fix conflicts

doc/doc_ch/recognition.md

@@ -324,7 +324,6 @@ Eval:
 
 评估数据集可以通过 `configs/rec/rec_icdar15_train.yml`  修改Eval中的 `label_file_path` 设置。
 
-*注意* 评估时必须确保配置文件中 infer_img 字段为空
 ```
 # GPU 评估， Global.checkpoints 为待测权重
 python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_icdar15_train.yml -o Global.checkpoints={path/to/weights}/best_accuracy
@@ -342,7 +341,7 @@ python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec
 
 ```
 # 预测英文结果
-python3 tools/infer_rec.py -c configs/rec/rec_icdar15_train.yml -o Global.checkpoints={path/to/weights}/best_accuracy Global.infer_img=doc/imgs_words/en/word_1.png
+python3 tools/infer_rec.py -c configs/rec/rec_icdar15_train.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.load_static_weights=false Global.infer_img=doc/imgs_words/en/word_1.png
 ```
 
 预测图片：
@@ -361,7 +360,7 @@ infer_img: doc/imgs_words/en/word_1.png
 
 ```
 # 预测中文结果
-python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.checkpoints={path/to/weights}/best_accuracy Global.infer_img=doc/imgs_words/ch/word_1.jpg
+python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.load_static_weights=false Global.infer_img=doc/imgs_words/ch/word_1.jpg
 ```
 
 预测图片：

doc/doc_ch/reference.md

@@ -11,11 +11,12 @@
 }
 
 2. DB:
-@article{liao2019real,
-  title={Real-time Scene Text Detection with Differentiable Binarization},
+@inproceedings{liao2020real,
+  title={Real-Time Scene Text Detection with Differentiable Binarization.},
   author={Liao, Minghui and Wan, Zhaoyi and Yao, Cong and Chen, Kai and Bai, Xiang},
-  journal={arXiv preprint arXiv:1911.08947},
-  year={2019}
+  booktitle={AAAI},
+  pages={11474--11481},
+  year={2020}
 }
 
 3. DTRB:
@@ -37,10 +38,11 @@
 }
 
 5. SRN:
-@article{yu2020towards,
-  title={Towards Accurate Scene Text Recognition with Semantic Reasoning Networks},
-  author={Yu, Deli and Li, Xuan and Zhang, Chengquan and Han, Junyu and Liu, Jingtuo and Ding, Errui},
-  journal={arXiv preprint arXiv:2003.12294},
+@inproceedings{yu2020towards,
+  title={Towards accurate scene text recognition with semantic reasoning networks},
+  author={Yu, Deli and Li, Xuan and Zhang, Chengquan and Liu, Tao and Han, Junyu and Liu, Jingtuo and Ding, Errui},
+  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
+  pages={12113--12122},
   year={2020}
 }
 
@@ -52,4 +54,62 @@
   pages={9086--9095},
   year={2019}
 }
+
+7. CRNN:
+@article{shi2016end,
+  title={An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition},
+  author={Shi, Baoguang and Bai, Xiang and Yao, Cong},
+  journal={IEEE transactions on pattern analysis and machine intelligence},
+  volume={39},
+  number={11},
+  pages={2298--2304},
+  year={2016},
+  publisher={IEEE}
+}
+
+8. FPGM:
+@inproceedings{he2019filter,
+  title={Filter pruning via geometric median for deep convolutional neural networks acceleration},
+  author={He, Yang and Liu, Ping and Wang, Ziwei and Hu, Zhilan and Yang, Yi},
+  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
+  pages={4340--4349},
+  year={2019}
+}
+
+9. PACT:
+@article{choi2018pact,
+  title={Pact: Parameterized clipping activation for quantized neural networks},
+  author={Choi, Jungwook and Wang, Zhuo and Venkataramani, Swagath and Chuang, Pierce I-Jen and Srinivasan, Vijayalakshmi and Gopalakrishnan, Kailash},
+  journal={arXiv preprint arXiv:1805.06085},
+  year={2018}
+}
+
+10.Rosetta
+@inproceedings{borisyuk2018rosetta,
+  title={Rosetta: Large scale system for text detection and recognition in images},
+  author={Borisyuk, Fedor and Gordo, Albert and Sivakumar, Viswanath},
+  booktitle={Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining},
+  pages={71--79},
+  year={2018}
+}
+
+11.STAR-Net
+@inproceedings{liu2016star,
+  title={STAR-Net: A SpaTial Attention Residue Network for Scene Text Recognition.},
+  author={Liu, Wei and Chen, Chaofeng and Wong, Kwan-Yee K and Su, Zhizhong and Han, Junyu},
+  booktitle={BMVC},
+  volume={2},
+  pages={7},
+  year={2016}
+}
+
+12.RARE
+@inproceedings{shi2016robust,
+  title={Robust scene text recognition with automatic rectification},
+  author={Shi, Baoguang and Wang, Xinggang and Lyu, Pengyuan and Yao, Cong and Bai, Xiang},
+  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
+  pages={4168--4176},
+  year={2016}
+}
+
 ```

doc/doc_en/algorithm_overview_en.md

@@ -11,9 +11,9 @@ This tutorial lists the text detection algorithms and text recognition algorithm
 ### 1. Text Detection Algorithm
 
 PaddleOCR open source text detection algorithms list:
-- [x]  EAST([paper](https://arxiv.org/abs/1704.03155))
-- [x]  DB([paper](https://arxiv.org/abs/1911.08947))
-- [x]  SAST([paper](https://arxiv.org/abs/1908.05498) )(Baidu Self-Research)
+- [x]  EAST([paper](https://arxiv.org/abs/1704.03155))[2]
+- [x]  DB([paper](https://arxiv.org/abs/1911.08947))[1]
+- [x]  SAST([paper](https://arxiv.org/abs/1908.05498))[4]
 
 On the ICDAR2015 dataset, the text detection result is as follows:
 
@@ -39,11 +39,11 @@ For the training guide and use of PaddleOCR text detection algorithms, please re
 ### 2. Text Recognition Algorithm
 
 PaddleOCR open-source text recognition algorithms list:
-- [x]  CRNN([paper](https://arxiv.org/abs/1507.05717))
-- [x]  Rosetta([paper](https://arxiv.org/abs/1910.05085))
-- [ ]  STAR-Net([paper](http://www.bmva.org/bmvc/2016/papers/paper043/index.html)) coming soon
-- [ ]  RARE([paper](https://arxiv.org/abs/1603.03915v1)) coming soon
-- [ ]  SRN([paper](https://arxiv.org/abs/2003.12294) )(Baidu Self-Research) coming soon
+- [x]  CRNN([paper](https://arxiv.org/abs/1507.05717))[7]
+- [x]  Rosetta([paper](https://arxiv.org/abs/1910.05085))[10]
+- [ ]  STAR-Net([paper](http://www.bmva.org/bmvc/2016/papers/paper043/index.html))[11] coming soon
+- [ ]  RARE([paper](https://arxiv.org/abs/1603.03915v1))[12] coming soon
+- [ ]  SRN([paper](https://arxiv.org/abs/2003.12294))[5] coming soon
 
 Refer to [DTRB](https://arxiv.org/abs/1904.01906), the training and evaluation result of these above text recognition (using MJSynth and SynthText for training, evaluate on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE) is as follow:
 

doc/doc_en/angle_class_en.md

@@ -119,7 +119,7 @@ Use `Global.infer_img` to specify the path of the predicted picture or folder, a
 
 ```
 # Predict English results
-python3 tools/infer_cls.py -c configs/cls/cls_mv3.yml -o Global.checkpoints={path/to/weights}/best_accuracy Global.infer_img=doc/imgs_words_en/word_10.png
+python3 tools/infer_cls.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.load_static_weights=false Global.infer_img=doc/imgs_words_en/word_10.png
 ```
 
 Input image:

doc/doc_en/detection_en.md

@@ -113,16 +113,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml  -o Global.checkpoints="{pat
 
 Test the detection result on a single image:
 ```shell
-python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.checkpoints="./output/det_db/best_accuracy"
+python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false
 ```
 
 When testing the DB model, adjust the post-processing threshold:
 ```shell
-python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.checkpoints="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
+python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
 ```
 
 
 Test the detection result on all images in the folder:
 ```shell
-python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.checkpoints="./output/det_db/best_accuracy"
+python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false
 ```

doc/doc_en/inference_en.md

@@ -255,15 +255,18 @@ The following will introduce the lightweight Chinese recognition model inference
 For lightweight Chinese recognition model inference, you can execute the following commands:
 
 ```
-python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/ch/word_4.jpg" --rec_model_dir="./inference/rec_crnn/"
+# download CRNN text recognition inference model
+wget  https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar
+tar xf ch_ppocr_mobile_v2.0_rec_infer.tar
+python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_10.png" --rec_model_dir="ch_ppocr_mobile_v2.0_rec_infer"
 ```
 
-![](../imgs_words/ch/word_4.jpg)
+![](../imgs_words_en/word_10.png)
 
 After executing the command, the prediction results (recognized text and score) of the above image will be printed on the screen.
 
 ```bash
-Predicts of ./doc/imgs_words/ch/word_4.jpg:('实力活力', 0.98458153)
+Predicts of ./doc/imgs_words_en/word_10.png:('PAIN', 0.9897658)
 ```
 
 <a name="CTC-BASED_RECOGNITION"></a>
@@ -339,7 +342,12 @@ For angle classification model inference, you can execute the following commands
 ```
 python3 tools/infer/predict_cls.py --image_dir="./doc/imgs_words_en/word_10.png" --cls_model_dir="./inference/cls/"
 ```
-
+```
+# download text angle class inference model：
+wget  https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar
+tar xf ch_ppocr_mobile_v2.0_cls_infer.tar
+python3 tools/infer/predict_cls.py --image_dir="./doc/imgs_words_en/word_10.png" --cls_model_dir="ch_ppocr_mobile_v2.0_cls_infer"
+```
 ![](../imgs_words_en/word_10.png)
 
 After executing the command, the prediction results (classification angle and score) of the above image will be printed on the screen.

doc/doc_en/recognition_en.md

@@ -317,11 +317,11 @@ Eval:
 <a name="EVALUATION"></a>
 ### EVALUATION
 
-The evaluation data set can be modified via `configs/rec/rec_icdar15_reader.yml` setting of `label_file_path` in EvalReader.
+The evaluation dataset can be set by modifying the `Eval.dataset.label_file_list` field in the `configs/rec/rec_icdar15_train.yml` file.
 
 ```
 # GPU evaluation, Global.checkpoints is the weight to be tested
-python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_icdar15_reader.yml -o Global.checkpoints={path/to/weights}/best_accuracy
+python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_icdar15_train.yml -o Global.checkpoints={path/to/weights}/best_accuracy
 ```
 
 <a name="PREDICTION"></a>
@@ -336,7 +336,7 @@ The default prediction picture is stored in `infer_img`, and the weight is speci
 
 ```
 # Predict English results
-python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.checkpoints={path/to/weights}/best_accuracy TestReader.infer_img=doc/imgs_words/en/word_1.jpg
+python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.load_static_weights=false Global.infer_img=doc/imgs_words/en/word_1.jpg
 ```
 
 Input image:
@@ -354,7 +354,7 @@ The configuration file used for prediction must be consistent with the training.
 
 ```
 # Predict Chinese results
-python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.checkpoints={path/to/weights}/best_accuracy TestReader.infer_img=doc/imgs_words/ch/word_1.jpg
+python3 tools/infer_rec.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.load_static_weights=false Global.infer_img=doc/imgs_words/ch/word_1.jpg
 ```
 
 Input image:

paddleocr.py

@@ -262,8 +262,8 @@ class PaddleOCR(predict_system.TextSystem):
             logger.error('rec_algorithm must in {}'.format(SUPPORT_REC_MODEL))
             sys.exit(0)
 
-        postprocess_params.rec_char_dict_path = Path(
-            __file__).parent / postprocess_params.rec_char_dict_path
+        postprocess_params.rec_char_dict_path = str(
+            Path(__file__).parent / postprocess_params.rec_char_dict_path)
 
         # init det_model and rec_model
         super().__init__(postprocess_params)

ppocr/losses/det_basic_loss.py

@@ -45,7 +45,6 @@ class BalanceLoss(nn.Layer):
         self.balance_loss = balance_loss
         self.main_loss_type = main_loss_type
         self.negative_ratio = negative_ratio
-        self.main_loss_type = main_loss_type
         self.return_origin = return_origin
         self.eps = eps
 

ppocr/losses/det_sast_loss.py

@@ -19,7 +19,6 @@ from __future__ import print_function
 import paddle
 from paddle import nn
 from .det_basic_loss import DiceLoss
-import paddle.fluid as fluid
 import numpy as np
 
 
@@ -27,9 +26,7 @@ class SASTLoss(nn.Layer):
     """
     """
 
-    def __init__(self,
-                 eps=1e-6,
-                 **kwargs):
+    def __init__(self, eps=1e-6, **kwargs):
         super(SASTLoss, self).__init__()
         self.dice_loss = DiceLoss(eps=eps)
 
@@ -53,10 +50,12 @@ class SASTLoss(nn.Layer):
         score_loss = 1.0 - 2 * intersection / (union + 1e-5)
 
         #border loss
-        l_border_split, l_border_norm = paddle.split(l_border, num_or_sections=[4, 1], axis=1)
+        l_border_split, l_border_norm = paddle.split(
+            l_border, num_or_sections=[4, 1], axis=1)
         f_border_split = f_border
         border_ex_shape = l_border_norm.shape * np.array([1, 4, 1, 1])
-        l_border_norm_split = paddle.expand(x=l_border_norm, shape=border_ex_shape)
+        l_border_norm_split = paddle.expand(
+            x=l_border_norm, shape=border_ex_shape)
         l_border_score = paddle.expand(x=l_score, shape=border_ex_shape)
         l_border_mask = paddle.expand(x=l_mask, shape=border_ex_shape)
 
@@ -72,7 +71,8 @@ class SASTLoss(nn.Layer):
                     (paddle.sum(l_border_score * l_border_mask) + 1e-5)
 
         #tvo_loss
-        l_tvo_split, l_tvo_norm = paddle.split(l_tvo, num_or_sections=[8, 1], axis=1)
+        l_tvo_split, l_tvo_norm = paddle.split(
+            l_tvo, num_or_sections=[8, 1], axis=1)
         f_tvo_split = f_tvo
         tvo_ex_shape = l_tvo_norm.shape * np.array([1, 8, 1, 1])
         l_tvo_norm_split = paddle.expand(x=l_tvo_norm, shape=tvo_ex_shape)
@@ -91,7 +91,8 @@ class SASTLoss(nn.Layer):
                     (paddle.sum(l_tvo_score * l_tvo_mask) + 1e-5)
 
         #tco_loss
-        l_tco_split, l_tco_norm = paddle.split(l_tco, num_or_sections=[2, 1], axis=1)
+        l_tco_split, l_tco_norm = paddle.split(
+            l_tco, num_or_sections=[2, 1], axis=1)
         f_tco_split = f_tco
         tco_ex_shape = l_tco_norm.shape * np.array([1, 2, 1, 1])
         l_tco_norm_split = paddle.expand(x=l_tco_norm, shape=tco_ex_shape)
@@ -109,7 +110,6 @@ class SASTLoss(nn.Layer):
         tco_loss = paddle.sum(tco_out_loss * l_tco_score * l_tco_mask) / \
                     (paddle.sum(l_tco_score * l_tco_mask) + 1e-5)
 
-
         # total loss
         tvo_lw, tco_lw = 1.5, 1.5
         score_lw, border_lw = 1.0, 1.0

setup.py

@@ -32,7 +32,7 @@ setup(
     package_dir={'paddleocr': ''},
     include_package_data=True,
     entry_points={"console_scripts": ["paddleocr= paddleocr.paddleocr:main"]},
-    version='2.0.1',
+    version='2.0.2',
     install_requires=requirements,
     license='Apache License 2.0',
     description='Awesome OCR toolkits based on PaddlePaddle （8.6M ultra-lightweight pre-trained model, support training and deployment among server, mobile, embeded and IoT devices',

tools/infer/predict_cls.py

@@ -24,7 +24,6 @@ import numpy as np
 import math
 import time
 import traceback
-import paddle.fluid as fluid
 
 import tools.infer.utility as utility
 from ppocr.postprocess import build_post_process
@@ -39,7 +38,6 @@ class TextClassifier(object):
         self.cls_image_shape = [int(v) for v in args.cls_image_shape.split(",")]
         self.cls_batch_num = args.cls_batch_num
         self.cls_thresh = args.cls_thresh
-        self.use_zero_copy_run = args.use_zero_copy_run
         postprocess_params = {
             'name': 'ClsPostProcess',
             "label_list": args.label_list,
@@ -99,12 +97,8 @@ class TextClassifier(object):
             norm_img_batch = norm_img_batch.copy()
             starttime = time.time()
 
-            if self.use_zero_copy_run:
             self.input_tensor.copy_from_cpu(norm_img_batch)
-                self.predictor.zero_copy_run()
-            else:
-                norm_img_batch = fluid.core.PaddleTensor(norm_img_batch)
-                self.predictor.run([norm_img_batch])
+            self.predictor.run()
             prob_out = self.output_tensors[0].copy_to_cpu()
             cls_result = self.postprocess_op(prob_out)
             elapse += time.time() - starttime
@@ -143,10 +137,11 @@ def main(args):
             "Please set --rec_image_shape='3,32,100' and --rec_char_type='en' ")
         exit()
     for ino in range(len(img_list)):
-        logger.info("Predicts of {}:{}".format(valid_image_file_list[ino], cls_res[
-            ino]))
+        logger.info("Predicts of {}:{}".format(valid_image_file_list[ino],
+                                               cls_res[ino]))
     logger.info("Total predict time for {} images, cost: {:.3f}".format(
         len(img_list), predict_time))
 
+
 if __name__ == "__main__":
     main(utility.parse_args())

tools/infer/predict_det.py

@@ -22,7 +22,6 @@ import cv2
 import numpy as np
 import time
 import sys
-import paddle
 
 import tools.infer.utility as utility
 from ppocr.utils.logging import get_logger
@@ -35,8 +34,8 @@ logger = get_logger()
 
 class TextDetector(object):
     def __init__(self, args):
+        self.args = args
         self.det_algorithm = args.det_algorithm
-        self.use_zero_copy_run = args.use_zero_copy_run
         pre_process_list = [{
             'DetResizeForTest': {
                 'limit_side_len': args.det_limit_side_len,
@@ -70,6 +69,11 @@ class TextDetector(object):
             postprocess_params["cover_thresh"] = args.det_east_cover_thresh
             postprocess_params["nms_thresh"] = args.det_east_nms_thresh
         elif self.det_algorithm == "SAST":
+            pre_process_list[0] = {
+                'DetResizeForTest': {
+                    'resize_long': args.det_limit_side_len
+                }
+            }
             postprocess_params['name'] = 'SASTPostProcess'
             postprocess_params["score_thresh"] = args.det_sast_score_thresh
             postprocess_params["nms_thresh"] = args.det_sast_nms_thresh
@@ -157,12 +161,8 @@ class TextDetector(object):
         img = img.copy()
         starttime = time.time()
 
-        if self.use_zero_copy_run:
         self.input_tensor.copy_from_cpu(img)
-            self.predictor.zero_copy_run()
-        else:
-            im = paddle.fluid.core.PaddleTensor(img)
-            self.predictor.run([im])
+        self.predictor.run()
         outputs = []
         for output_tensor in self.output_tensors:
             output = output_tensor.copy_to_cpu()

tools/infer/predict_rec.py

@@ -23,7 +23,6 @@ import numpy as np
 import math
 import time
 import traceback
-import paddle.fluid as fluid
 
 import tools.infer.utility as utility
 from ppocr.postprocess import build_post_process
@@ -39,7 +38,6 @@ class TextRecognizer(object):
         self.character_type = args.rec_char_type
         self.rec_batch_num = args.rec_batch_num
         self.rec_algorithm = args.rec_algorithm
-        self.use_zero_copy_run = args.use_zero_copy_run
         postprocess_params = {
             'name': 'CTCLabelDecode',
             "character_type": args.rec_char_type,
@@ -101,12 +99,8 @@ class TextRecognizer(object):
             norm_img_batch = np.concatenate(norm_img_batch)
             norm_img_batch = norm_img_batch.copy()
             starttime = time.time()
-            if self.use_zero_copy_run:
             self.input_tensor.copy_from_cpu(norm_img_batch)
-                self.predictor.zero_copy_run()
-            else:
-                norm_img_batch = fluid.core.PaddleTensor(norm_img_batch)
-                self.predictor.run([norm_img_batch])
+            self.predictor.run()
             outputs = []
             for output_tensor in self.output_tensors:
                 output = output_tensor.copy_to_cpu()
@@ -145,8 +139,8 @@ def main(args):
             "Please set --rec_image_shape='3,32,100' and --rec_char_type='en' ")
         exit()
     for ino in range(len(img_list)):
-        logger.info("Predicts of {}:{}".format(valid_image_file_list[ino], rec_res[
-            ino]))
+        logger.info("Predicts of {}:{}".format(valid_image_file_list[ino],
+                                               rec_res[ino]))
     logger.info("Total predict time for {} images, cost: {:.3f}".format(
         len(img_list), predict_time))
 

tools/infer/utility.py

@@ -20,8 +20,7 @@ import numpy as np
 import json
 from PIL import Image, ImageDraw, ImageFont
 import math
-from paddle.fluid.core import AnalysisConfig
-from paddle.fluid.core import create_paddle_predictor
+from paddle import inference
 
 
 def parse_args():
@@ -33,6 +32,7 @@ def parse_args():
     parser.add_argument("--use_gpu", type=str2bool, default=True)
     parser.add_argument("--ir_optim", type=str2bool, default=True)
     parser.add_argument("--use_tensorrt", type=str2bool, default=False)
+    parser.add_argument("--use_fp16", type=str2bool, default=False)
     parser.add_argument("--gpu_mem", type=int, default=8000)
 
     # params for text detector
@@ -46,7 +46,7 @@ def parse_args():
     parser.add_argument("--det_db_thresh", type=float, default=0.3)
     parser.add_argument("--det_db_box_thresh", type=float, default=0.5)
     parser.add_argument("--det_db_unclip_ratio", type=float, default=1.6)
-
+    parser.add_argument("--max_batch_size", type=int, default=10)
     # EAST parmas
     parser.add_argument("--det_east_score_thresh", type=float, default=0.8)
     parser.add_argument("--det_east_cover_thresh", type=float, default=0.1)
@@ -62,7 +62,7 @@ def parse_args():
     parser.add_argument("--rec_model_dir", type=str)
     parser.add_argument("--rec_image_shape", type=str, default="3, 32, 320")
     parser.add_argument("--rec_char_type", type=str, default='ch')
-    parser.add_argument("--rec_batch_num", type=int, default=6)
+    parser.add_argument("--rec_batch_num", type=int, default=1)
     parser.add_argument("--max_text_length", type=int, default=25)
     parser.add_argument(
         "--rec_char_dict_path",
@@ -78,12 +78,10 @@ def parse_args():
     parser.add_argument("--cls_model_dir", type=str)
     parser.add_argument("--cls_image_shape", type=str, default="3, 48, 192")
     parser.add_argument("--label_list", type=list, default=['0', '180'])
-    parser.add_argument("--cls_batch_num", type=int, default=30)
+    parser.add_argument("--cls_batch_num", type=int, default=6)
     parser.add_argument("--cls_thresh", type=float, default=0.9)
 
     parser.add_argument("--enable_mkldnn", type=str2bool, default=False)
-    parser.add_argument("--use_zero_copy_run", type=str2bool, default=False)
-
     parser.add_argument("--use_pdserving", type=str2bool, default=False)
 
     return parser.parse_args()
@@ -109,10 +107,15 @@ def create_predictor(args, mode, logger):
         logger.info("not find params file path {}".format(params_file_path))
         sys.exit(0)
 
-    config = AnalysisConfig(model_file_path, params_file_path)
+    config = inference.Config(model_file_path, params_file_path)
 
     if args.use_gpu:
         config.enable_use_gpu(args.gpu_mem, 0)
+        if args.use_tensorrt:
+            config.enable_tensorrt_engine(
+                precision_mode=inference.PrecisionType.Half
+                if args.use_fp16 else inference.PrecisionType.Float32,
+                max_batch_size=args.max_batch_size)
     else:
         config.disable_gpu()
         config.set_cpu_math_library_num_threads(6)
@@ -124,20 +127,18 @@ def create_predictor(args, mode, logger):
     # config.enable_memory_optim()
     config.disable_glog_info()
 
-    if args.use_zero_copy_run:
     config.delete_pass("conv_transpose_eltwiseadd_bn_fuse_pass")
     config.switch_use_feed_fetch_ops(False)
-    else:
-        config.switch_use_feed_fetch_ops(True)
 
-    predictor = create_paddle_predictor(config)
+    # create predictor
+    predictor = inference.create_predictor(config)
     input_names = predictor.get_input_names()
     for name in input_names:
-        input_tensor = predictor.get_input_tensor(name)
+        input_tensor = predictor.get_input_handle(name)
     output_names = predictor.get_output_names()
     output_tensors = []
     for output_name in output_names:
-        output_tensor = predictor.get_output_tensor(output_name)
+        output_tensor = predictor.get_output_handle(output_name)
         output_tensors.append(output_tensor)
     return predictor, input_tensor, output_tensors
 

tools/program.py

@@ -131,7 +131,7 @@ def check_gpu(use_gpu):
           "model on CPU"
 
     try:
-        if use_gpu and not paddle.fluid.is_compiled_with_cuda():
+        if use_gpu and not paddle.is_compiled_with_cuda():
             print(err)
             sys.exit(1)
     except Exception as e:
@@ -332,7 +332,7 @@ def eval(model, valid_dataloader, post_process_class, eval_class):
     return metirc
 
 
-def preprocess():
+def preprocess(is_train=False):
     FLAGS = ArgsParser().parse_args()
     config = load_config(FLAGS.config)
     merge_config(FLAGS.opt)
@@ -350,15 +350,17 @@ def preprocess():
     device = paddle.set_device(device)
 
     config['Global']['distributed'] = dist.get_world_size() != 1
-
+    if is_train:
         # save_config
         save_model_dir = config['Global']['save_model_dir']
         os.makedirs(save_model_dir, exist_ok=True)
         with open(os.path.join(save_model_dir, 'config.yml'), 'w') as f:
-        yaml.dump(dict(config), f, default_flow_style=False, sort_keys=False)
-
-    logger = get_logger(
-        name='root', log_file='{}/train.log'.format(save_model_dir))
+            yaml.dump(
+                dict(config), f, default_flow_style=False, sort_keys=False)
+        log_file = '{}/train.log'.format(save_model_dir)
+    else:
+        log_file = None
+    logger = get_logger(name='root', log_file=log_file)
     if config['Global']['use_visualdl']:
         from visualdl import LogWriter
         vdl_writer_path = '{}/vdl/'.format(save_model_dir)

tools/train.py

@@ -110,6 +110,6 @@ def test_reader(config, device, logger):
 
 
 if __name__ == '__main__':
-    config, device, logger, vdl_writer = program.preprocess()
+    config, device, logger, vdl_writer = program.preprocess(is_train=True)
     main(config, device, logger, vdl_writer)
     # test_reader(config, device, logger)