TensorFlow-gpu ImportError: DLL load failed: 找不到指定的程序。

最新推荐文章于 2024-11-05 14:25:41 发布

走错路的程序员

最新推荐文章于 2024-11-05 14:25:41 发布

阅读量310

点赞数

分类专栏：机器视觉 python 文章标签： tensorflow python pycharm

本文链接：https://blog.youkuaiyun.com/phker/article/details/120453284

版权

python 同时被 2 个专栏收录

56 篇文章

订阅专栏

机器视觉

40 篇文章

订阅专栏

本文记录了一次使用RKNN工具将YOLOv5模型从ONNX格式转换为RKNN格式过程中遇到的问题及解决方法。主要问题是由于TensorFlow GPU版本与相关依赖不匹配导致的DLL加载失败。解决方案包括调整依赖项版本、重新安装cudnn和cuda以及最终选择使用TensorFlow CPU版本。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

 Traceback (most recent call last):
  File "f:/project/yolov5/modelconvert/onnx2rknn.py", line 10, in <module>  
    rknn = RKNN()
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\rknn\api\rknn.py", line 47, in __init__
    self.config()
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\rknn\api\rknn.py", line 695, in config
    return self.rknn_base.config(args)
  File "rknn\api\rknn_base.py", line 74, in rknn.api.rknn_base.RKNNBase.config
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\rknn\base\acuitylib\__init__.py", line 1, in <module>
    from acuitylib.optimize.optimizer import Optimizer
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\rknn\base\acuitylib\__init__.py", line 1, in <module>
    from acuitylib.optimize.optimizer import Optimizer
  File "rknn\base\acuitylib\optimize\optimizer.py", line 14, in init rknn.base.acuitylib.optimize.optimizer
  File "rknn\base\acuitylib\optimize\rules\merge_layer.py", line 3, in init rknn.base.acuitylib.optimize.rules.merge_layer
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\rknn\base\acuitylib\layer\__init__.py", line 2, in <module>
    from acuitylib.layer.add import *
  File "rknn\base\acuitylib\layer\add.py", line 1, in init rknn.base.acuitylib.layer.add
  File "rknn\base\acuitylib\layer\acuitylayer.py", line 3, in init rknn.base.acuitylib.layer.acuitylayer
  File "rknn\base\acuitylib\xtf.py", line 1, in init rknn.base.acuitylib.xtf
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\__init__.py", line 28, in <module>
    from tensorflow.python import pywrap_tensorflow  # pylint: disable=unused-import
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\python\__init__.py", line 49, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\pytho    raise ImportError(msg)
ImportError: Traceback (most recent call last):
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "F:\project\yolov5\.env_yolov5_5.0\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)
  File "H:\Program Files\Python36\lib\imp.py", line 243, in load_module
    return load_dynamic(name, filename, file)
  File "H:\Program Files\Python36\lib\imp.py", line 343, in load_dynamic
    return _load(spec)
ImportError: DLL load failed: 找不到指定的程序。

如果装的是TensorFlow-gpu版, 应该是缺少文件 cudnn相关的文件, 重新安装一下cudnn和cuda
好像要把cudnn的文件复制到cuda下面.
参考: https://blog.youkuaiyun.com/qq_43511039/article/details/105590841

因为我也不需要用gpu进行训练, 只是用它转换模型, 所以我就抛弃了gpu版, 重新装了TensorFlow 的 cpu版,
问题就少了很多.

另外我感觉AI学术界的更新速度太快了, 版本问题太多了. 工程界面对各种版本更新带来的问题头大的要命, 要是能停1-2个月等一等就好了…

更新速度太快也是非常非常坑的事情.