Toybrick

标题: Tensorflow1.14 转换pb模型输出不支持报错？？ [打印本页]

作者: zengwubin 时间: 2020-4-7 10:47
标题: Tensorflow1.14 转换pb模型输出不支持报错？？
rknn.config(channel_mean_value='127.5 127.5 127.5 127.5', reorder_channel='0 1 2')

# Load TensorFlow Model
print('--> Loading model')
rknn.load_tensorflow(tf_pb='./face_slim_pc.pb',
                  inputs=['input0'],
                  outputs=['444', 'concat_50', 'output0'],
                  input_size_list=[[3, INPUT_SIZE, INPUT_SIZE]])
print('done')

# Build Model
print('--> Building model')
rknn.build(do_quantization=True, dataset='./dataset.txt')
print('done')E 32 outputs, more than 16 outputs is not support in split layer!
W ----------------Warning(13)----------------
E Catch exception when loading tensorflow model: ./face_slim_pc.pb!
E Traceback (most recent call last):
E File "rknn/api/rknn_base.py", line 337, in rknn.api.rknn_base.RKNNBase.load_tensorflow
E File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 286, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build
E File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 305, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 305, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 305, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E [Previous line repeated 88 more times]
E File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 311, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E File "rknn/base/RKNNlib/layer/RKNNlayer.py", line 262, in rknn.base.RKNNlib.layer.RKNNlayer.RKNNLayer.compute_shape
E File "rknn/base/RKNNlib/layer/split.py", line 44, in rknn.base.RKNNlib.layer.split.Split.compute_out_shape
E File "rknn/api/rknn_log.py", line 312, in rknn.api.rknn_log.RKNNLog.e
E ValueError: 32 outputs, more than 16 outputs is not support in split layer!

作者: jefferyzhang 时间: 2020-4-7 11:54
你的output和rknn配置的output数量不符，而且输出超多
32 outputs, more than 16 outputs is not support in split layer!

欢迎光临 Toybrick (https://t.rock-chips.com/)