虚拟机上可以跑通，上板子就挂了。。。。求指导！

[复制链接] · 发表于 2019-3-11 20:54:59

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：

只看该作者 · 发表于 2019-3-12 09:05:59

代码方便贴出来看下？

只看该作者 · 发表于 2019-3-12 09:16:06

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。

import numpy as np

import cv2

from rknn.api import RKNN



def show_outputs(outputs):

    output = outputs[0][0]

    output_sorted = sorted(output, reverse=True)

    top5_str = 'face_det_new\n-----TOP 5-----\n'

    for i in range(5):

        value = output_sorted[i]

        index = np.where(output == value)

        for j in range(len(index)):

            if (i + j) >= 5:

                break

            if value > 0:

                topi = '{}: {}\n'.format(index[j], value)

            else:

                topi = '-1: 0.0\n'

            top5_str += topi

    print(top5_str)



if __name__ == '__main__':



    # Create RKNN object

    rknn = RKNN()

    

    # pre-process config

    print('--> config model')

    rknn.config(channel_mean_value='103.94 116.78 123.68 58.82', reorder_channel='2 1 0')

    print('done')



    # Load tensorflow model

    print('--> Loading model')

    ret = rknn.load_caffe(model='./face_det_new.prototxt',

                          proto='caffe',

                          blobs='./face_det_new.caffemodel')

    if ret != 0:

        print('Load face_det_new failed! Ret = {}'.format(ret))

        exit(ret)

    print('done')



    # Build model

    print('--> Building model')

    ret = rknn.build(do_quantization=True, dataset='./dataset.txt')

    if ret != 0:

        print('Build face_det_new failed!')

        exit(ret)

    print('done')



    # Export rknn model

    print('--> Export RKNN model')

    ret = rknn.export_rknn('./face_det_new.rknn')

    if ret != 0:

        print('Export face_det_new.rknn failed!')

        exit(ret)

    print('done')



    # Set inputs

    img = cv2.imread('./cat_224x224.jpg')

    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)



    print('--> Init runtime environment')

    ret = rknn.init_runtime()

    if ret != 0:

        print('Init runtime environment failed')

        exit(ret)

    print('done')



    # Inference

    print('--> Running model')

    outputs = rknn.inference(inputs=[img])

    show_outputs(outputs)

    print('done')



    # perf

    print('--> Begin evaluate model performance')

    perf_results = rknn.eval_perf(inputs=[img])

    print('done')



    rknn.release()

复制代码

只看该作者 · 发表于 2019-3-12 11:51:03

顶一下。。。。。

只看该作者 · 发表于 2019-3-12 15:43:58

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?

只看该作者 · 发表于 2019-3-12 17:59:04

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.

虚拟机上可以跑通，上板子就挂了。。。。求指导！

本帖子中包含更多资源