Toybrick

标题: 虚拟机上可以跑通，上板子就挂了。。。。求指导！ [打印本页]
作者: zhouzhouzlove    时间: 2019-3-11 20:54
标题: 虚拟机上可以跑通，上板子就挂了。。。。求指导！
本帖最后由 zhouzhouzlove 于 2019-3-11 20:54 编辑

按照toolkit工具手册的流程图，比对并照搬example文件夹中的Mobilenet_V2代码，把*.prototxt和*.caffemodel换成了自己的caffe文件，然后上虚拟机上跑了一遍，没问题，能够跑通，
但是把文件复制到3399pro板子上的时候，就挂了，输出信息如下：
[toybrick@localhost 111]$ sudo python3 ./face_det_new.py
/usr/lib64/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
--> config model
done
--> Loading model
done
--> Building model
done
--> Export RKNN model
done
--> Init runtime environment
done
--> Running model
E Catch exception when setting inputs.
T Traceback (most recent call last):
T   File "rknn/api/rknn_base.py", line 614, in rknn.api.rknn_base.RKNNBase.inference
T   File "rknn/api/rknn_base.py", line 532, in rknn.api.rknn_base.RKNNBase.set_inputs
T   File "rknn/api/rknn_runtime.py", line 318, in rknn.api.rknn_runtime.RKNNRuntime.set_inputs
T Exception: Set inputs failed. error code: RKNN_ERR_PARAM_INVALID
Traceback (most recent call last):
  File "./face_det_new.py", line 72, in <module>
    show_outputs(outputs)
  File "./face_det_new.py", line 6, in show_outputs
    output = outputs[0][0]
TypeError: 'NoneType' object is not subscriptable
[toybrick@localhost 111]$




ubuntu虚拟机上结果如下：

作者: zhangzj    时间: 2019-3-12 09:05
代码方便贴出来看下？
作者: zhouzhouzlove    时间: 2019-3-12 09:16
这是照搬的test.py的内容，只替换了一下文件名，然后在虚拟机上可以运行。


  1. import numpy as np
  2. import cv2
  3. from rknn.api import RKNN

  5. def show_outputs(outputs):
  6.     output = outputs[0][0]
  7.     output_sorted = sorted(output, reverse=True)
  8.     top5_str = 'face_det_new\n-----TOP 5-----\n'
  9.     for i in range(5):
  10.         value = output_sorted[i]
  11.         index = np.where(output == value)
  12.         for j in range(len(index)):
  13.             if (i + j) >= 5:
  14.                 break
  15.             if value > 0:
  16.                 topi = '{}: {}\n'.format(index[j], value)
  17.             else:
  18.                 topi = '-1: 0.0\n'
  19.             top5_str += topi
  20.     print(top5_str)

  22. if __name__ == '__main__':

  24.     # Create RKNN object
  25.     rknn = RKNN()
  26.    
  27.     # pre-process config
  28.     print('--> config model')
  29.     rknn.config(channel_mean_value='103.94 116.78 123.68 58.82', reorder_channel='2 1 0')
  30.     print('done')

  32.     # Load tensorflow model
  33.     print('--> Loading model')
  34.     ret = rknn.load_caffe(model='./face_det_new.prototxt',
  35.                           proto='caffe',
  36.                           blobs='./face_det_new.caffemodel')
  37.     if ret != 0:
  38.         print('Load face_det_new failed! Ret = {}'.format(ret))
  39.         exit(ret)
  40.     print('done')

  42.     # Build model
  43.     print('--> Building model')
  44.     ret = rknn.build(do_quantization=True, dataset='./dataset.txt')
  45.     if ret != 0:
  46.         print('Build face_det_new failed!')
  47.         exit(ret)
  48.     print('done')

  50.     # Export rknn model
  51.     print('--> Export RKNN model')
  52.     ret = rknn.export_rknn('./face_det_new.rknn')
  53.     if ret != 0:
  54.         print('Export face_det_new.rknn failed!')
  55.         exit(ret)
  56.     print('done')

  58.     # Set inputs
  59.     img = cv2.imread('./cat_224x224.jpg')
  60.     img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

  62.     print('--> Init runtime environment')
  63.     ret = rknn.init_runtime()
  64.     if ret != 0:
  65.         print('Init runtime environment failed')
  66.         exit(ret)
  67.     print('done')

  69.     # Inference
  70.     print('--> Running model')
  71.     outputs = rknn.inference(inputs=[img])
  72.     show_outputs(outputs)
  73.     print('done')

  75.     # perf
  76.     print('--> Begin evaluate model performance')
  77.     perf_results = rknn.eval_perf(inputs=[img])
  78.     print('done')

  80.     rknn.release()
复制代码
作者: zhouzhouzlove    时间: 2019-3-12 11:51
顶一下。。。。。
作者: elooon    时间: 2019-3-12 15:43
interesting.
The exception below seems that somthing is wrong in inputs data
T Exception: Set inputs failed. error code: RKNN_ERR_PARAM_INVALID

So *.prototxt and *.caffemodel files are also mobilenet_v2 model? Can they be shared out?
作者: zhouzhouzlove    时间: 2019-3-12 17:59
elooon 发表于 2019-3-12 15:43
interesting.
The exception below seems that somthing is wrong in inputs data
T Exception: Set inputs ...

灵异事件~~~虚拟机和板子上（错误）的推理结果不一致！
http://t.rock-chips.com/forum.ph ... 235&fromuid=208
(出处: Toybrick)
This is the new problem，the results are not identical.



欢迎光临 Toybrick (https://t.rock-chips.com/) Powered by Discuz! X3.3