paddlex 3.0.0rc0__py3-none-any.whl → 3.0.0rc1__py3-none-any.whl

paddlex/inference/models/doc_vlm/result.py

@@ -0,0 +1,21 @@
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
+
+from ...common.result import BaseResult
+
+
+class DocVLMResult(BaseResult):
+
+    def __init__(self, data: dict) -> None:
+        super().__init__(data)

paddlex/inference/models/face_feature/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.

paddlex/inference/models/face_feature/predictor.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -15,6 +15,7 @@
 from typing import Any, Dict, List, Union
 
 import numpy as np
+
 from ....modules.face_recognition.model_list import MODELS
 from ..image_feature import ImageFeaturePredictor
 

paddlex/inference/models/formula_recognition/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.

paddlex/inference/models/formula_recognition/predictor.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -13,33 +13,30 @@
 # limitations under the License.
 
 import numpy as np
+
+from ....modules.formula_recognition.model_list import MODELS
 from ....utils import logging
 from ....utils.func_register import FuncRegister
-from ....modules.formula_recognition.model_list import MODELS
 from ...common.batch_sampler import ImageBatchSampler
 from ...common.reader import ReadImage
-from ..common import (
-    StaticInfer,
-)
-from ..base import BasicPredictor
+from ..base import BasePredictor
 from .processors import (
-    MinMaxResize,
-    LatexTestTransform,
     LatexImageFormat,
     LaTeXOCRDecode,
+    LatexTestTransform,
+    MinMaxResize,
     NormalizeImage,
     ToBatch,
-    UniMERNetImgDecode,
     UniMERNetDecode,
-    UniMERNetTestTransform,
     UniMERNetImageFormat,
+    UniMERNetImgDecode,
+    UniMERNetTestTransform,
 )
-
 from .result import FormulaRecResult
 
 
-class FormulaRecPredictor(BasicPredictor):
-    """FormulaRecPredictor that inherits from BasicPredictor."""
+class FormulaRecPredictor(BasePredictor):
+    """FormulaRecPredictor that inherits from BasePredictor."""
 
     entities = MODELS
 
@@ -84,20 +81,7 @@ class FormulaRecPredictor(BasicPredictor):
                 pre_tfs[name] = op
         pre_tfs["ToBatch"] = ToBatch()
 
-        if self.model_name in ("LaTeX_OCR_rec") and self.pp_option.device in ("cpu"):
-            import cpuinfo
-
-            if "GenuineIntel" in cpuinfo.get_cpu_info().get("vendor_id_raw", ""):
-                self.pp_option.run_mode = "mkldnn"
-                logging.warning(
-                    "Now, the `LaTeX_OCR_rec` model only support `mkldnn` mode when running on Intel CPU devices. So using `mkldnn` instead."
-                )
-
-        infer = StaticInfer(
-            model_dir=self.model_dir,
-            model_prefix=self.MODEL_FILE_PREFIX,
-            option=self.pp_option,
-        )
+        infer = self.create_static_infer()
 
         post_op = self.build_postprocess(**self.config["PostProcess"])
         return pre_tfs, infer, post_op

paddlex/inference/models/formula_recognition/processors.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -13,23 +13,28 @@
 # limitations under the License.
 
 
+import json
+import math
 import os
-import os.path as osp
-
 import re
-import numpy as np
-from PIL import Image, ImageOps, ImageDraw
-import cv2
-import math
-import json
 import tempfile
-from tokenizers import Tokenizer as TokenizerFast
-from tokenizers import AddedToken
-from typing import List, Tuple, Optional, Any, Dict, Union
+from typing import Any, Dict, List, Optional, Tuple, Union
+
+import numpy as np
+from PIL import Image, ImageOps
 
-from ....utils import logging
+from ....utils.deps import class_requires_deps, is_dep_available
+from ...utils.benchmark import benchmark
 
+if is_dep_available("opencv-contrib-python"):
+    import cv2
+if is_dep_available("tokenizers"):
+    from tokenizers import AddedToken
+    from tokenizers import Tokenizer as TokenizerFast
 
+
+@benchmark.timeit
+@class_requires_deps("opencv-contrib-python")
 class MinMaxResize:
     """Class for resizing images to be within specified minimum and maximum dimensions, with padding and normalization."""
 
@@ -154,6 +159,8 @@ class MinMaxResize:
         return [self.resize(img) for img in imgs]
 
 
+@benchmark.timeit
+@class_requires_deps("opencv-contrib-python")
 class LatexTestTransform:
     """
     A transform class for processing images according to Latex test requirements.
@@ -194,6 +201,7 @@ class LatexTestTransform:
         return [self.transform(img) for img in imgs]
 
 
+@benchmark.timeit
 class LatexImageFormat:
     """Class for formatting images to a specific format suitable for LaTeX."""
 
@@ -232,6 +240,7 @@ class LatexImageFormat:
         return [self.format(img) for img in imgs]
 
 
+@benchmark.timeit
 class NormalizeImage(object):
     """Normalize an image by subtracting the mean and dividing by the standard deviation.
 
@@ -280,6 +289,7 @@ class NormalizeImage(object):
         return [self.normalize(img) for img in imgs]
 
 
+@benchmark.timeit
 class ToBatch(object):
     """A class for batching images."""
 
@@ -302,6 +312,8 @@ class ToBatch(object):
         return x
 
 
+@benchmark.timeit
+@class_requires_deps("tokenizers")
 class LaTeXOCRDecode(object):
     """Class for decoding LaTeX OCR tokens based on a provided character list."""
 
@@ -312,8 +324,6 @@ class LaTeXOCRDecode(object):
             character_list (list): The list of characters to use for tokenization.
             **kwargs: Additional keyword arguments for initialization.
         """
-        from tokenizers import Tokenizer as TokenizerFast
-
         super(LaTeXOCRDecode, self).__init__()
         temp_path = tempfile.gettempdir()
         rec_char_dict_path = os.path.join(temp_path, "latexocr_tokenizer.json")
@@ -402,6 +412,8 @@ class LaTeXOCRDecode(object):
         return text, label
 
 
+@benchmark.timeit
+@class_requires_deps("opencv-contrib-python")
 class UniMERNetImgDecode(object):
     """Class for decoding images for UniMERNet, including cropping margins, resizing, and padding."""
 
@@ -554,6 +566,8 @@ class UniMERNetImgDecode(object):
         return [self.img_decode(img) for img in imgs]
 
 
+@benchmark.timeit
+@class_requires_deps("tokenizers")
 class UniMERNetDecode(object):
     """Class for decoding tokenized inputs using UniMERNet tokenizer.
 
@@ -687,8 +701,8 @@ class UniMERNetDecode(object):
                         self._add_tokens(tokens, special_tokens=is_last_special)
 
     def _add_tokens(
-        self, new_tokens: List[Union[AddedToken, str]], special_tokens: bool = False
-    ) -> List[Union[AddedToken, str]]:
+        self, new_tokens: "List[Union[AddedToken, str]]", special_tokens: bool = False
+    ) -> "List[Union[AddedToken, str]]":
         """Adds new tokens to the tokenizer.
 
         Args:
@@ -704,7 +718,7 @@ class UniMERNetDecode(object):
         return self.tokenizer.add_tokens(new_tokens)
 
     def added_tokens_encoder(
-        self, added_tokens_decoder: Dict[int, AddedToken]
+        self, added_tokens_decoder: "Dict[int, AddedToken]"
     ) -> Dict[str, int]:
         """Creates an encoder dictionary from added tokens.
 
@@ -730,7 +744,7 @@ class UniMERNetDecode(object):
         return all_toks
 
     @property
-    def all_special_tokens_extended(self) -> List[Union[str, AddedToken]]:
+    def all_special_tokens_extended(self) -> "List[Union[str, AddedToken]]":
         """Retrieves all special tokens, including extended ones.
 
         Returns:
@@ -900,6 +914,8 @@ class UniMERNetDecode(object):
         return text, label
 
 
+@benchmark.timeit
+@class_requires_deps("opencv-contrib-python")
 class UniMERNetTestTransform:
     """
     A class for transforming images according to UniMERNet test specifications.
@@ -947,13 +963,13 @@ class UniMERNetTestTransform:
         return [self.transform(img) for img in imgs]
 
 
+@benchmark.timeit
 class UniMERNetImageFormat:
     """Class for formatting images to UniMERNet's required format."""
 
     def __init__(self, **kwargs) -> None:
         """Initializes the UniMERNetImageFormat instance."""
         # your init code
-        pass
 
     def format(self, img: np.ndarray) -> np.ndarray:
         """Formats a single image to UniMERNet's required format.

paddlex/inference/models/formula_recognition/result.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -12,24 +12,28 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import os, sys
-from typing import Any, Dict, Optional, List
-import cv2
-import PIL
-import fitz
 import copy
 import math
-import random
-import tempfile
+import os
 import subprocess
-import numpy as np
+import tempfile
 from pathlib import Path
+from typing import List, Optional
+
+import numpy as np
+import PIL
 from PIL import Image, ImageDraw, ImageFont
 
-from ...common.result import BaseCVResult, StrMixin, JsonMixin
 from ....utils import logging
-from ....utils.fonts import PINGFANG_FONT_FILE_PATH
+from ....utils.deps import function_requires_deps, is_dep_available
 from ....utils.file_interface import custom_open
+from ....utils.fonts import PINGFANG_FONT_FILE_PATH
+from ...common.result import BaseCVResult, JsonMixin
+
+if is_dep_available("opencv-contrib-python"):
+    import cv2
+if is_dep_available("PyMuPDF"):
+    import fitz
 
 
 class FormulaRecResult(BaseCVResult):
@@ -199,12 +203,13 @@ def generate_pdf_file(
         if is_debug:
             subprocess.check_call(command, shell=True)
         else:
-            devNull = custom_open(os.devnull, "w")
+            custom_open(os.devnull, "w")
             subprocess.check_call(
                 command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, shell=True
             )
 
 
+@function_requires_deps("opencv-contrib-python")
 def crop_white_area(image: np.ndarray) -> Optional[List[int]]:
     """
     Finds and returns the bounding box of the non-white area in an image.
@@ -231,6 +236,7 @@ def crop_white_area(image: np.ndarray) -> Optional[List[int]]:
         return None
 
 
+@function_requires_deps("PyMuPDF", "opencv-contrib-python")
 def pdf2img(pdf_path: str, img_path: str, is_padding: bool = False):
     """
     Converts a single-page PDF to an image, optionally cropping white areas and adding padding.
@@ -326,6 +332,7 @@ def env_valid() -> bool:
             formula_img = pdf2img(pdf_file_path, img_file_path, is_padding=False)
 
 
+@function_requires_deps("opencv-contrib-python")
 def draw_box_txt_fine(img_size: tuple, box: list, txt: str, font_path: str):
     """
     Draw box text.

paddlex/inference/models/image_classification/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.

paddlex/inference/models/image_classification/predictor.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -12,28 +12,22 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Union, Dict, List, Tuple
+from typing import Any, Dict, List, Tuple, Union
+
 import numpy as np
 
-from ....utils.func_register import FuncRegister
 from ....modules.image_classification.model_list import MODELS
+from ....utils.func_register import FuncRegister
 from ...common.batch_sampler import ImageBatchSampler
 from ...common.reader import ReadImage
-from ..common import (
-    Resize,
-    ResizeByShort,
-    Normalize,
-    ToCHWImage,
-    ToBatch,
-    StaticInfer,
-)
-from ..base import BasicPredictor
+from ..base import BasePredictor
+from ..common import Normalize, Resize, ResizeByShort, ToBatch, ToCHWImage
 from .processors import Crop, Topk
 from .result import TopkResult
 
 
-class ClasPredictor(BasicPredictor):
-    """ClasPredictor that inherits from BasicPredictor."""
+class ClasPredictor(BasePredictor):
+    """ClasPredictor that inherits from BasePredictor."""
 
     entities = MODELS
 
@@ -85,11 +79,7 @@ class ClasPredictor(BasicPredictor):
             preprocessors[name] = op
         preprocessors["ToBatch"] = ToBatch()
 
-        infer = StaticInfer(
-            model_dir=self.model_dir,
-            model_prefix=self.MODEL_FILE_PREFIX,
-            option=self.pp_option,
-        )
+        infer = self.create_static_infer()
 
         postprocessors = {}
         for key in self.config["PostProcess"]:

paddlex/inference/models/image_classification/processors.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -14,10 +14,11 @@
 
 import numpy as np
 
-from ....utils import logging
+from ...utils.benchmark import benchmark
 from ..common.vision import F
 
 
+@benchmark.timeit
 class Crop:
     """Crop region from the image."""
 
@@ -64,6 +65,7 @@ class Crop:
         return img
 
 
+@benchmark.timeit
 class Topk:
     """Topk Transform"""
 

paddlex/inference/models/image_classification/result.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -14,13 +14,14 @@
 
 
 import copy
+
+import numpy as np
 import PIL
 from PIL import Image, ImageDraw, ImageFont
-import numpy as np
 
 from ....utils.fonts import PINGFANG_FONT_FILE_PATH
+from ...common.result import BaseCVResult, JsonMixin
 from ...utils.color_map import get_colormap
-from ...common.result import BaseCVResult, StrMixin, JsonMixin
 
 
 class TopkResult(BaseCVResult):

paddlex/inference/models/image_feature/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.

paddlex/inference/models/image_feature/predictor.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -12,28 +12,22 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Union, Dict, List, Tuple
+from typing import Any, Dict, List, Tuple, Union
+
 import numpy as np
 
-from ....utils.func_register import FuncRegister
 from ....modules.general_recognition.model_list import MODELS
+from ....utils.func_register import FuncRegister
 from ...common.batch_sampler import ImageBatchSampler
 from ...common.reader import ReadImage
-from ..common import (
-    Resize,
-    ResizeByShort,
-    Normalize,
-    ToCHWImage,
-    ToBatch,
-    StaticInfer,
-)
-from ..base import BasicPredictor
+from ..base import BasePredictor
+from ..common import Normalize, Resize, ResizeByShort, ToBatch, ToCHWImage
 from .processors import NormalizeFeatures
 from .result import IdentityResult
 
 
-class ImageFeaturePredictor(BasicPredictor):
-    """ImageFeaturePredictor that inherits from BasicPredictor."""
+class ImageFeaturePredictor(BasePredictor):
+    """ImageFeaturePredictor that inherits from BasePredictor."""
 
     entities = MODELS
 
@@ -83,11 +77,7 @@ class ImageFeaturePredictor(BasicPredictor):
             preprocessors[name] = op
         preprocessors["ToBatch"] = ToBatch()
 
-        infer = StaticInfer(
-            model_dir=self.model_dir,
-            model_prefix=self.MODEL_FILE_PREFIX,
-            option=self.pp_option,
-        )
+        infer = self.create_static_infer()
 
         postprocessors = {}
         for key in self.config["PostProcess"]:

paddlex/inference/models/image_feature/processors.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -14,7 +14,10 @@
 
 import numpy as np
 
+from ...utils.benchmark import benchmark
 
+
+@benchmark.timeit
 class NormalizeFeatures:
     """Normalize Features Transform"""
 

paddlex/inference/models/image_feature/result.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -13,9 +13,8 @@
 # limitations under the License.
 
 import copy
-from PIL import Image
 
-from ...common.result import BaseResult, StrMixin, JsonMixin
+from ...common.result import BaseResult, JsonMixin
 
 
 class IdentityResult(BaseResult):

paddlex/inference/models/image_multilabel_classification/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.

paddlex/inference/models/image_multilabel_classification/predictor.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -12,17 +12,18 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Union, Dict, List
+from typing import Any, Dict, List, Union
+
 import numpy as np
 
-from .result import MLClassResult
-from .processors import MultiLabelThreshOutput
-from ..image_classification import ClasPredictor
 from ....modules.multilabel_classification.model_list import MODELS
+from ..image_classification import ClasPredictor
+from .processors import MultiLabelThreshOutput
+from .result import MLClassResult
 
 
 class MLClasPredictor(ClasPredictor):
-    """MLClasPredictor that inherits from BasicPredictor."""
+    """MLClasPredictor that inherits from BasePredictor."""
 
     entities = MODELS
 

paddlex/inference/models/image_multilabel_classification/processors.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -12,10 +12,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import numpy as np
 from typing import Union
 
+import numpy as np
+
+from ...utils.benchmark import benchmark
+
 
+@benchmark.timeit
 class MultiLabelThreshOutput:
     """MultiLabelThresh Transform"""
 

paddlex/inference/models/image_multilabel_classification/result.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.
@@ -13,13 +13,14 @@
 # limitations under the License.
 
 import copy
+
+import numpy as np
 import PIL
 from PIL import Image, ImageDraw, ImageFont
-import numpy as np
 
 from ....utils.fonts import PINGFANG_FONT_FILE_PATH
+from ...common.result import BaseCVResult, JsonMixin
 from ...utils.color_map import get_colormap
-from ...common.result import BaseCVResult, StrMixin, JsonMixin
 
 
 class MLClassResult(BaseCVResult):

paddlex/inference/models/image_unwarping/__init__.py

@@ -1,4 +1,4 @@
-# copyright (c) 2024 PaddlePaddle Authors. All Rights Reserve.
+# Copyright (c) 2024 PaddlePaddle Authors. 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.