Add vision models to doc tests (#15905)

* Add vision models to doc tests

* Apply suggestions from code review

* Add more models

Co-authored-by: Niels Rogge <nielsrogge@Nielss-MacBook-Pro.local>

src/transformers/models/beit/modeling_beit.py

@@ -754,6 +754,7 @@ class BeitForMaskedImageModeling(BeitPreTrainedModel):
 
         ```python
         >>> from transformers import BeitFeatureExtractor, BeitForMaskedImageModeling
+        >>> import torch
         >>> from PIL import Image
         >>> import requests
 
@@ -763,9 +764,15 @@ class BeitForMaskedImageModeling(BeitPreTrainedModel):
         >>> feature_extractor = BeitFeatureExtractor.from_pretrained("microsoft/beit-base-patch16-224-pt22k")
         >>> model = BeitForMaskedImageModeling.from_pretrained("microsoft/beit-base-patch16-224-pt22k")
 
-        >>> inputs = feature_extractor(images=image, return_tensors="pt")
-        >>> outputs = model(**inputs)
-        >>> logits = outputs.logits
+        >>> num_patches = (model.config.image_size // model.config.patch_size) ** 2
+        >>> pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
+        >>> # create random boolean mask of shape (batch_size, num_patches)
+        >>> bool_masked_pos = torch.randint(low=0, high=2, size=(1, num_patches)).bool()
+
+        >>> outputs = model(pixel_values, bool_masked_pos=bool_masked_pos)
+        >>> loss, logits = outputs.loss, outputs.logits
+        >>> list(logits.shape)
+        [1, 196, 8192]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 

src/transformers/models/deit/modeling_deit.py

@@ -587,19 +587,25 @@ class DeiTForMaskedImageModeling(DeiTPreTrainedModel):
         Examples:
         ```python
         >>> from transformers import DeiTFeatureExtractor, DeiTForMaskedImageModeling
+        >>> import torch
         >>> from PIL import Image
         >>> import requests
 
         >>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
         >>> image = Image.open(requests.get(url, stream=True).raw)
 
-        >>> feature_extractor = DeiTFeatureExtractor.from_pretrained("google/vit-base-patch16-224-in21k")
-        >>> model = DeiTForMaskedImageModeling.from_pretrained("google/vit-base-patch16-224-in21k")
+        >>> feature_extractor = DeiTFeatureExtractor.from_pretrained("facebook/deit-base-distilled-patch16-224")
+        >>> model = DeiTForMaskedImageModeling.from_pretrained("facebook/deit-base-distilled-patch16-224")
 
-        >>> inputs = feature_extractor(images=image, return_tensors="pt")
+        >>> num_patches = (model.config.image_size // model.config.patch_size) ** 2
+        >>> pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
+        >>> # create random boolean mask of shape (batch_size, num_patches)
+        >>> bool_masked_pos = torch.randint(low=0, high=2, size=(1, num_patches)).bool()
 
-        >>> outputs = model(**inputs)
-        >>> last_hidden_states = outputs.last_hidden_state
+        >>> outputs = model(pixel_values, bool_masked_pos=bool_masked_pos)
+        >>> loss, reconstructed_pixel_values = outputs.loss, outputs.logits
+        >>> list(reconstructed_pixel_values.shape)
+        [1, 3, 224, 224]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 

src/transformers/models/swin/modeling_swin.py

@@ -810,6 +810,7 @@ class SwinForMaskedImageModeling(SwinPreTrainedModel):
         Examples:
         ```python
         >>> from transformers import AutoFeatureExtractor, SwinForMaskedImageModeling
+        >>> import torch
         >>> from PIL import Image
         >>> import requests
 
@@ -819,10 +820,15 @@ class SwinForMaskedImageModeling(SwinPreTrainedModel):
         >>> feature_extractor = AutoFeatureExtractor.from_pretrained("microsoft/swin-tiny-patch4-window7-224")
         >>> model = SwinForMaskedImageModeling.from_pretrained("microsoft/swin-tiny-patch4-window7-224")
 
-        >>> inputs = feature_extractor(images=image, return_tensors="pt")
+        >>> num_patches = (model.config.image_size // model.config.patch_size) ** 2
+        >>> pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
+        >>> # create random boolean mask of shape (batch_size, num_patches)
+        >>> bool_masked_pos = torch.randint(low=0, high=2, size=(1, num_patches)).bool()
 
-        >>> outputs = model(**inputs)
-        >>> last_hidden_states = outputs.last_hidden_state
+        >>> outputs = model(pixel_values, bool_masked_pos=bool_masked_pos)
+        >>> loss, reconstructed_pixel_values = outputs.loss, outputs.logits
+        >>> list(reconstructed_pixel_values.shape)
+        [1, 3, 224, 224]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 

src/transformers/models/vit/modeling_vit.py

@@ -624,6 +624,7 @@ class ViTForMaskedImageModeling(ViTPreTrainedModel):
         Examples:
         ```python
         >>> from transformers import ViTFeatureExtractor, ViTForMaskedImageModeling
+        >>> import torch
         >>> from PIL import Image
         >>> import requests
 
@@ -633,10 +634,15 @@ class ViTForMaskedImageModeling(ViTPreTrainedModel):
         >>> feature_extractor = ViTFeatureExtractor.from_pretrained("google/vit-base-patch16-224-in21k")
         >>> model = ViTForMaskedImageModeling.from_pretrained("google/vit-base-patch16-224-in21k")
 
-        >>> inputs = feature_extractor(images=image, return_tensors="pt")
+        >>> num_patches = (model.config.image_size // model.config.patch_size) ** 2
+        >>> pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
+        >>> # create random boolean mask of shape (batch_size, num_patches)
+        >>> bool_masked_pos = torch.randint(low=0, high=2, size=(1, num_patches)).bool()
 
-        >>> outputs = model(**inputs)
-        >>> last_hidden_states = outputs.last_hidden_state
+        >>> outputs = model(pixel_values, bool_masked_pos=bool_masked_pos)
+        >>> loss, reconstructed_pixel_values = outputs.loss, outputs.logits
+        >>> list(reconstructed_pixel_values.shape)
+        [1, 3, 224, 224]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 

utils/documentation_tests.txt

@@ -9,7 +9,16 @@ src/transformers/models/sew/modeling_sew.py
 src/transformers/models/sew_d/modeling_sew_d.py
 src/transformers/models/speech_to_text_2/modeling_speech_to_text_2.py
 src/transformers/models/speech_to_text/modeling_speech_to_text.py
-src/transformers/models/speech_encoder_decoder/modeling_speech_enocder_decoder.py
+src/transformers/models/speech_encoder_decoder/modeling_speech_encoder_decoder.py
 src/transformers/models/data2vec/modeling_data2vec_audio.py
+src/transformers/models/vit/modeling_vit.py
+src/transformers/models/beit/modeling_beit.py
+src/transformers/models/deit/modeling_deit.py
+src/transformers/models/swin/modeling_swin.py
+src/transformers/models/convnext/modeling_convnext.py
+src/transformers/models/poolformer/modeling_poolformer.py
+src/transformers/models/vit_mae/modeling_vit_mae.py
+src/transformers/models/segformer/modeling_segformer.py
+src/transformers/models/vision_encoder_decoder/modeling_vision_encoder_decoder.py
 docs/source/quicktour.mdx
-docs/source/task_summary.mdx
+docs/source/task_summary.mdx