Created by: fimo723

你好，我采用PaddleDetection的tools/export_model.py执行导出模型相关操作，出下如下错误，未查找出相关问题 运行环境为ubuntu16.04 Cuda9+Cudnn7.3 使用PaddleDetection yolov3+mobilenetv1训练正常 使用 python tools/export_model.py -c configs/fruveg/yolov3_mobilenet_v1_voc.yml --output_dir=./inference_model -o weights=output/yolov3_mobilenet_v1_voc/80000 YoloTestFeed.image_shape=[3,608,608] 出现如下错误 Traceback (most recent call last): File "tools/export_model.py", line 118, in main() File "tools/export_model.py", line 88, in main test_feed = create(cfg.test_feed) File "/home/ryg/FruVeg/PaddleDetection/ppdet/core/workspace.py", line 160, in create "the module {} is not registered".format(name) AssertionError: the module YoloTestFee is not registered 使用 ython tools/export_model.py -c configs/fruveg/yolov3_mobilenet_v1_voc.yml --output_dir=./inference_model -o weights=output/yolov3_mobilenet_v1_voc/80000 同样出现错误 Traceback (most recent call last): File "tools/export_model.py", line 118, in main() File "tools/export_model.py", line 88, in main test_feed = create(cfg.test_feed) File "/home/ryg/FruVeg/PaddleDetection/ppdet/core/workspace.py", line 160, in create "the module {} is not registered".format(name) AssertionError: the module YoloTestFee is not registered 使用python tools/configure.py list检查 Available modules in the category 'op':

AnchorGenerator Wrapper for anchor_generator OP RPNTargetAssign Wrapper for rpn_target_assign OP GenerateProposals Wrapper for generate_proposals OP MaskAssigner Wrapper for generate_mask_labels OP MultiClassNMS Wrapper for multiclass_nms OP BBoxAssigner Wrapper for generate_proposal_labels OP RoIAlign Wrapper for roi_align OP RoIPool Wrapper for roi_pool OP MultiBoxHead Wrapper for multi_box_head OP SSDOutputDecoder Wrapper for detection_output OP RetinaTargetAssign Wrapper for retinanet_target_assign OP RetinaOutputDecoder Wrapper for retinanet_detection_output OP BoxCoder Wrapper for box_coder OP

Available modules in the category 'module':

MultiClassSoftNMS
RPNHead RPN Head FPNRPNHead RPN Head that supports FPN input YOLOv3Head Head block for YOLOv3 network RetinaHead Retina Head ResNet Residual Network, see https://arxiv.org/abs/1512.03385 ResNetC5 Residual Network, see https://arxiv.org/abs/1512.03385 ResNeXt ResNeXt, see https://arxiv.org/abs/1611.05431 ResNeXtC5 ResNeXt, see https://arxiv.org/abs/1611.05431 DarkNet DarkNet, see https://pjreddie.com/darknet/yolo/ MobileNet MobileNet v1, see https://arxiv.org/abs/1704.04861 SENet Squeeze-and-Excitation Networks, see https://arxiv.org/abs/1709.01507 SENetC5 Squeeze-and-Excitation Networks, see https://arxiv.org/abs/1709.01507 FPN Feature Pyramid Network, see https://arxiv.org/abs/1612.03144 VGG VGG, see https://arxiv.org/abs/1409.1556 BlazeNet BlazeFace, see https://arxiv.org/abs/1907.05047 FaceBoxNet FaceBoxes, see https://https://arxiv.org/abs/1708.05234 CBResNet CBNet, see https://arxiv.org/abs/1909.03625 FPNRoIAlign RoI align pooling for FPN feature maps XConvNormHead RCNN head with serveral convolution layers TwoFCHead RCNN head with two Fully Connected layers BBoxHead RCNN bbox head MaskHead RCNN mask head CascadeBBoxHead Cascade RCNN bbox head CascadeXConvNormHead RCNN head with serveral convolution layers CascadeTwoFCHead RCNN head with serveral convolution layers CascadeBBoxAssigner

Available modules in the category 'architecture':

FasterRCNN Faster R-CNN architecture, see https://arxiv.org/abs/1506.01497 MaskRCNN Mask R-CNN architecture, see https://arxiv.org/abs/1703.06870 CascadeRCNN Cascade R-CNN architecture, see https://arxiv.org/abs/1712.00726 CascadeMaskRCNN Cascade Mask R-CNN architecture, see https://arxiv.org/abs/1712.00726 CascadeRCNNClsAware Cascade R-CNN architecture, see https://arxiv.org/abs/1712.00726 YOLOv3 YOLOv3 network, see https://arxiv.org/abs/1804.02767 SSD Single Shot MultiBox Detector, see https://arxiv.org/abs/1512.02325 RetinaNet RetinaNet architecture, see https://arxiv.org/abs/1708.02002 BlazeFace BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs, FaceBoxes FaceBoxes: Sub-millisecond Neural Face Detection on Mobile GPUs,

Available modules in the category 'optim':

LearningRate Learning Rate configuration OptimizerBuilder Build optimizer handles

Available modules in the category 'data':

TrainFeed DataFeed encompasses all data loading related settings EvalFeed DataFeed encompasses all data loading related settings TestFeed DataFeed encompasses all data loading related settings FasterRCNNTrainFeed DataFeed encompasses all data loading related settings FasterRCNNEvalFeed DataFeed encompasses all data loading related settings FasterRCNNTestFeed DataFeed encompasses all data loading related settings MaskRCNNTrainFeed DataFeed encompasses all data loading related settings MaskRCNNEvalFeed DataFeed encompasses all data loading related settings MaskRCNNTestFeed DataFeed encompasses all data loading related settings SSDTrainFeed DataFeed encompasses all data loading related settings SSDEvalFeed DataFeed encompasses all data loading related settings SSDTestFeed DataFeed encompasses all data loading related settings YoloTrainFeed DataFeed encompasses all data loading related settings YoloEvalFeed DataFeed encompasses all data loading related settings YoloTestFeed DataFeed encompasses all data loading related settings 不存在缺少模块 盼复🙏