fix_docs (#2178)

docs/module_usage/tutorials/ocr_modules/layout_detection.md

@@ -13,11 +13,11 @@
 |模型|mAP(0.5)（%）|GPU推理耗时（ms）|CPU推理耗时 (ms)|模型存储大小（M)|介绍|
 |-|-|-|-|-|-|
 |PicoDet-L_layout_3cls|89.3|15.7|159.8|22.6|基于PicoDet-L的高效率版面区域定位模型，包含3个类别：表格，图像和印章|
-|PicoDet_layout_1x|86.8|13.0|91.3|7.4|基于PicoDet-1x的高效率版面区域定位模型，包含文字、标题、表格、图片、列表|
-|RT-DETR-H_layout_17cls|92.6|115.1|3827.2|470.2|基于RT-DETR-H的的高精度版面区域定位模型，包含17个版面常见类别。|
+|PicoDet_layout_1x|86.8|13.0|91.3|7.4|基于PicoDet-1x的高效率版面区域定位模型，包含文字、标题、表格、图片、列表名|
+|RT-DETR-H_layout_17cls|92.6|115.1|3827.2|470.2|基于RT-DETR-H的的高精度版面区域定位模型，包含17个版面常见类别，分别是：段落标题、图片、文本、数字、摘要、内容、图表标题、公式、表格、表格标题、参考文献、文档标题、脚注、页眉、算法、页脚、印章|
 |RT-DETR-H_layout_3cls|95.9|114.6|3832.6|470.1|基于RT-DETR-H的的高精度版面区域定位模型，包含3个类别：表格，图像和印章|
 
-**注：以上精度指标的评估集是 PaddleOCR 自建的版面区域分析数据集，包含 1w 张图片。GPU 推理耗时基于 NVIDIA Tesla T4 机器，精度类型为 FP32， CPU 推理速度基于 Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz，线程数为 8，精度类型为 FP32。**
+**注：以上精度指标的评估集是 PaddleOCR 自建的版面区域分析数据集，包含中英文论文、杂志和研报等常见的 1w 张文档类型图片。GPU 推理耗时基于 NVIDIA Tesla T4 机器，精度类型为 FP32， CPU 推理速度基于 Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz，线程数为 8，精度类型为 FP32。**
 </details>
 
 ## 三、快速集成

docs/module_usage/tutorials/ocr_modules/layout_detection_en.md

@@ -14,10 +14,10 @@ The core task of structure analysis is to parse and segment the content of input
 |-|-|-|-|-|-|
 | PicoDet-L_layout_3cls | 89.3 | 15.7 | 159.8 | 22.6 | High-efficiency structure analysis model based on PicoDet-L, including 3 classes: table, image, and seal |
 | PicoDet_layout_1x | 86.8 | 13.0 | 91.3 | 7.4 | High-efficiency structure analysis model based on PicoDet-1x, including text, title, table, image, and list |
-| RT-DETR-H_layout_17cls | 92.6 | 115.1 | 3827.2 | 470.2 | High-precision structure analysis model based on RT-DETR-H, including 17 common layout categories. |
+| RT-DETR-H_layout_17cls | 92.6 | 115.1 | 3827.2 | 470.2 | High-precision structure analysis model based on RT-DETR-H, containing 17 common layout categories, namely: paragraph title, image, text, number, abstract, content, figure title, formula, table, table title, reference, document title, footnote, header, algorithm, footer, and seal. |
 | RT-DETR-H_layout_3cls | 95.9 | 114.6 | 3832.6 | 470.1 | High-precision structure analysis model based on RT-DETR-H, including 3 classes: table, image, and seal |
 
-**Note: The evaluation set for the above accuracy metrics is PaddleOCR's self-built layout region analysis dataset, containing 10,000 images. GPU inference time is based on an NVIDIA Tesla T4 machine with FP32 precision. CPU inference speed is based on an Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz with 8 threads and FP32 precision.**
+**Note: The evaluation set for the above accuracy metrics is PaddleOCR's self-built layout region analysis dataset, containing 10,000 images of common document types, including English and Chinese papers, magazines, research reports, etc. GPU inference time is based on an NVIDIA Tesla T4 machine with FP32 precision. CPU inference speed is based on an Intel(R) Xeon(R) Gold 5117 CPU @ 2.00GHz with 8 threads and FP32 precision.**
 </details>
 
 ## III. Quick Integration  <a id="quick"> </a> 

paddlex/configs/doc_text_orientation/PP-LCNet_x1_0_doc_ori.yaml

@@ -1,7 +1,7 @@
 Global:
   model: PP-LCNet_x1_0_doc_ori
   mode: check_dataset # check_dataset/train/evaluate/predict
-  dataset_dir: "/paddle/dataset/paddlex/cls/cls_flowers_examples"
+  dataset_dir: "/paddle/dataset/paddlex/cls/text_image_orientation"
   device: gpu:0,1,2,3
   output: "output"
 
@@ -15,7 +15,7 @@ CheckDataset:
     val_percent: null
 
 Train:
-  num_classes: 102
+  num_classes: 4
   epochs_iters: 50
   batch_size: 16
   learning_rate: 0.08