could not create a primitive descriptor for a reorder primitive

环境

【FastDeploy版本】： fastdeploy-gpu-python 1.0.7 paddlepaddle-gpu 2.5.2
【编译命令】pip install
【系统平台】: Windows x64(Windows11)
【硬件】： CUDA 11.8 CUDNN 8.6
【编译语言】： Python(3.8）

问题日志及出现问题的操作流程

import cv2
import fastdeploy.vision as vision

model = vision.detection.MaskRCNN("E:/PaddlePaddle/PaddleDetection/output_dir_inference_model/mask_rcnn_r50_vd_fpn_ssld_1x_coco/model.pdmodel",
                                 "E:/PaddlePaddle/PaddleDetection/output_dir_inference_model/mask_rcnn_r50_vd_fpn_ssld_1x_coco/model.pdiparams",
                                 "E:/PaddlePaddle/PaddleDetection/output_dir_inference_model/mask_rcnn_r50_vd_fpn_ssld_1x_coco/infer_cfg.yml")

im = cv2.imread(r"E:/20231214_10010118.jpg")

result = model.predict(im)
print(result)

vis_im = vision.vis_detection(im, result, score_threshold=0.5)
cv2.imwrite("vis_image.jpg", vis_im)

[INFO] fastdeploy/vision/common/processors/transform.cc(159)::fastdeploy::vision::FuseNormalizeColorConvert     BGR2RGB and Normalize are fused to Normalize with swap_rb=1
WARNING: Logging before InitGoogleLogging() is written to STDERR
W1219 12:03:14.043656 17448 analysis_config.cc:971] It is detected that mkldnn and memory_optimize_pass are enabled at the same time, but they are not supported yet. Currently, memory_optimize_pass is explicitly disabled
[INFO] fastdeploy/runtime/runtime.cc(273)::fastdeploy::Runtime::CreatePaddleBackend     Runtime initialized with Backend::PDINFER in Device::CPU.
Traceback (most recent call last):
  File "e:/2023_Code/py_2023/FD/cs.py", line 12, in <module>
    result = model.predict(im)
  File "E:\anaconda3\envs\paddlepaddle\lib\site-packages\fastdeploy\vision\detection\ppdet\__init__.py", line 126, in predict
    return self._model.predict(im)
RuntimeError: could not create a primitive descriptor for a reorder primitive

PaddlePaddle / FastDeploy

could not create a primitive descriptor for a reorder primitive #2328

环境

问题日志及出现问题的操作流程