trt-yolo app 测试“try increasing the workspace size with IBuilder::setMaxWorkspaceSize()” 错误处理

最新推荐文章于 2024-07-22 22:15:05 发布

原创最新推荐文章于 2024-07-22 22:15:05 发布 · 3.1k 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#deepstream #trt-yolo

C/C++ 专栏收录该内容

12 篇文章

订阅专栏

解决Jetson Nano上运行TRT-YOLO App时的内存不足错误，通过调整workspace size参数，成功生成engine并实现正常运行。

部署运行你感兴趣的模型镜像

环境：

  - Board:                                                                                                                  Author: Raffaello Bonghi
    * Name:           NVIDIA Jetson NANO/TX1                                                                                e-mail: raffaello@rnext.it
    * Type:           NANO/TX1
    * Jetpack:        4.3 [L4T 32.3.1]
    * GPU-Arch:       5.3
    * SN:             1422219037419
  - Libraries:
    * CUDA:           10.0.326
    * cuDNN:          7.6.3.28-1+cuda10.0
    * TensorRT:       6.0.1.10-1+cuda10.0
    * VisionWorks:    1.6.0.500n
    * OpenCV:         3.4.6 compiled CUDA: YES

在jetson nano 上测试deepstream/trt-yolo-app，在第一次运行生成engine时，出现错误：

UNKNOWN: --------------- Timing Runner: mm1_19 (MatrixMultiply)
UNKNOWN: Tactic: 0 skipped. Scratch requested: 2621440, available: 1048576
UNKNOWN: Tactic: 1 skipped. Scratch requested: 2621440, available: 1048576
UNKNOWN: Fastest Tactic: -3360065831133338131 Time: 3.40282e+38
ERROR: Internal error: could not find any implementation for node mm1_19, try increasing the workspace size with IBuilder::setMaxWorkspaceSize()
ERROR: ../builder/tacticOptimizer.cpp (1461) - OutOfMemory Error in computeCosts: 0
trt-yolo-app: /root/soft/deepstream_reference_apps/yolo/lib/yolo.cpp:458: void Yolo::createYOLOEngine(nvinfer1::DataType, Int8EntropyCalibrator*): Assertion `m_Engine != nullptr' failed.
Aborted (core dumped)

处理：

修改yolo.cpp，增加workspace size

MAX_WorkspaceSize(1 << 30)

重新编译trt-yolo-app，可以正常生成engine，测试正常

您可能感兴趣的与本文相关的镜像