paddlex 3.0.0rc0__py3-none-any.whl → 3.0.1__py3-none-any.whl

paddlex/inference/pipelines/open_vocabulary_segmentation/pipeline.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.
@@ -12,16 +12,20 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional, Union, Tuple, List
+from typing import Any, Dict, List, Optional, Union
+
 import numpy as np
-from ...utils.pp_option import PaddlePredictorOption
-from ..base import BasePipeline
 
+from ....utils.deps import pipeline_requires_extra
 from ...models.open_vocabulary_segmentation.results import SAMSegResult
+from ...utils.hpi import HPIConfig
+from ...utils.pp_option import PaddlePredictorOption
+from ..base import BasePipeline
 
 Number = Union[int, float]
 
 
+@pipeline_requires_extra("multimodal")
 class OpenVocabularySegmentationPipeline(BasePipeline):
     """Open Vocabulary Segmentation pipeline"""
 
@@ -33,6 +37,7 @@ class OpenVocabularySegmentationPipeline(BasePipeline):
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
     ) -> None:
         """
         Initializes the class with given configurations and options.
@@ -41,9 +46,15 @@ class OpenVocabularySegmentationPipeline(BasePipeline):
             config (Dict): Configuration dictionary containing model and other parameters.
             device (str): The device to run the prediction on. Default is None.
             pp_option (PaddlePredictorOption): Options for PaddlePaddle predictor. Default is None.
-            use_hpip (bool): Whether to use high-performance inference (hpip) for prediction. Defaults to False.
+            use_hpip (bool, optional): Whether to use the high-performance
+                inference plugin (HPIP) by default. Defaults to False.
+            hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional):
+                The default high-performance inference configuration dictionary.
+                Defaults to None.
         """
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
         # create box-prompted SAM-H
         box_prompted_model_cfg = config.get("SubModules", {}).get(

paddlex/inference/pipelines/pp_chatocr/__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/pipelines/pp_chatocr/pipeline_base.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 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional
-from ..base import BasePipeline
+from typing import Any, Dict, Optional, Union
+
 from ....utils import logging
+from ...utils.hpi import HPIConfig
 from ...utils.pp_option import PaddlePredictorOption
+from ..base import BasePipeline
 
 
 class PP_ChatOCR_Pipeline(BasePipeline):
@@ -26,6 +28,7 @@ class PP_ChatOCR_Pipeline(BasePipeline):
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
     ) -> None:
         """Initializes the pp-chatocrv3-doc pipeline.
 
@@ -33,10 +36,16 @@ class PP_ChatOCR_Pipeline(BasePipeline):
             config (Dict): Configuration dictionary containing various settings.
             device (str, optional): Device to run the predictions on. Defaults to None.
             pp_option (PaddlePredictorOption, optional): PaddlePredictor options. Defaults to None.
-            use_hpip (bool, optional): Whether to use high-performance inference (hpip) for prediction. Defaults to False.
+            use_hpip (bool, optional): Whether to use the high-performance
+                inference plugin (HPIP) by default. Defaults to False.
+            hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional):
+                The default high-performance inference configuration dictionary.
+                Defaults to None.
         """
 
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
     def visual_predict(self):
         """

paddlex/inference/pipelines/pp_chatocr/pipeline_v3.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.
@@ -12,22 +12,27 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional, Union, List, Tuple
-import os
-import re
 import copy
 import json
+import os
+import re
+from typing import Any, Dict, List, Optional, Tuple, Union
+
 import numpy as np
-from .pipeline_base import PP_ChatOCR_Pipeline
-from ...common.reader import ReadImage
-from ...common.batch_sampler import ImageBatchSampler
+
 from ....utils import logging
+from ....utils.deps import pipeline_requires_extra
 from ....utils.file_interface import custom_open
+from ...common.batch_sampler import ImageBatchSampler
+from ...common.reader import ReadImage
+from ...utils.hpi import HPIConfig
 from ...utils.pp_option import PaddlePredictorOption
-from ..layout_parsing.result import LayoutParsingResult
 from ..components.chat_server import BaseChat
+from ..layout_parsing.result import LayoutParsingResult
+from .pipeline_base import PP_ChatOCR_Pipeline
 
 
+@pipeline_requires_extra("ie")
 class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
     """PP-ChatOCR Pipeline"""
 
@@ -39,6 +44,7 @@ class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
         initial_predictor: bool = True,
     ) -> None:
         """Initializes the pp-chatocrv3-doc pipeline.
@@ -47,12 +53,17 @@ class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
             config (Dict): Configuration dictionary containing various settings.
             device (str, optional): Device to run the predictions on. Defaults to None.
             pp_option (PaddlePredictorOption, optional): PaddlePredictor options. Defaults to None.
-            use_hpip (bool, optional): Whether to use high-performance inference (hpip) for prediction. Defaults to False.
-            use_layout_parsing (bool, optional): Whether to use layout parsing. Defaults to True.
+            use_hpip (bool, optional): Whether to use the high-performance
+                inference plugin (HPIP) by default. Defaults to False.
+            hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional):
+                The default high-performance inference configuration dictionary.
+                Defaults to None.
             initial_predictor (bool, optional): Whether to initialize the predictor. Defaults to True.
         """
 
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
         self.pipeline_name = config["pipeline_name"]
         self.config = config
@@ -195,7 +206,6 @@ class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
         input: Union[str, List[str], np.ndarray, List[np.ndarray]],
         use_doc_orientation_classify: Optional[bool] = None,
         use_doc_unwarping: Optional[bool] = None,
-        use_general_ocr: Optional[bool] = None,
         use_seal_recognition: Optional[bool] = None,
         use_table_recognition: Optional[bool] = None,
         layout_threshold: Optional[Union[float, dict]] = None,
@@ -226,7 +236,6 @@ class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
                                                                         numpy array of an image, or list of numpy arrays.
             use_doc_orientation_classify (bool): Flag to use document orientation classification.
             use_doc_unwarping (bool): Flag to use document unwarping.
-            use_general_ocr (bool): Flag to use general OCR.
             use_seal_recognition (bool): Flag to use seal recognition.
             use_table_recognition (bool): Flag to use table recognition.
             layout_threshold (Optional[float]): The threshold value to filter out low-confidence predictions. Default is None.
@@ -269,7 +278,6 @@ class PP_ChatOCRv3_Pipeline(PP_ChatOCR_Pipeline):
             input,
             use_doc_orientation_classify=use_doc_orientation_classify,
             use_doc_unwarping=use_doc_unwarping,
-            use_general_ocr=use_general_ocr,
             use_seal_recognition=use_seal_recognition,
             use_table_recognition=use_table_recognition,
             layout_threshold=layout_threshold,

paddlex/inference/pipelines/pp_chatocr/pipeline_v4.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 +12,35 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional, Union, List, Tuple
-import os
-import re
-import cv2
+import base64
 import copy
 import json
-import base64
+import os
+import re
+from typing import Any, Dict, List, Optional, Tuple, Union
+
 import numpy as np
-from .pipeline_base import PP_ChatOCR_Pipeline
-from ...common.reader import ReadImage
-from ...common.batch_sampler import ImageBatchSampler
+
 from ....utils import logging
+from ....utils.deps import (
+    function_requires_deps,
+    is_dep_available,
+    pipeline_requires_extra,
+)
 from ....utils.file_interface import custom_open
+from ...common.batch_sampler import ImageBatchSampler
+from ...common.reader import ReadImage
+from ...utils.hpi import HPIConfig
 from ...utils.pp_option import PaddlePredictorOption
-from ..layout_parsing.result import LayoutParsingResult
 from ..components.chat_server import BaseChat
+from ..layout_parsing.result import LayoutParsingResult
+from .pipeline_base import PP_ChatOCR_Pipeline
+
+if is_dep_available("opencv-contrib-python"):
+    import cv2
 
 
+@pipeline_requires_extra("ie")
 class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
     """PP-ChatOCRv4 Pipeline"""
 
@@ -41,6 +52,7 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
         initial_predictor: bool = True,
     ) -> None:
         """Initializes the pp-chatocrv3-doc pipeline.
@@ -49,12 +61,17 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
             config (Dict): Configuration dictionary containing various settings.
             device (str, optional): Device to run the predictions on. Defaults to None.
             pp_option (PaddlePredictorOption, optional): PaddlePredictor options. Defaults to None.
-            use_hpip (bool, optional): Whether to use high-performance inference (hpip) for prediction. Defaults to False.
-            use_layout_parsing (bool, optional): Whether to use layout parsing. Defaults to True.
+            use_hpip (bool, optional): Whether to use the high-performance
+                inference plugin (HPIP) by default. Defaults to False.
+            hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional):
+                The default high-performance inference configuration dictionary.
+                Defaults to None.
             initial_predictor (bool, optional): Whether to initialize the predictor. Defaults to True.
         """
 
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
         self.pipeline_name = config["pipeline_name"]
         self.config = config
@@ -232,7 +249,7 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
         input: Union[str, List[str], np.ndarray, List[np.ndarray]],
         use_doc_orientation_classify: Optional[bool] = None,
         use_doc_unwarping: Optional[bool] = None,
-        use_general_ocr: Optional[bool] = None,
+        use_textline_orientation: Optional[bool] = None,
         use_seal_recognition: Optional[bool] = None,
         use_table_recognition: Optional[bool] = None,
         layout_threshold: Optional[Union[float, dict]] = None,
@@ -263,7 +280,7 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
                                                                         numpy array of an image, or list of numpy arrays.
             use_doc_orientation_classify (bool): Flag to use document orientation classification.
             use_doc_unwarping (bool): Flag to use document unwarping.
-            use_general_ocr (bool): Flag to use general OCR.
+            use_textline_orientation (Optional[bool]): Whether to use textline orientation prediction.
             use_seal_recognition (bool): Flag to use seal recognition.
             use_table_recognition (bool): Flag to use table recognition.
             layout_threshold (Optional[float]): The threshold value to filter out low-confidence predictions. Default is None.
@@ -305,7 +322,7 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
             input,
             use_doc_orientation_classify=use_doc_orientation_classify,
             use_doc_unwarping=use_doc_unwarping,
-            use_general_ocr=use_general_ocr,
+            use_textline_orientation=use_textline_orientation,
             use_seal_recognition=use_seal_recognition,
             use_table_recognition=use_table_recognition,
             layout_threshold=layout_threshold,
@@ -574,6 +591,7 @@ class PP_ChatOCRv4_Pipeline(PP_ChatOCR_Pipeline):
 
         return []
 
+    @function_requires_deps("opencv-contrib-python")
     def mllm_pred(
         self,
         input: Union[str, np.ndarray],

paddlex/inference/pipelines/pp_shitu_v2/__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/pipelines/pp_shitu_v2/pipeline.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.
@@ -12,18 +12,19 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional
+from typing import Any, Dict, Optional, Union
 
-import numpy as np
-
-from ...utils.pp_option import PaddlePredictorOption
-from ...common.reader import ReadImage
+from ....utils.deps import pipeline_requires_extra
 from ...common.batch_sampler import ImageBatchSampler
-from ..components import CropByBoxes, FaissIndexer, FaissBuilder, IndexData
+from ...common.reader import ReadImage
+from ...utils.hpi import HPIConfig
+from ...utils.pp_option import PaddlePredictorOption
 from ..base import BasePipeline
+from ..components import CropByBoxes, FaissBuilder, FaissIndexer
 from .result import ShiTuResult
 
 
+@pipeline_requires_extra("cv")
 class ShiTuV2Pipeline(BasePipeline):
     """ShiTuV2 Pipeline"""
 
@@ -35,8 +36,11 @@ class ShiTuV2Pipeline(BasePipeline):
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
     ):
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
         self._topk, self._rec_threshold, self._hamming_radius, self._det_threshold = (
             config.get("rec_topk", 5),

paddlex/inference/pipelines/pp_shitu_v2/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.
@@ -17,9 +17,9 @@ import copy
 import PIL
 from PIL import Image, ImageDraw, ImageFont
 
-from ....utils.fonts import PINGFANG_FONT_FILE_PATH, create_font
-from ...common.result import BaseCVResult, StrMixin, JsonMixin
-from ...utils.color_map import get_colormap, font_colormap
+from ....utils.fonts import PINGFANG_FONT_FILE_PATH
+from ...common.result import BaseCVResult, JsonMixin
+from ...utils.color_map import font_colormap, get_colormap
 
 
 def draw_box(img, boxes):

paddlex/inference/pipelines/rotated_object_detection/__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/pipelines/rotated_object_detection/pipeline.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.
@@ -12,25 +12,28 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Optional, Union, List
+from typing import Any, Dict, List, Optional, Union
+
 import numpy as np
-from ...utils.pp_option import PaddlePredictorOption
-from ..base import BasePipeline
 
+from ....utils.deps import pipeline_requires_extra
 from ...models.object_detection.result import DetResult
+from ...utils.hpi import HPIConfig
+from ...utils.pp_option import PaddlePredictorOption
+from .._parallel import AutoParallelImageSimpleInferencePipeline
+from ..base import BasePipeline
 
 
-class RotatedObjectDetectionPipeline(BasePipeline):
+class _RotatedObjectDetectionPipeline(BasePipeline):
     """Rotated Object Detection Pipeline"""
 
-    entities = "rotated_object_detection"
-
     def __init__(
         self,
         config: Dict,
         device: str = None,
         pp_option: PaddlePredictorOption = None,
         use_hpip: bool = False,
+        hpi_config: Optional[Union[Dict[str, Any], HPIConfig]] = None,
     ) -> None:
         """
         Initializes the class with given configurations and options.
@@ -39,9 +42,15 @@ class RotatedObjectDetectionPipeline(BasePipeline):
             config (Dict): Configuration dictionary containing model and other parameters.
             device (str): The device to run the prediction on. Default is None.
             pp_option (PaddlePredictorOption): Options for PaddlePaddle predictor. Default is None.
-            use_hpip (bool): Whether to use high-performance inference (hpip) for prediction. Defaults to False.
+            use_hpip (bool, optional): Whether to use the high-performance
+                inference plugin (HPIP) by default. Defaults to False.
+            hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional):
+                The default high-performance inference configuration dictionary.
+                Defaults to None.
         """
-        super().__init__(device=device, pp_option=pp_option, use_hpip=use_hpip)
+        super().__init__(
+            device=device, pp_option=pp_option, use_hpip=use_hpip, hpi_config=hpi_config
+        )
 
         rotated_object_detection_model_config = config["SubModules"][
             "RotatedObjectDetection"
@@ -72,3 +81,15 @@ class RotatedObjectDetectionPipeline(BasePipeline):
             DetResult: The predicted rotated object detection results.
         """
         yield from self.rotated_object_detection_model(input, threshold=threshold)
+
+
+@pipeline_requires_extra("cv")
+class RotatedObjectDetectionPipeline(AutoParallelImageSimpleInferencePipeline):
+    entities = "rotated_object_detection"
+
+    @property
+    def _pipeline_cls(self):
+        return _RotatedObjectDetectionPipeline
+
+    def _get_batch_size(self, config):
+        return config["SubModules"]["RotatedObjectDetection"].get("batch_size", 1)

paddlex/inference/pipelines/seal_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.