paddlex 3.0.0rc1__py3-none-any.whl → 3.0.1__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (233) hide show
  1. paddlex/.version +1 -1
  2. paddlex/__init__.py +1 -1
  3. paddlex/configs/modules/chart_parsing/PP-Chart2Table.yaml +13 -0
  4. paddlex/configs/modules/doc_vlm/PP-DocBee2-3B.yaml +14 -0
  5. paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-L.yaml +40 -0
  6. paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-M.yaml +40 -0
  7. paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-S.yaml +40 -0
  8. paddlex/configs/modules/layout_detection/PP-DocBlockLayout.yaml +40 -0
  9. paddlex/configs/modules/layout_detection/PP-DocLayout-L.yaml +2 -2
  10. paddlex/configs/modules/layout_detection/PP-DocLayout-M.yaml +2 -2
  11. paddlex/configs/modules/layout_detection/PP-DocLayout-S.yaml +2 -2
  12. paddlex/configs/modules/layout_detection/PP-DocLayout_plus-L.yaml +40 -0
  13. paddlex/configs/modules/text_detection/PP-OCRv5_mobile_det.yaml +40 -0
  14. paddlex/configs/modules/text_detection/PP-OCRv5_server_det.yaml +40 -0
  15. paddlex/configs/modules/text_recognition/PP-OCRv5_mobile_rec.yaml +39 -0
  16. paddlex/configs/modules/text_recognition/PP-OCRv5_server_rec.yaml +39 -0
  17. paddlex/configs/modules/textline_orientation/PP-LCNet_x1_0_textline_ori.yaml +41 -0
  18. paddlex/configs/pipelines/OCR.yaml +7 -6
  19. paddlex/configs/pipelines/PP-ChatOCRv3-doc.yaml +3 -1
  20. paddlex/configs/pipelines/PP-ChatOCRv4-doc.yaml +91 -34
  21. paddlex/configs/pipelines/PP-StructureV3.yaml +72 -72
  22. paddlex/configs/pipelines/doc_understanding.yaml +1 -1
  23. paddlex/configs/pipelines/formula_recognition.yaml +2 -2
  24. paddlex/configs/pipelines/layout_parsing.yaml +3 -2
  25. paddlex/configs/pipelines/seal_recognition.yaml +1 -0
  26. paddlex/configs/pipelines/table_recognition.yaml +2 -1
  27. paddlex/configs/pipelines/table_recognition_v2.yaml +7 -1
  28. paddlex/hpip_links.html +20 -20
  29. paddlex/inference/common/batch_sampler/doc_vlm_batch_sampler.py +33 -10
  30. paddlex/inference/common/batch_sampler/image_batch_sampler.py +34 -25
  31. paddlex/inference/common/result/mixin.py +19 -12
  32. paddlex/inference/models/base/predictor/base_predictor.py +2 -8
  33. paddlex/inference/models/common/static_infer.py +11 -59
  34. paddlex/inference/models/common/tokenizer/__init__.py +2 -0
  35. paddlex/inference/models/common/tokenizer/clip_tokenizer.py +1 -1
  36. paddlex/inference/models/common/tokenizer/gpt_tokenizer.py +2 -2
  37. paddlex/inference/models/common/tokenizer/qwen2_5_tokenizer.py +112 -0
  38. paddlex/inference/models/common/tokenizer/qwen2_tokenizer.py +7 -1
  39. paddlex/inference/models/common/tokenizer/qwen_tokenizer.py +288 -0
  40. paddlex/inference/models/common/tokenizer/tokenizer_utils.py +13 -13
  41. paddlex/inference/models/common/tokenizer/tokenizer_utils_base.py +3 -3
  42. paddlex/inference/models/common/tokenizer/vocab.py +7 -7
  43. paddlex/inference/models/common/vlm/conversion_utils.py +99 -0
  44. paddlex/inference/models/common/vlm/fusion_ops.py +205 -0
  45. paddlex/inference/models/common/vlm/generation/configuration_utils.py +1 -1
  46. paddlex/inference/models/common/vlm/generation/logits_process.py +1 -1
  47. paddlex/inference/models/common/vlm/generation/utils.py +1 -1
  48. paddlex/inference/models/common/vlm/transformers/configuration_utils.py +3 -3
  49. paddlex/inference/models/common/vlm/transformers/conversion_utils.py +3 -3
  50. paddlex/inference/models/common/vlm/transformers/model_outputs.py +2 -2
  51. paddlex/inference/models/common/vlm/transformers/model_utils.py +7 -31
  52. paddlex/inference/models/doc_vlm/modeling/GOT_ocr_2_0.py +830 -0
  53. paddlex/inference/models/doc_vlm/modeling/__init__.py +2 -0
  54. paddlex/inference/models/doc_vlm/modeling/qwen2.py +1606 -0
  55. paddlex/inference/models/doc_vlm/modeling/qwen2_5_vl.py +3006 -0
  56. paddlex/inference/models/doc_vlm/modeling/qwen2_vl.py +0 -105
  57. paddlex/inference/models/doc_vlm/predictor.py +79 -24
  58. paddlex/inference/models/doc_vlm/processors/GOT_ocr_2_0.py +97 -0
  59. paddlex/inference/models/doc_vlm/processors/__init__.py +2 -0
  60. paddlex/inference/models/doc_vlm/processors/common.py +189 -0
  61. paddlex/inference/models/doc_vlm/processors/qwen2_5_vl.py +548 -0
  62. paddlex/inference/models/doc_vlm/processors/qwen2_vl.py +21 -176
  63. paddlex/inference/models/formula_recognition/predictor.py +7 -1
  64. paddlex/inference/models/formula_recognition/processors.py +92 -79
  65. paddlex/inference/models/formula_recognition/result.py +28 -27
  66. paddlex/inference/models/image_feature/processors.py +3 -4
  67. paddlex/inference/models/keypoint_detection/predictor.py +3 -0
  68. paddlex/inference/models/object_detection/predictor.py +2 -0
  69. paddlex/inference/models/object_detection/processors.py +28 -3
  70. paddlex/inference/models/object_detection/utils.py +2 -0
  71. paddlex/inference/models/table_structure_recognition/result.py +0 -10
  72. paddlex/inference/models/text_detection/predictor.py +8 -0
  73. paddlex/inference/models/text_detection/processors.py +44 -10
  74. paddlex/inference/models/text_detection/result.py +0 -10
  75. paddlex/inference/pipelines/__init__.py +9 -5
  76. paddlex/inference/pipelines/_parallel.py +172 -0
  77. paddlex/inference/pipelines/anomaly_detection/pipeline.py +16 -6
  78. paddlex/inference/pipelines/attribute_recognition/pipeline.py +11 -1
  79. paddlex/inference/pipelines/base.py +14 -4
  80. paddlex/inference/pipelines/components/faisser.py +1 -1
  81. paddlex/inference/pipelines/doc_preprocessor/pipeline.py +53 -27
  82. paddlex/inference/pipelines/formula_recognition/pipeline.py +120 -82
  83. paddlex/inference/pipelines/formula_recognition/result.py +1 -11
  84. paddlex/inference/pipelines/image_classification/pipeline.py +16 -6
  85. paddlex/inference/pipelines/image_multilabel_classification/pipeline.py +16 -6
  86. paddlex/inference/pipelines/instance_segmentation/pipeline.py +16 -6
  87. paddlex/inference/pipelines/keypoint_detection/pipeline.py +16 -6
  88. paddlex/inference/pipelines/layout_parsing/pipeline.py +34 -47
  89. paddlex/inference/pipelines/layout_parsing/pipeline_v2.py +893 -260
  90. paddlex/inference/pipelines/layout_parsing/result.py +4 -17
  91. paddlex/inference/pipelines/layout_parsing/result_v2.py +523 -245
  92. paddlex/inference/pipelines/layout_parsing/setting.py +87 -0
  93. paddlex/inference/pipelines/layout_parsing/utils.py +565 -1998
  94. paddlex/inference/pipelines/layout_parsing/xycut_enhanced/__init__.py +16 -0
  95. paddlex/inference/pipelines/layout_parsing/xycut_enhanced/utils.py +1144 -0
  96. paddlex/inference/pipelines/layout_parsing/xycut_enhanced/xycuts.py +563 -0
  97. paddlex/inference/pipelines/m_3d_bev_detection/pipeline.py +2 -2
  98. paddlex/inference/pipelines/multilingual_speech_recognition/pipeline.py +2 -2
  99. paddlex/inference/pipelines/object_detection/pipeline.py +16 -6
  100. paddlex/inference/pipelines/ocr/pipeline.py +127 -70
  101. paddlex/inference/pipelines/ocr/result.py +19 -16
  102. paddlex/inference/pipelines/open_vocabulary_detection/pipeline.py +2 -2
  103. paddlex/inference/pipelines/open_vocabulary_segmentation/pipeline.py +2 -2
  104. paddlex/inference/pipelines/pp_chatocr/pipeline_base.py +2 -2
  105. paddlex/inference/pipelines/pp_chatocr/pipeline_v3.py +2 -5
  106. paddlex/inference/pipelines/pp_chatocr/pipeline_v4.py +5 -5
  107. paddlex/inference/pipelines/rotated_object_detection/pipeline.py +16 -6
  108. paddlex/inference/pipelines/seal_recognition/pipeline.py +109 -53
  109. paddlex/inference/pipelines/semantic_segmentation/pipeline.py +16 -6
  110. paddlex/inference/pipelines/small_object_detection/pipeline.py +16 -6
  111. paddlex/inference/pipelines/table_recognition/pipeline.py +26 -18
  112. paddlex/inference/pipelines/table_recognition/pipeline_v2.py +624 -53
  113. paddlex/inference/pipelines/table_recognition/result.py +1 -1
  114. paddlex/inference/pipelines/table_recognition/table_recognition_post_processing_v2.py +9 -5
  115. paddlex/inference/pipelines/ts_anomaly_detection/pipeline.py +2 -2
  116. paddlex/inference/pipelines/ts_classification/pipeline.py +2 -2
  117. paddlex/inference/pipelines/ts_forecasting/pipeline.py +2 -2
  118. paddlex/inference/pipelines/video_classification/pipeline.py +2 -2
  119. paddlex/inference/pipelines/video_detection/pipeline.py +2 -2
  120. paddlex/inference/serving/basic_serving/_pipeline_apps/_common/common.py +5 -1
  121. paddlex/inference/serving/basic_serving/_pipeline_apps/layout_parsing.py +0 -1
  122. paddlex/inference/serving/basic_serving/_pipeline_apps/pp_chatocrv3_doc.py +0 -1
  123. paddlex/inference/serving/basic_serving/_pipeline_apps/pp_chatocrv4_doc.py +1 -1
  124. paddlex/inference/serving/basic_serving/_pipeline_apps/pp_structurev3.py +6 -2
  125. paddlex/inference/serving/basic_serving/_pipeline_apps/table_recognition.py +1 -5
  126. paddlex/inference/serving/basic_serving/_pipeline_apps/table_recognition_v2.py +4 -5
  127. paddlex/inference/serving/infra/utils.py +20 -22
  128. paddlex/inference/serving/schemas/formula_recognition.py +1 -1
  129. paddlex/inference/serving/schemas/layout_parsing.py +1 -2
  130. paddlex/inference/serving/schemas/pp_chatocrv3_doc.py +1 -2
  131. paddlex/inference/serving/schemas/pp_chatocrv4_doc.py +2 -2
  132. paddlex/inference/serving/schemas/pp_structurev3.py +10 -6
  133. paddlex/inference/serving/schemas/seal_recognition.py +1 -1
  134. paddlex/inference/serving/schemas/table_recognition.py +2 -6
  135. paddlex/inference/serving/schemas/table_recognition_v2.py +5 -6
  136. paddlex/inference/utils/hpi.py +8 -1
  137. paddlex/inference/utils/hpi_model_info_collection.json +81 -2
  138. paddlex/inference/utils/io/readers.py +12 -12
  139. paddlex/inference/utils/mkldnn_blocklist.py +25 -0
  140. paddlex/inference/utils/official_models.py +14 -0
  141. paddlex/inference/utils/pp_option.py +29 -8
  142. paddlex/model.py +2 -2
  143. paddlex/modules/__init__.py +1 -1
  144. paddlex/modules/anomaly_detection/evaluator.py +2 -2
  145. paddlex/modules/base/__init__.py +1 -1
  146. paddlex/modules/base/evaluator.py +5 -5
  147. paddlex/modules/base/trainer.py +1 -1
  148. paddlex/modules/doc_vlm/dataset_checker.py +2 -2
  149. paddlex/modules/doc_vlm/evaluator.py +2 -2
  150. paddlex/modules/doc_vlm/exportor.py +2 -2
  151. paddlex/modules/doc_vlm/model_list.py +1 -1
  152. paddlex/modules/doc_vlm/trainer.py +2 -2
  153. paddlex/modules/face_recognition/evaluator.py +2 -2
  154. paddlex/modules/formula_recognition/evaluator.py +5 -2
  155. paddlex/modules/formula_recognition/model_list.py +3 -0
  156. paddlex/modules/formula_recognition/trainer.py +3 -0
  157. paddlex/modules/general_recognition/evaluator.py +1 -1
  158. paddlex/modules/image_classification/evaluator.py +2 -2
  159. paddlex/modules/image_classification/model_list.py +1 -0
  160. paddlex/modules/instance_segmentation/evaluator.py +1 -1
  161. paddlex/modules/keypoint_detection/evaluator.py +1 -1
  162. paddlex/modules/m_3d_bev_detection/evaluator.py +2 -2
  163. paddlex/modules/multilabel_classification/evaluator.py +2 -2
  164. paddlex/modules/object_detection/dataset_checker/dataset_src/convert_dataset.py +4 -4
  165. paddlex/modules/object_detection/evaluator.py +2 -2
  166. paddlex/modules/object_detection/model_list.py +2 -0
  167. paddlex/modules/semantic_segmentation/evaluator.py +2 -2
  168. paddlex/modules/table_recognition/evaluator.py +2 -2
  169. paddlex/modules/text_detection/evaluator.py +2 -2
  170. paddlex/modules/text_detection/model_list.py +2 -0
  171. paddlex/modules/text_recognition/evaluator.py +2 -2
  172. paddlex/modules/text_recognition/model_list.py +2 -0
  173. paddlex/modules/ts_anomaly_detection/evaluator.py +2 -2
  174. paddlex/modules/ts_classification/dataset_checker/dataset_src/split_dataset.py +1 -1
  175. paddlex/modules/ts_classification/evaluator.py +2 -2
  176. paddlex/modules/ts_forecast/evaluator.py +2 -2
  177. paddlex/modules/video_classification/evaluator.py +2 -2
  178. paddlex/modules/video_detection/evaluator.py +2 -2
  179. paddlex/ops/__init__.py +2 -2
  180. paddlex/paddlex_cli.py +19 -13
  181. paddlex/repo_apis/Paddle3D_api/bev_fusion/model.py +2 -2
  182. paddlex/repo_apis/PaddleClas_api/cls/config.py +1 -1
  183. paddlex/repo_apis/PaddleClas_api/cls/model.py +1 -1
  184. paddlex/repo_apis/PaddleClas_api/cls/register.py +10 -0
  185. paddlex/repo_apis/PaddleClas_api/cls/runner.py +1 -1
  186. paddlex/repo_apis/PaddleDetection_api/instance_seg/model.py +1 -1
  187. paddlex/repo_apis/PaddleDetection_api/instance_seg/runner.py +1 -1
  188. paddlex/repo_apis/PaddleDetection_api/object_det/config.py +1 -1
  189. paddlex/repo_apis/PaddleDetection_api/object_det/model.py +1 -1
  190. paddlex/repo_apis/PaddleDetection_api/object_det/official_categories.py +25 -0
  191. paddlex/repo_apis/PaddleDetection_api/object_det/register.py +30 -0
  192. paddlex/repo_apis/PaddleDetection_api/object_det/runner.py +1 -1
  193. paddlex/repo_apis/PaddleOCR_api/formula_rec/config.py +3 -3
  194. paddlex/repo_apis/PaddleOCR_api/formula_rec/model.py +5 -9
  195. paddlex/repo_apis/PaddleOCR_api/formula_rec/register.py +27 -0
  196. paddlex/repo_apis/PaddleOCR_api/formula_rec/runner.py +1 -1
  197. paddlex/repo_apis/PaddleOCR_api/table_rec/model.py +1 -1
  198. paddlex/repo_apis/PaddleOCR_api/table_rec/runner.py +1 -1
  199. paddlex/repo_apis/PaddleOCR_api/text_det/model.py +1 -1
  200. paddlex/repo_apis/PaddleOCR_api/text_det/register.py +18 -0
  201. paddlex/repo_apis/PaddleOCR_api/text_det/runner.py +1 -1
  202. paddlex/repo_apis/PaddleOCR_api/text_rec/config.py +3 -3
  203. paddlex/repo_apis/PaddleOCR_api/text_rec/model.py +5 -9
  204. paddlex/repo_apis/PaddleOCR_api/text_rec/register.py +18 -0
  205. paddlex/repo_apis/PaddleOCR_api/text_rec/runner.py +1 -1
  206. paddlex/repo_apis/PaddleSeg_api/seg/model.py +1 -1
  207. paddlex/repo_apis/PaddleSeg_api/seg/runner.py +1 -1
  208. paddlex/repo_apis/PaddleTS_api/ts_ad/config.py +3 -3
  209. paddlex/repo_apis/PaddleTS_api/ts_cls/config.py +2 -2
  210. paddlex/repo_apis/PaddleTS_api/ts_fc/config.py +4 -4
  211. paddlex/repo_apis/PaddleVideo_api/video_cls/config.py +1 -1
  212. paddlex/repo_apis/PaddleVideo_api/video_cls/model.py +1 -1
  213. paddlex/repo_apis/PaddleVideo_api/video_cls/runner.py +1 -1
  214. paddlex/repo_apis/PaddleVideo_api/video_det/config.py +1 -1
  215. paddlex/repo_apis/PaddleVideo_api/video_det/model.py +1 -1
  216. paddlex/repo_apis/PaddleVideo_api/video_det/runner.py +1 -1
  217. paddlex/repo_apis/base/config.py +1 -1
  218. paddlex/repo_manager/core.py +3 -3
  219. paddlex/repo_manager/meta.py +6 -2
  220. paddlex/repo_manager/repo.py +17 -16
  221. paddlex/utils/custom_device_list.py +26 -2
  222. paddlex/utils/deps.py +1 -1
  223. paddlex/utils/device.py +15 -8
  224. paddlex/utils/env.py +4 -0
  225. paddlex/utils/flags.py +2 -4
  226. paddlex/utils/fonts/__init__.py +34 -4
  227. paddlex/utils/misc.py +1 -1
  228. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/METADATA +52 -56
  229. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/RECORD +233 -206
  230. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/WHEEL +1 -1
  231. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/entry_points.txt +0 -0
  232. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/licenses/LICENSE +0 -0
  233. {paddlex-3.0.0rc1.dist-info → paddlex-3.0.1.dist-info}/top_level.txt +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: paddlex
3
- Version: 3.0.0rc1
3
+ Version: 3.0.1
4
4
  Summary: Low-code development tool based on PaddlePaddle.
5
5
  Author: PaddlePaddle Authors
6
6
  Author-email:
@@ -46,7 +46,7 @@ Requires-Dist: faiss-cpu; extra == "cv"
46
46
  Requires-Dist: matplotlib; extra == "cv"
47
47
  Requires-Dist: opencv-contrib-python==4.10.0.84; extra == "cv"
48
48
  Requires-Dist: pycocotools; extra == "cv"
49
- Requires-Dist: PyMuPDF; extra == "cv"
49
+ Requires-Dist: pypdfium2>=4; extra == "cv"
50
50
  Requires-Dist: scikit-image; extra == "cv"
51
51
  Provides-Extra: base
52
52
  Requires-Dist: chinese-calendar; extra == "base"
@@ -69,12 +69,13 @@ Requires-Dist: openpyxl; extra == "base"
69
69
  Requires-Dist: premailer; extra == "base"
70
70
  Requires-Dist: pyclipper; extra == "base"
71
71
  Requires-Dist: pycocotools; extra == "base"
72
- Requires-Dist: PyMuPDF; extra == "base"
72
+ Requires-Dist: pypdfium2>=4; extra == "base"
73
73
  Requires-Dist: regex; extra == "base"
74
74
  Requires-Dist: scikit-image; extra == "base"
75
75
  Requires-Dist: scikit-learn; extra == "base"
76
76
  Requires-Dist: shapely; extra == "base"
77
77
  Requires-Dist: soundfile; extra == "base"
78
+ Requires-Dist: tiktoken; extra == "base"
78
79
  Requires-Dist: tokenizers==0.19.1; extra == "base"
79
80
  Requires-Dist: tqdm; extra == "base"
80
81
  Provides-Extra: multimodal
@@ -82,8 +83,9 @@ Requires-Dist: einops; extra == "multimodal"
82
83
  Requires-Dist: ftfy; extra == "multimodal"
83
84
  Requires-Dist: Jinja2; extra == "multimodal"
84
85
  Requires-Dist: opencv-contrib-python==4.10.0.84; extra == "multimodal"
85
- Requires-Dist: PyMuPDF; extra == "multimodal"
86
+ Requires-Dist: pypdfium2>=4; extra == "multimodal"
86
87
  Requires-Dist: regex; extra == "multimodal"
88
+ Requires-Dist: tiktoken; extra == "multimodal"
87
89
  Provides-Extra: ie
88
90
  Requires-Dist: ftfy; extra == "ie"
89
91
  Requires-Dist: imagesize; extra == "ie"
@@ -97,7 +99,7 @@ Requires-Dist: opencv-contrib-python==4.10.0.84; extra == "ie"
97
99
  Requires-Dist: openpyxl; extra == "ie"
98
100
  Requires-Dist: premailer; extra == "ie"
99
101
  Requires-Dist: pyclipper; extra == "ie"
100
- Requires-Dist: PyMuPDF; extra == "ie"
102
+ Requires-Dist: pypdfium2>=4; extra == "ie"
101
103
  Requires-Dist: scikit-learn; extra == "ie"
102
104
  Requires-Dist: shapely; extra == "ie"
103
105
  Requires-Dist: tokenizers==0.19.1; extra == "ie"
@@ -109,7 +111,7 @@ Requires-Dist: opencv-contrib-python==4.10.0.84; extra == "ocr"
109
111
  Requires-Dist: openpyxl; extra == "ocr"
110
112
  Requires-Dist: premailer; extra == "ocr"
111
113
  Requires-Dist: pyclipper; extra == "ocr"
112
- Requires-Dist: PyMuPDF; extra == "ocr"
114
+ Requires-Dist: pypdfium2>=4; extra == "ocr"
113
115
  Requires-Dist: scikit-learn; extra == "ocr"
114
116
  Requires-Dist: shapely; extra == "ocr"
115
117
  Requires-Dist: tokenizers==0.19.1; extra == "ocr"
@@ -168,13 +170,14 @@ Requires-Dist: openpyxl; extra == "all"
168
170
  Requires-Dist: premailer; extra == "all"
169
171
  Requires-Dist: pyclipper; extra == "all"
170
172
  Requires-Dist: pycocotools; extra == "all"
171
- Requires-Dist: PyMuPDF; extra == "all"
173
+ Requires-Dist: pypdfium2>=4; extra == "all"
172
174
  Requires-Dist: regex; extra == "all"
173
175
  Requires-Dist: scikit-image; extra == "all"
174
176
  Requires-Dist: scikit-learn; extra == "all"
175
177
  Requires-Dist: shapely; extra == "all"
176
178
  Requires-Dist: soundfile; extra == "all"
177
179
  Requires-Dist: starlette>=0.36; extra == "all"
180
+ Requires-Dist: tiktoken; extra == "all"
178
181
  Requires-Dist: tokenizers==0.19.1; extra == "all"
179
182
  Requires-Dist: tqdm; extra == "all"
180
183
  Requires-Dist: uvicorn>=0.16; extra == "all"
@@ -214,16 +217,10 @@ Dynamic: summary
214
217
 
215
218
  PaddleX 3.0 是基于飞桨框架构建的低代码开发工具,它集成了众多**开箱即用的预训练模型**,可以实现模型从训练到推理的**全流程开发**,支持国内外**多款主流硬件**,助力AI 开发者进行产业实践。
216
219
 
217
- | [**通用图像分类**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/image_classification.html) | [**图像多标签分类**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/image_multi_label_classification.html) | [**通用目标检测**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/object_detection.html) | [**通用实例分割**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/instance_segmentation.html) |
218
- |:--------------------------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------------------------------------:|
219
- | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/b302cd7e-e027-4ea6-86d0-8a4dd6d61f39" height="126px" width="180px"> | <img src="https://raw.githubusercontent.com/cuicheng01/PaddleX_doc_images/main/images/multilabel_cls.png" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/099e2b00-0bbe-4b20-9c5a-96b69e473bd2" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/09f683b4-27df-4c24-b8a7-84da20fdd182" height="126px" width="180px"> |
220
- | [**通用语义分割**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/semantic_segmentation.html) | [**图像异常检测**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/cv_pipelines/image_anomaly_detection.html) | [ **通用OCR**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/OCR.html) | [**通用表格识别**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition.html) |
221
- | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/02637f8c-f248-415b-89ab-1276505f198c" height="126px" width="180px"> | <img src="https://raw.githubusercontent.com/cuicheng01/PaddleX_doc_images/main/images/image_anomaly_detection.png" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/1ef48536-48d4-484b-a6fb-0d6631ba2386" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/1e798e05-dee7-4b41-9cc4-6708b6014efa" height="126px" width="180px"> |
222
- | [**文本图像智能分析**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/information_extraction_pipelines/document_scene_information_extraction_v3.html) | [**时序预测**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/time_series_pipelines/time_series_forecasting.html) | [**时序异常检测**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/time_series_pipelines/time_series_anomaly_detection.html) | [**时序分类**](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/time_series_pipelines/time_series_classification.html) |
223
- | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/e3d97f4e-ab46-411c-8155-494c61492b0a" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/6e897bf6-35fe-45e6-a040-e9a1a20cfdf2" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/c54c66cc-da4f-4631-877b-43b0fbb192a6" height="126px" width="180px"> | <img src="https://github.com/PaddlePaddle/PaddleX/assets/142379845/0ce925b2-3776-4dde-8ce0-5156d5a2476e" height="126px" width="180px"> |
220
+ ![PaddleX](https://raw.githubusercontent.com/cuicheng01/PaddleX_doc_images/main/images/PaddleX_ch.png)
224
221
 
225
222
  ## 🌟 特性
226
- 🎨 **模型丰富一键调用**:将覆盖文本图像智能分析、OCR、目标检测、时序预测等多个关键领域的 **200+ 飞桨模型**整合为 **33 条模型产线**,通过极简的 Python API 一键调用,快速体验模型效果。同时支持 **38 种单功能模块**,方便开发者进行模型组合使用。
223
+ 🎨 **模型丰富一键调用**:将覆盖文本图像智能分析、OCR、目标检测、时序预测等多个关键领域的 **200+ 飞桨模型**整合为 **33 条模型产线**,通过极简的 Python API 一键调用,快速体验模型效果。同时支持 **39 种单功能模块**,方便开发者进行模型组合使用。
227
224
 
228
225
  🚀 **提高效率降低门槛**:实现基于统一命令和图形界面的模型**全流程开发**,打造大小模型结合、大模型半监督学习和多模型融合的[**8 条特色模型产线**](https://aistudio.baidu.com/intro/paddlex),大幅度降低迭代模型的成本。
229
226
 
@@ -233,46 +230,31 @@ PaddleX 3.0 是基于飞桨框架构建的低代码开发工具,它集成了
233
230
 
234
231
  ## 📣 近期更新
235
232
 
236
- 🔥🔥 **2025.4.22,发布 PaddleX v3.0.0rc1 。** 本次版本全面适配 PaddlePaddle 3.0正式版,核心升级如下:
237
233
 
238
- - **全面适配飞桨框架3.0新特性**:支持编译器训练,训练命令通过追加 `-o Global.dy2st=True` 即可开启编译器训练,在 GPU 上,多数模型训练速度可提升 10% 以上,少部分模型训练速度可以提升 30% 以上。推理方面,模型整体适配飞桨 3.0 中间表示技术(PIR),拥有更加灵活的扩展能力和兼容性,静态图模型存储文件名由 `xxx.pdmodel` 改为 `xxx.json`。
239
- - **新增飞桨自研文档图像理解多模态大模型 PP-DocBee**:在学术界及内部业务场景文档理解评测榜单上,PP-DocBee 均达到同参数量级别模型的 SOTA 水平。可应用到财报、研报、合同、说明书、法律法规等文档 QA 场景。
240
- - **全面支持 ONNX 格式模型,支持通过Paddle2ONNX插件转换模型格式。**
241
- - **升级高性能推理:**
242
- - **新增对 ONNX、OM 格式模型的支持:** PaddleX 可以根据需要智能选择模型格式;
243
- - **扩展支持的产线和模块:** 所有静态图推理的单功能模块与产线均可使用高性能推理插件来提升推理性能;
244
- - **支持 CLI、API、配置文件 3 种配置方式:** 支持更精细的配置,用户可以在子产线、子模块粒度启用和禁用高性能推理插件。
234
+ 🔥🔥 **2025.5.20,发布 PaddleX v3.0.0**,相比PaddleX v2.x,核心升级如下:
245
235
 
246
- - **多硬件支持扩展:**
247
- - **NPU:昇腾全面验证的模型数量提升到 200 个。此外,通用 OCR、图像分类、目标检测等常用产线支持 OM 模型格式推理,推理速度能够提升 113.8%-226.4%,支持在 Atlas 200、Atlas 300 系列产品上推理部署。**
248
- - **GCU:燧原正式纳入飞桨例行发版体系,完成了 PaddleX 生态适配。支持 90 个模型的训练和推理。**
236
+ **丰富的模型库:**
237
+ - **模型丰富:** PaddleX3.0 包含270+模型,涵盖了图像(视频)分类/检测/分割、OCR、语音识别、时序等多种场景。
238
+ - **方案成熟:** PaddleX3.0 基于丰富的模型库,**提供了通用文档解析、关键信息抽取、文档理解、表格识别、通用图像识别等多种重要且成熟的AI解决方案。**
249
239
 
240
+ **统一推理接口,重构部署能力:**
241
+ - **推理接口标准化**,降低不同种类模型带来的API接口差异,减少用户学习成本,提升企业落地效率。
242
+ - **提供多模型组合能力**,复杂任务可以通过不同的模型方便地进行组合使用,实现1+1>2 的能力。
243
+ - **部署能力升级,多种模型部署可以使用统一的命令管理,支持多卡推理,支持多卡多实例服务化部署。**
250
244
 
251
- 🔥 **2025.2.14,PaddleX v3.0.0rc0 重磅升级。** 本次版本全面适配 PaddlePaddle 3.0rc0及以上版本,核心升级如下:
245
+ **全面适配飞桨框架3.0:**
246
+ - **全面适配飞桨框架3.0新特性:** 支持编译器训练,训练命令通过追加 `-o Global.dy2st=True` 即可开启编译器训练,在 GPU 上,多数模型训练速度可提升 10% 以上,少部分模型训练速度可以提升 30% 以上。推理方面,模型整体适配飞桨 3.0 中间表示技术(PIR),拥有更加灵活的扩展能力和兼容性,静态图模型存储文件名由 `xxx.pdmodel` 改为 `xxx.json`。
247
+ - **全面支持 ONNX 格式模型:** 支持通过Paddle2ONNX插件转换模型格式。
252
248
 
253
- - **新增 12 条高价值产线,重磅推出自研 [通用版面解析v3产线](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/PP-StructureV3.html)、[PP-ChatOCRv4-doc产线](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/information_extraction_pipelines/document_scene_information_extraction_v4.html)、[表格识别v2产线](https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition_v2.html)**。此外新增了文档处理、旋转框检测、开放词汇检测/分割、视频分析、多语种语音识别、3D 等场景的产线。
254
-
255
- - **扩充 48 个前沿模型,包括重磅推出的 OCR 领域的版面区域检测模型 [PP-DocLayout](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/layout_detection.html)、公式识别模型 [PP-FormulaNet](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/formula_recognition.html),表格结构识别模型 [SLANeXt](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/table_structure_recognition.html),文本识别模型 [PP-OCRv4_server_rec_doc](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/text_recognition.html)**。CV 领域的 3D 检测、人体关键点、开放词汇检测/分割模型,以及语音识别领域的 Whisper 系列等模型。
256
-
257
- - **优化和升级模型和产线的推理 API:** 支持更多参数的配置,提升模型和产线推理的灵活性,[详情](docs/API_change_log/v3.0.0rc.md)。
258
-
259
- - **多硬件支持扩展:** 新增燧原 GCU 支持(90+模型),昇腾 NPU/昆仑芯 XPU/寒武纪 MLU/海光 DCU 模型数量显著提升。
260
-
261
- - **全场景部署能力升级:**
262
- - **高性能推理支持一键安装、Windows 系统及 220+ 模型,核心库 ultra-infer 开源;**
263
- - **服务化部署新增高稳定性方案,支持动态配置优化。**
264
-
265
- - **系统兼容性增强:** 适配 Windows 训练/推理,全面支持 Python 3.11/3.12。
266
-
267
- 🔥 **2024.11.15**,PaddleX 3.0 Beta2 开源版正式发布,全面适配 PaddlePaddle 3.0b2 版本。**新增通用图像识别、人脸识别、车辆属性识别和行人属性识别产线,同时新增 42 个模型开发全流程适配昇腾 910B,并全面支持[GitHub 站点文档](https://paddlepaddle.github.io/PaddleX/latest/index.html)。**
268
-
269
- 🔥 **2024.9.30**,PaddleX 3.0 Beta1 开源版正式发布,提供 **200+ 模型** 通过极简的 Python API 一键调用;实现基于统一命令的模型全流程开发,并开源 **PP-ChatOCRv3** 特色模型产线基础能力;支持 **100+ 模型高性能推理和服务化部署**(持续迭代中),**4条模型产线8个重点视觉模型端侧部署**;**100+ 模型开发全流程适配昇腾 910B**,**39+ 模型开发全流程适配昆仑芯和寒武纪**。
270
-
271
-
272
- 🔥 **2024.6.27**,PaddleX 3.0 Beta 开源版正式发布,支持以低代码的方式在本地端使用多种主流硬件进行产线和模型开发。
273
-
274
- 🔥 **2024.3.25**,PaddleX 3.0 云端发布,支持在 AI Studio 星河社区 以零代码的方式【创建产线】使用。
249
+ **重磅能力支撑:**
250
+ - **支撑PP-OCRv5的串联逻辑和多硬件推理、多后端推理、服务化部署能力。**
251
+ - **支撑PP-StructureV3的复杂模型串联和并联的逻辑,首次串联并联共15个模型,实现多模型协同的复杂pipeline。精度在 OmniDocBench 榜单上达到 SOTA 水平。**
252
+ - **支撑PP-ChatOCRv4的大模型串联逻辑,结合文心大模型4.5Turbo,结合新增的PP-DocBee2,关键信息抽取精度相比上一代提升15.7个百分点。**
275
253
 
254
+ **多硬件支持:**
255
+ - **整体支持英伟达、英特尔、苹果M系列、昆仑芯、昇腾、寒武纪、海光、燧原等芯片的训练和推理。**
256
+ - **在昇腾上,全面适配的模型达到200个,** 支持OM高性能推理的模型达到21个。此外支持PP-OCRv5、PP-StructureV3等重要模型方案。
257
+ - 在昆仑芯上支持重要分类、检测、OCR类模型(含PP-OCRv5)。
276
258
 
277
259
  ## 🔠 模型产线说明
278
260
 
@@ -309,7 +291,7 @@ PaddleX的各个产线均支持本地**快速推理**,部分模型支持在[AI
309
291
  <td>✅</td>
310
292
  </tr>
311
293
  <tr>
312
- <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/information_extraction_pipelines/document_scene_information_extraction.html">文档场景信息抽取v3</a></td>
294
+ <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/information_extraction_pipelines/document_scene_information_extraction_v3.html">文档场景信息抽取v3</a></td>
313
295
  <td><a href = "https://aistudio.baidu.com/community/app/182491/webUI?source=appCenter">链接</a></td>
314
296
  <td>✅</td>
315
297
  <td>✅</td>
@@ -318,6 +300,16 @@ PaddleX的各个产线均支持本地**快速推理**,部分模型支持在[AI
318
300
  <td>✅</td>
319
301
  <td>✅</td>
320
302
  </tr>
303
+ <tr>
304
+ <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/information_extraction_pipelines/document_scene_information_extraction_v4.html">文档场景信息抽取v4</a></td>
305
+ <td><a href = "https://aistudio.baidu.com/community/app/518493/webUI?source=appCenter">链接</a></td>
306
+ <td>✅</td>
307
+ <td>✅</td>
308
+ <td>✅</td>
309
+ <td>🚧</td>
310
+ <td>✅</td>
311
+ <td>✅</td>
312
+ </tr>
321
313
  <tr>
322
314
  <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition.html">通用表格识别</a></td>
323
315
  <td><a href = "https://aistudio.baidu.com/community/app/91661?source=appMineRecent">链接</a></td>
@@ -520,13 +512,13 @@ PaddleX的各个产线均支持本地**快速推理**,部分模型支持在[AI
520
512
  </tr>
521
513
  <tr>
522
514
  <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/table_recognition_v2.html">通用表格识别v2</a></td>
523
- <td>🚧</td>
515
+ <td><a href = "https://aistudio.baidu.com/community/app/518495/webUI?source=appCenter">链接</a></td>
524
516
  <td>✅</td>
525
517
  <td>✅</td>
526
518
  <td>✅</td>
527
519
  <td>🚧</td>
528
520
  <td>✅</td>
529
- <td>🚧</td>
521
+ <td>✅</td>
530
522
  </tr>
531
523
  <tr>
532
524
  <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/layout_parsing.html">通用版面解析</a></td>
@@ -540,13 +532,13 @@ PaddleX的各个产线均支持本地**快速推理**,部分模型支持在[AI
540
532
  </tr>
541
533
  <tr>
542
534
  <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/PP-StructureV3.html">通用版面解析v3</a></td>
543
- <td>🚧</td>
535
+ <td><a href = "https://aistudio.baidu.com/community/app/518494/webUI?source=appCente">链接</a></td>
544
536
  <td>✅</td>
545
537
  <td>✅</td>
546
538
  <td>✅</td>
547
539
  <td>🚧</td>
548
540
  <td>🚧</td>
549
- <td>🚧</td>
541
+ <td>✅</td>
550
542
  </tr>
551
543
  <tr>
552
544
  <td><a href="https://paddlepaddle.github.io/PaddleX/latest/pipeline_usage/tutorials/ocr_pipelines/doc_preprocessor.html">文档图像预处理</a></td>
@@ -752,7 +744,7 @@ PaddleX的各个产线均支持本地**快速推理**,部分模型支持在[AI
752
744
 
753
745
  ### 🛠️ 安装
754
746
 
755
- > ❗在安装 PaddleX 之前,请确保您已具备基本的 **Python 运行环境**(注:目前支持 Python 3.8 至 Python 3.12)。PaddleX 3.0-rc1 版本依赖的 PaddlePaddle 版本为 3.0.0 及以上版本,请在使用前务必保证版本的对应关系。
747
+ > ❗在安装 PaddleX 之前,请确保您已具备基本的 **Python 运行环境**(注:目前支持 Python 3.8 至 Python 3.12)。PaddleX 3.0.x 版本依赖的 PaddlePaddle 版本为 3.0.0 及以上版本,请在使用前务必保证版本的对应关系。
756
748
 
757
749
  * **安装 PaddlePaddle**
758
750
  ```bash
@@ -770,7 +762,7 @@ python -m pip install paddlepaddle==3.0.0 -i https://www.paddlepaddle.org.cn/pac
770
762
  * **安装PaddleX**
771
763
 
772
764
  ```bash
773
- pip install paddlex==3.0rc1
765
+ pip install "paddlex[base]==3.0.1"
774
766
  ```
775
767
 
776
768
  > ❗ 更多安装方式参考 [PaddleX 安装教程](https://paddlepaddle.github.io/PaddleX/latest/installation/installation.html)
@@ -1095,22 +1087,26 @@ for res in output:
1095
1087
  <summary> <b> 📦 3D </b></summary>
1096
1088
 
1097
1089
  * [📦 3D多模态融合检测模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/cv_modules/3d_bev_detection.html)
1090
+ </details>
1098
1091
 
1099
1092
  * <details open>
1100
1093
  <summary> <b> 🎤 语音识别 </b></summary>
1101
1094
 
1102
1095
  * [🌐 多语种语音识别模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/speech_modules/multilingual_speech_recognition.html)
1096
+ </details>
1103
1097
 
1104
1098
  * <details open>
1105
1099
  <summary> <b> 🎥 视频识别 </b></summary>
1106
1100
 
1107
1101
  * [📈 视频分类模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/video_modules/video_classification.html)
1108
1102
  * [🔍 视频检测模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/video_modules/video_detection.html)
1103
+ </details>
1109
1104
 
1110
1105
  * <details open>
1111
1106
  <summary> <b> 🌐 多模态视觉语言模型 </b></summary>
1112
1107
 
1113
1108
  * [📝 文档类视觉语言模型模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/vlm_modules/doc_vlm.html)
1109
+ * [📈 图表解析模块使用教程](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/vlm_modules/chart_parsing.html)
1114
1110
  </details>
1115
1111
 
1116
1112
  * <details>