模型转化失败，非量化时模型可以转成功，量化时失败

[复制链接] · 发表于 2020-8-17 14:28:32

本帖最后由 liuwenhua 于 2020-8-17 14:29 编辑

模型转化时，量化未开启时，可以转成功，量化开启开启时。报错如下：
E [Previous line repeated 4 more times]
E File "rknn\base\RKNNlib\RKNNnetbuilder.py", line 331, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E File "rknn\base\RKNNlib\RKNNnetbuilder.py", line 336, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E File "rknn\base\RKNNlib\layer\RKNNlayer.py", line 287, in rknn.base.RKNNlib.layer.RKNNlayer.RKNNLayer.compute_tensor
E File "rknn\base\RKNNlib\layer\reshapelayer.py", line 139, in rknn.base.RKNNlib.layer.reshapelayer.Reshape.compute_out_tensor
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\ops\gen_array_ops.py", line 7546, in reshape
E    "Reshape", tensor=tensor, shape=shape, name=name)
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 787, in _apply_op_helper
E    op_def=op_def)
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\util\deprecation.py", line 488, in new_func
E    return func(*args, **kwargs)
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\framework\ops.py", line 3272, in create_op
E    op_def=op_def)
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\framework\ops.py", line 1790, in __init__
E    control_input_ops)
E File "E:\usb_test\example\rknn_win_env\lib\site-packages\tensorflow\python\framework\ops.py", line 1629, in _create_c_op
E    raise ValueError(str(e))
E ValueError: Cannot reshape a tensor with 4096 elements to shape [1,2048] (2048 elements) for 'Reshape_510_5/Reshape_510_5' (op: 'Reshape') with input shapes: [2,2048,1,1], [2] and with input tensors computed as partial shapes: input[1] = [1,2048].
Build model failed!

运行代买如下

from rknn.api import RKNN
import cv2
import time
import numpy as np

if __name__ == '__main__':

# Create RKNN object
rknn = RKNN(verbose=True)
rknn.config(channel_mean_value='0 0 0 1', reorder_channel='0 1 2',batch_size = 1)
print('done')

# Load tensorflow model
ret = rknn.load_onnx(model = './model_embding_gpu_1.1.0.onnx')
if ret != 0:
      print('Load model failed!')
      exit(ret)
print('done')

# Build model
print('--> Building model')

#ret = rknn.build(do_quantization=False)
ret = rknn.build(do_quantization=True, dataset='./dataset.txt')
if ret != 0:
      print('Build model failed!')
      exit(ret)
print('done')
# Export rknn model
print('--> Export RKNN model')
ret = rknn.export_rknn('./person_reid_quantization.rknn')
if ret != 0:
      print('Export model failed!')
      exit(ret)
print('done')

rknn.release()