paddlex 3.0.0rc1__py3-none-any.whl → 3.0.2__py3-none-any.whl

paddlex/.version

@@ -1 +1 @@
-3.0.0.rc1
+3.0.2

paddlex/__init__.py

@@ -24,7 +24,7 @@ for mod in _SPECIAL_MODS:
 from . import version
 from .inference import create_pipeline, create_predictor
 from .model import create_model
-from .modules import build_dataset_checker, build_evaluater, build_trainer
+from .modules import build_dataset_checker, build_evaluator, build_trainer
 
 
 def _initialize():

paddlex/configs/modules/chart_parsing/PP-Chart2Table.yaml

@@ -0,0 +1,13 @@
+Global:
+  model: PP-Chart2Table
+  mode: predict # only support predict
+  device: gpu:0
+  output: "output"
+
+Predict:
+  batch_size: 1
+  model_dir: "/path/to/PP-Chart2Table"
+  input:
+    image: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/chart_parsing_02.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/doc_vlm/PP-DocBee2-3B.yaml

@@ -0,0 +1,14 @@
+Global:
+  model: PP-DocBee2-3B
+  mode: predict # only support predict
+  device: gpu:0
+  output: "output"
+
+Predict:
+  batch_size: 1
+  model_dir: "/path/to/PP-DocBee2-3B"
+  input:
+    image: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/medal_table.png"
+    query: "识别这份表格的内容, 以markdown格式输出"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-L.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-FormulaNet_plus-L
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "./dataset/ocr_rec_latexocr_dataset_example"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert: 
+    enable: False
+    src_dataset_type: FormulaRecDataset
+  split: 
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 20
+  batch_size: 3
+  learning_rate: 0.0001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-L_pretrained.pdparams
+  resume_path: null
+  log_interval: 20
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  batch_size: 3
+  weight_path: output/best_accuracy/best_accuracy.pdparams
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-L_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_formula_rec_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-M.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-FormulaNet_plus-M
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "./dataset/ocr_rec_latexocr_dataset_example"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert: 
+    enable: False
+    src_dataset_type: FormulaRecDataset
+  split: 
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 20
+  batch_size: 14
+  learning_rate: 0.0001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-M_pretrained.pdparams
+  resume_path: null
+  log_interval: 20
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  batch_size: 20
+  weight_path: output/best_accuracy/best_accuracy.pdparams
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-M_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_formula_rec_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/formula_recognition/PP-FormulaNet_plus-S.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-FormulaNet_plus-S
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "./dataset/ocr_rec_latexocr_dataset_example"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert: 
+    enable: False
+    src_dataset_type: FormulaRecDataset
+  split: 
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 20
+  batch_size: 14
+  learning_rate: 0.0001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-S_pretrained.pdparams
+  resume_path: null
+  log_interval: 20
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  batch_size: 20
+  weight_path: output/best_accuracy/best_accuracy.pdparams
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-FormulaNet_plus-S_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_formula_rec_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/layout_detection/PP-DocBlockLayout.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-DocBlockLayout
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "/paddle/dataset/paddlex/layout/det_layout_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  num_classes: 11
+  epochs_iters: 100
+  batch_size: 1
+  learning_rate: 0.0001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocBlockLayout_pretrained.pdparams
+  warmup_steps: 100
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+
+Evaluate:
+  weight_path: "output/best_model/best_model.pdparams"
+  log_interval: 10
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocBlockLayout_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_model/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/layout.jpg"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/layout_detection/PP-DocLayout-L.yaml

@@ -19,7 +19,7 @@ Train:
   epochs_iters: 100
   batch_size: 1
   learning_rate: 0.0001
-  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-L_pretrain.pdparams
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-L_pretrained.pdparams
   warmup_steps: 100
   resume_path: null
   log_interval: 10
@@ -30,7 +30,7 @@ Evaluate:
   log_interval: 10
 
 Export:
-  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-L_pretrain.pdparams
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-L_pretrained.pdparams
 
 Predict:
   batch_size: 1

paddlex/configs/modules/layout_detection/PP-DocLayout-M.yaml

@@ -19,7 +19,7 @@ Train:
   epochs_iters: 100
   batch_size: 1
   learning_rate: 0.0001
-  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-M_pretrain.pdparams
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-M_pretrained.pdparams
   warmup_steps: 100
   resume_path: null
   log_interval: 10
@@ -30,7 +30,7 @@ Evaluate:
   log_interval: 10
 
 Export:
-  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-M_pretrain.pdparams
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-M_pretrained.pdparams
 
 Predict:
   batch_size: 1

paddlex/configs/modules/layout_detection/PP-DocLayout-S.yaml

@@ -19,7 +19,7 @@ Train:
   epochs_iters: 100
   batch_size: 1
   learning_rate: 0.0001
-  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-S_pretrain.pdparams
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-S_pretrained.pdparams
   warmup_steps: 100
   resume_path: null
   log_interval: 10
@@ -30,7 +30,7 @@ Evaluate:
   log_interval: 10
 
 Export:
-  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-S_pretrain.pdparams
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout-S_pretrained.pdparams
 
 Predict:
   batch_size: 1

paddlex/configs/modules/layout_detection/PP-DocLayout_plus-L.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-DocLayout_plus-L
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "/paddle/dataset/paddlex/layout/det_layout_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  num_classes: 11
+  epochs_iters: 100
+  batch_size: 1
+  learning_rate: 0.0001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout_plus-L_pretrained.pdparams
+  warmup_steps: 100
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+
+Evaluate:
+  weight_path: "output/best_model/best_model.pdparams"
+  log_interval: 10
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-DocLayout_plus-L_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_model/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/layout.jpg"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/text_detection/PP-OCRv5_mobile_det.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-OCRv5_mobile_det
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  module: text_det
+  dataset_dir: "/paddle/dataset/paddlex/ocr_det/ocr_det_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 100
+  batch_size: 4
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_det_pretrained.pdparams
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_det_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/text_detection/PP-OCRv5_server_det.yaml

@@ -0,0 +1,40 @@
+Global:
+  model: PP-OCRv5_server_det
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  module: text_det
+  dataset_dir: "/paddle/dataset/paddlex/ocr_det/ocr_det_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 100
+  batch_size: 4
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_det_pretrained.pdparams
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_det_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/text_recognition/PP-OCRv5_mobile_rec.yaml

@@ -0,0 +1,39 @@
+Global:
+  model: PP-OCRv5_mobile_rec
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "/paddle/dataset/paddlex/ocr_rec/ocr_rec_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert: 
+    enable: False
+    src_dataset_type: null
+  split: 
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 20
+  batch_size: 8
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_rec_pretrained.pdparams
+  resume_path: null
+  log_interval: 20
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_mobile_rec_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_rec_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/text_recognition/PP-OCRv5_server_rec.yaml

@@ -0,0 +1,39 @@
+Global:
+  model: PP-OCRv5_server_rec
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "/paddle/dataset/paddlex/ocr_rec/ocr_rec_dataset_examples"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert: 
+    enable: False
+    src_dataset_type: null
+  split: 
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  epochs_iters: 20
+  batch_size: 8
+  learning_rate: 0.001
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_rec_pretrained.pdparams
+  resume_path: null
+  log_interval: 20
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_accuracy/best_accuracy.pdparams"
+  log_interval: 1
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-OCRv5_server_rec_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_accuracy/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_rec_001.png"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/modules/textline_orientation/PP-LCNet_x1_0_textline_ori.yaml

@@ -0,0 +1,41 @@
+Global:
+  model: PP-LCNet_x1_0_textline_ori
+  mode: check_dataset # check_dataset/train/evaluate/predict
+  dataset_dir: "/paddle/dataset/paddlex/cls/textline_orientation_example_data"
+  device: gpu:0,1,2,3
+  output: "output"
+
+CheckDataset:
+  convert:
+    enable: False
+    src_dataset_type: null
+  split:
+    enable: False
+    train_percent: null
+    val_percent: null
+
+Train:
+  num_classes: 2
+  epochs_iters: 20
+  batch_size: 32
+  learning_rate: 0.8
+  pretrain_weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-LCNet_x1_0_textline_ori_pretrained.pdparams
+  warmup_steps: 100
+  resume_path: null
+  log_interval: 10
+  eval_interval: 1
+  save_interval: 1
+
+Evaluate:
+  weight_path: "output/best_model/best_model.pdparams"
+  log_interval: 10
+
+Export:
+  weight_path: https://paddle-model-ecology.bj.bcebos.com/paddlex/official_pretrained_model/PP-LCNet_x1_0_textline_ori_pretrained.pdparams
+
+Predict:
+  batch_size: 1
+  model_dir: "output/best_model/inference"
+  input: "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/img_textline180_demo.jpg"
+  kernel_option:
+    run_mode: paddle

paddlex/configs/pipelines/OCR.yaml

@@ -24,21 +24,22 @@ SubPipelines:
 SubModules:
   TextDetection:
     module_name: text_detection
-    model_name: PP-OCRv4_mobile_det
+    model_name: PP-OCRv5_server_det
     model_dir: null
-    limit_side_len: 960
-    limit_type: max
+    limit_side_len: 64
+    limit_type: min
+    max_side_limit: 4000
     thresh: 0.3
     box_thresh: 0.6
-    unclip_ratio: 2.0
+    unclip_ratio: 1.5
   TextLineOrientation:
     module_name: textline_orientation
-    model_name: PP-LCNet_x0_25_textline_ori 
+    model_name: PP-LCNet_x1_0_textline_ori 
     model_dir: null
     batch_size: 6    
   TextRecognition:
     module_name: text_recognition
-    model_name: PP-OCRv4_mobile_rec 
+    model_name: PP-OCRv5_server_rec 
     model_dir: null
     batch_size: 6
     score_thresh: 0.0

paddlex/configs/pipelines/PP-ChatOCRv3-doc.yaml

@@ -99,9 +99,10 @@ SubPipelines:
             model_dir: null
             limit_side_len: 960
             limit_type: max
+            max_side_limit: 4000
             thresh: 0.3
             box_thresh: 0.6
-            unclip_ratio: 2.0
+            unclip_ratio: 1.5
             
           TextRecognition:
             module_name: text_recognition
@@ -138,6 +139,7 @@ SubPipelines:
                 model_dir: null
                 limit_side_len: 736
                 limit_type: min
+                max_side_limit: 4000
                 thresh: 0.2
                 box_thresh: 0.6
                 unclip_ratio: 0.5