoder_config.output_hidden_states = True
         # TODO: A generalizable way to determine this attribute
-        encoder_decoder_config.output_attentions = True
+        encoder_decoder_config.output_attentions = False
 
         tf_model = TFVisionEncoderDecoderModel(encoder_decoder_config)
         # Make sure model is built before saving
@@ -496,6 +506,8 @@ def check_tf_to_pt_equivalence(self, config, decoder_config, tf_inputs_dict):
             pt_model = VisionEncoderDecoderModel.from_pretrained(tmpdirname, from_tf=True)
 
         self.check_pt_tf_equivalence(tf_model, pt_model, tf_inputs_dict)
+        # Revert mutable objet modification
+        config.output_attentions = config_output_attention
 
     def test_encoder_decoder_model(self):
         config_inputs_dict = self.prepare_config_and_inputs()

tests/models/vision_encoder_decoder/test_modeling_vision_encoder_decoder.py

@@ -324,6 +324,7 @@ def test_save_and_load_from_encoder_decoder_pretrained(self):
         input_ids_dict = self.prepare_config_and_inputs()
         self.check_save_and_load_encoder_decoder_model(**input_ids_dict)
 
+    @slow
     def test_encoder_decoder_model_output_attentions(self):
         input_ids_dict = self.prepare_config_and_inputs()
         self.check_encoder_decoder_model_output_attentions(**input_ids_dict)
@@ -412,9 +413,6 @@ def check_encoder_decoder_model_output_attentions(
         pixel_values=None,
         **kwargs,
     ):
-        if not _run_slow_tests:
-            return
-
         # make the decoder inputs a different shape from the encoder inputs to harden the test
         decoder_input_ids = decoder_input_ids[:, :-1]
         decoder_attention_mask = decoder_attention_mask[:, :-1]
@@ -461,6 +459,8 @@ def check_encoder_decoder_model_output_attentions(
         )
 
     def get_encoder_decoder_model(self, config, decoder_config):
+        if _run_slow_tests:
+            config._attn_implementation = "eager"
         encoder_model = DeiTModel(config).eval()
         decoder_model = BertLMHeadModel(decoder_config).eval()
         return encoder_model, decoder_model
@@ -526,6 +526,8 @@ def get_pretrained_model_and_inputs(self):
         return model, inputs
 
     def get_encoder_decoder_model(self, config, decoder_config):
+        if _run_slow_tests:
+            config._attn_implementation = "eager"
         encoder_model = ViTModel(config).eval()
         decoder_model = BertLMHeadModel(decoder_config).eval()
         return encoder_model, decoder_model
@@ -565,29 +567,6 @@ def prepare_config_and_inputs(self):
             "labels": decoder_token_labels,
         }
 
-    def check_encoder_decoder_model_output_attentions(
-        self,
-        config,
-        decoder_config,
-        decoder_input_ids,
-        decoder_attention_mask,
-        labels=None,
-        pixel_values=None,
-        **kwargs,
-    ):
-        if not _run_slow_tests:
-            return
-
-        super().check_encoder_decoder_model_output_attentions(
-            config,
-            decoder_config,
-            decoder_input_ids,
-            decoder_attention_mask,
-            labels=None,
-            pixel_values=None,
-            **kwargs,
-        )
-
 
 @require_torch
 class Swin2BartModelTest(EncoderDecoderMixin, unittest.TestCase):
@@ -677,6 +656,8 @@ def test_real_model_save_load_from_pretrained(self):
 @require_torch
 class ViT2TrOCR(EncoderDecoderMixin, unittest.TestCase):
     def get_encoder_decoder_model(self, config, decoder_config):
+        if _run_slow_tests:
+            config._attn_implementation = "eager"
         encoder_model = ViTModel(config).eval()
         decoder_model = TrOCRForCausalLM(decoder_config).eval()
         return encoder_model, decoder_model
@@ -704,29 +685,6 @@ def prepare_config_and_inputs(self):
             "labels": decoder_input_ids,
         }
 
-    def check_encoder_decoder_model_output_attentions(
-        self,
-        config,
-        decoder_config,
-        decoder_input_ids,
-        decoder_attention_mask,
-        labels=None,
-        pixel_values=None,
-        **kwargs,
-    ):
-        if not _run_slow_tests:
-            return
-
-        super().check_encoder_decoder_model_output_attentions(
-            config,
-            decoder_config,
-            decoder_input_ids,
-            decoder_attention_mask,
-            labels=None,
-            pixel_values=None,
-            **kwargs,
-        )
-
     # there are no published pretrained TrOCR checkpoints for now
     def test_real_model_save_load_from_pretrained(self):
         pass