add uniform processors for altclip + chinese_clip

src/transformers/models/align/processing_align.py

@@ -16,12 +16,13 @@
 Image/Text processor class for ALIGN
 """
 
+import sys
 from typing import List, Union
 
 
-try:
+if sys.version_info >= (3, 11):
     from typing import Unpack
-except ImportError:
+else:
     from typing_extensions import Unpack
 
 from ...image_utils import ImageInput

src/transformers/models/altclip/processing_altclip.py

@@ -16,10 +16,23 @@
 Image/Text processor class for AltCLIP
 """
 
-import warnings
+import sys
+from typing import List, Union
 
-from ...processing_utils import ProcessorMixin
-from ...tokenization_utils_base import BatchEncoding
+
+if sys.version_info >= (3, 11):
+    from typing import Unpack
+else:
+    from typing_extensions import Unpack
+
+from ...image_utils import ImageInput
+from ...processing_utils import ProcessingKwargs, ProcessorMixin
+from ...tokenization_utils_base import BatchEncoding, PreTokenizedInput, TextInput
+from ...utils.deprecation import deprecate_kwarg
+
+
+class AltClipProcessorKwargs(ProcessingKwargs, total=False):
+    _defaults = {}
 
 
 class AltCLIPProcessor(ProcessorMixin):
@@ -41,25 +54,23 @@ class AltCLIPProcessor(ProcessorMixin):
     image_processor_class = "CLIPImageProcessor"
     tokenizer_class = ("XLMRobertaTokenizer", "XLMRobertaTokenizerFast")
 
-    def __init__(self, image_processor=None, tokenizer=None, **kwargs):
-        feature_extractor = None
-        if "feature_extractor" in kwargs:
-            warnings.warn(
-                "The `feature_extractor` argument is deprecated and will be removed in v5, use `image_processor`"
-                " instead.",
-                FutureWarning,
-            )
-            feature_extractor = kwargs.pop("feature_extractor")
-
-        image_processor = image_processor if image_processor is not None else feature_extractor
+    @deprecate_kwarg(old_name="feature_extractor", version="5.0.0", new_name="image_processor")
+    def __init__(self, image_processor=None, tokenizer=None):
         if image_processor is None:
             raise ValueError("You need to specify an `image_processor`.")
         if tokenizer is None:
             raise ValueError("You need to specify a `tokenizer`.")
 
         super().__init__(image_processor, tokenizer)
 
-    def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
+    def __call__(
+        self,
+        images: ImageInput = None,
+        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
+        audio=None,
+        videos=None,
+        **kwargs: Unpack[AltClipProcessorKwargs],
+    ) -> BatchEncoding:
         """
         Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text`
         and `kwargs` arguments to XLMRobertaTokenizerFast's [`~XLMRobertaTokenizerFast.__call__`] if `text` is not
@@ -68,22 +79,20 @@ def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
         of the above two methods for more information.
 
         Args:
-            text (`str`, `List[str]`, `List[List[str]]`):
+
+            images (`ImageInput`):
+                The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch
+                tensor. Both channels-first and channels-last formats are supported.
+            text (`TextInput`, `PreTokenizedInput`, `List[TextInput]`, `List[PreTokenizedInput]`):
                 The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
                 (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
                 `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
-            images (`PIL.Image.Image`, `np.ndarray`, `torch.Tensor`, `List[PIL.Image.Image]`, `List[np.ndarray]`, `List[torch.Tensor]`):
-                The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch
-                tensor. Both channels-first and channels-last formats are supported.
-
             return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors of a particular framework. Acceptable values are:
-
-                - `'tf'`: Return TensorFlow `tf.constant` objects.
-                - `'pt'`: Return PyTorch `torch.Tensor` objects.
-                - `'np'`: Return NumPy `np.ndarray` objects.
-                - `'jax'`: Return JAX `jnp.ndarray` objects.
-
+                    - `'tf'`: Return TensorFlow `tf.constant` objects.
+                    - `'pt'`: Return PyTorch `torch.Tensor` objects.
+                    - `'np'`: Return NumPy `np.ndarray` objects.
+                    - `'jax'`: Return JAX `jnp.ndarray` objects.
         Returns:
             [`BatchEncoding`]: A [`BatchEncoding`] with the following fields:
 
@@ -95,13 +104,24 @@ def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
         """
 
         if text is None and images is None:
-            raise ValueError("You have to specify either text or images. Both cannot be none.")
+            raise ValueError("You must specify either text or images.")
 
-        if text is not None:
-            encoding = self.tokenizer(text, return_tensors=return_tensors, **kwargs)
+        if text is None and images is None:
+            raise ValueError("You must specify either text or images.")
+        output_kwargs = self._merge_kwargs(
+            AltClipProcessorKwargs,
+            tokenizer_init_kwargs=self.tokenizer.init_kwargs,
+            **kwargs,
+        )
 
+        if text is not None:
+            encoding = self.tokenizer(text, **output_kwargs["text_kwargs"])
         if images is not None:
-            image_features = self.image_processor(images, return_tensors=return_tensors, **kwargs)
+            image_features = self.image_processor(images, **output_kwargs["images_kwargs"])
+
+        # BC for explicit return_tensors
+        if "return_tensors" in output_kwargs["common_kwargs"]:
+            return_tensors = output_kwargs["common_kwargs"].pop("return_tensors", None)
 
         if text is not None and images is not None:
             encoding["pixel_values"] = image_features.pixel_values

src/transformers/models/chinese_clip/image_processing_chinese_clip.py

@@ -249,6 +249,7 @@ def preprocess(
                 - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format.
                 - `"none"` or `ChannelDimension.NONE`: image in (height, width) format.
         """
+
         do_resize = do_resize if do_resize is not None else self.do_resize
         size = size if size is not None else self.size
         size = get_size_dict(size, default_to_square=False)

src/transformers/models/chinese_clip/processing_chinese_clip.py

@@ -16,10 +16,24 @@
 Image/Text processor class for Chinese-CLIP
 """
 
+import sys
 import warnings
+from typing import List, Union
 
-from ...processing_utils import ProcessorMixin
-from ...tokenization_utils_base import BatchEncoding
+from ...image_utils import ImageInput
+from ...processing_utils import ProcessingKwargs, ProcessorMixin
+
+
+if sys.version_info >= (3, 11):
+    from typing import Unpack
+else:
+    from typing_extensions import Unpack
+
+from ...tokenization_utils_base import BatchEncoding, PreTokenizedInput, TextInput
+
+
+class ChineseClipProcessorKwargs(ProcessingKwargs, total=False):
+    _defaults = {}
 
 
 class ChineseCLIPProcessor(ProcessorMixin):
@@ -60,7 +74,14 @@ def __init__(self, image_processor=None, tokenizer=None, **kwargs):
         super().__init__(image_processor, tokenizer)
         self.current_processor = self.image_processor
 
-    def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
+    def __call__(
+        self,
+        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
+        images: ImageInput = None,
+        audio=None,
+        videos=None,
+        **kwargs: Unpack[ChineseClipProcessorKwargs],
+    ) -> BatchEncoding:
         """
         Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text`
         and `kwargs` arguments to BertTokenizerFast's [`~BertTokenizerFast.__call__`] if `text` is not `None` to encode
@@ -79,12 +100,10 @@ def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
 
             return_tensors (`str` or [`~utils.TensorType`], *optional*):
                 If set, will return tensors of a particular framework. Acceptable values are:
-
-                - `'tf'`: Return TensorFlow `tf.constant` objects.
-                - `'pt'`: Return PyTorch `torch.Tensor` objects.
-                - `'np'`: Return NumPy `np.ndarray` objects.
-                - `'jax'`: Return JAX `jnp.ndarray` objects.
-
+                    - `'tf'`: Return TensorFlow `tf.constant` objects.
+                    - `'pt'`: Return PyTorch `torch.Tensor` objects.
+                    - `'np'`: Return NumPy `np.ndarray` objects.
+                    - `'jax'`: Return JAX `jnp.ndarray` objects.
         Returns:
             [`BatchEncoding`]: A [`BatchEncoding`] with the following fields:
 
@@ -97,12 +116,20 @@ def __call__(self, text=None, images=None, return_tensors=None, **kwargs):
 
         if text is None and images is None:
             raise ValueError("You have to specify either text or images. Both cannot be none.")
+        output_kwargs = self._merge_kwargs(
+            ChineseClipProcessorKwargs,
+            tokenizer_init_kwargs=self.tokenizer.init_kwargs,
+            **kwargs,
+        )
 
         if text is not None:
-            encoding = self.tokenizer(text, return_tensors=return_tensors, **kwargs)
-
+            encoding = self.tokenizer(text, **output_kwargs["text_kwargs"])
         if images is not None:
-            image_features = self.image_processor(images, return_tensors=return_tensors, **kwargs)
+            image_features = self.image_processor(images, **output_kwargs["images_kwargs"])
+
+        # BC for explicit return_tensors
+        if "return_tensors" in output_kwargs["common_kwargs"]:
+            return_tensors = output_kwargs["common_kwargs"].pop("return_tensors", None)
 
         if text is not None and images is not None:
             encoding["pixel_values"] = image_features.pixel_values

src/transformers/processing_utils.py

@@ -150,6 +150,8 @@ class methods and docstrings.
             Standard deviation to use if normalizing the image.
         do_pad (`bool`, *optional*):
             Whether to pad the image to the `(max_height, max_width)` of the images in the batch.
+        pad_size (`Dict[str, int]`, *optional*):
+            The size `{"height": int, "width" int}` to pad the images to.
         do_center_crop (`bool`, *optional*):
             Whether to center crop the image.
         data_format (`ChannelDimension` or `str`, *optional*):
@@ -169,6 +171,7 @@ class methods and docstrings.
     image_mean: Optional[Union[float, List[float]]]
     image_std: Optional[Union[float, List[float]]]
     do_pad: Optional[bool]
+    pad_size: Optional[Dict[str, int]]
     do_center_crop: Optional[bool]
     data_format: Optional[ChannelDimension]
     input_data_format: Optional[Union[str, ChannelDimension]]
@@ -753,7 +756,8 @@ class MyProcessingKwargs(ProcessingKwargs, CommonKwargs, TextKwargs, ImagesKwarg
                     # check if this key was passed as a flat kwarg.
                     if kwarg_value != "__empty__" and modality_key in non_modality_kwargs:
                         raise ValueError(
-                            f"Keyword argument {modality_key} was passed two times: in a dictionary for {modality} and as a **kwarg."
+                            f"Keyword argument {modality_key} was passed two times:\n"
+                            f"in a dictionary for {modality} and as a **kwarg."
                         )
                 elif modality_key in kwargs:
                     kwarg_value = kwargs.pop(modality_key, "__empty__")

tests/models/altclip/test_processor_altclip.py

@@ -0,0 +1,82 @@
+# coding=utf-8
+# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+
+import unittest
+
+import numpy as np
+
+from transformers import XLMRobertaTokenizer, XLMRobertaTokenizerFast
+from transformers.testing_utils import require_torch, require_vision
+from transformers.utils import is_vision_available
+
+from ...test_processing_common import ProcessorTesterMixin
+
+
+if is_vision_available():
+    from PIL import Image
+
+    from transformers import AltCLIPProcessor, CLIPImageProcessor
+
+
+@require_vision
+class AltClipProcessorTest(ProcessorTesterMixin, unittest.TestCase):
+    processor_class = AltCLIPProcessor
+
+    def setUp(self):
+        self.model_id = "BAAI/AltCLIP"
+
+    def get_tokenizer(self, **kwargs):
+        return XLMRobertaTokenizer.from_pretrained(self.model_id, **kwargs)
+
+    def get_rust_tokenizer(self, **kwargs):
+        return XLMRobertaTokenizerFast.from_pretrained(self.model_id, **kwargs)
+
+    def get_image_processor(self, **kwargs):
+        return CLIPImageProcessor.from_pretrained(self.model_id, **kwargs)
+
+    def prepare_image_inputs(self):
+        """This function prepares a list of PIL images, or a list of numpy arrays if one specifies numpify=True,
+        or a list of PyTorch tensors if one specifies torchify=True.
+        """
+
+        image_inputs = [np.random.randint(255, size=(3, 30, 400), dtype=np.uint8)]
+        image_inputs = [Image.fromarray(np.moveaxis(x, 0, -1)) for x in image_inputs]
+        return image_inputs
+
+    @require_torch
+    @require_vision
+    def test_unstructured_kwargs_batched(self):
+        if "image_processor" not in self.processor_class.attributes:
+            self.skipTest(f"image_processor attribute not present in {self.processor_class}")
+        image_processor = self.get_component("image_processor")
+        tokenizer = self.get_component("tokenizer")
+
+        processor = self.processor_class(tokenizer=tokenizer, image_processor=image_processor)
+        self.skip_processor_without_typed_kwargs(processor)
+
+        input_str = ["lower newer", "upper older longer string"]
+        image_input = self.prepare_image_inputs() * 2
+        inputs = processor(
+            text=input_str,
+            images=image_input,
+            return_tensors="pt",
+            crop_size={"height": 214, "width": 214},
+            padding="longest",
+            max_length=76,
+        )
+        self.assertEqual(inputs["pixel_values"].shape[2], 214)
+
+        self.assertEqual(len(inputs["input_ids"][0]), 7)

tests/models/chinese_clip/test_processor_chinese_clip.py

@@ -26,6 +26,8 @@
 from transformers.testing_utils import require_vision
 from transformers.utils import FEATURE_EXTRACTOR_NAME, is_vision_available
 
+from ...test_processing_common import ProcessorTesterMixin
+
 
 if is_vision_available():
     from PIL import Image
@@ -34,7 +36,9 @@
 
 
 @require_vision
-class ChineseCLIPProcessorTest(unittest.TestCase):
+class ChineseCLIPProcessorTest(ProcessorTesterMixin, unittest.TestCase):
+    processor_class = ChineseCLIPProcessor
+
     def setUp(self):
         self.tmpdirname = tempfile.mkdtemp()