Toybrick

标题: RKNN模型转换问题 [打印本页]

作者: xupeng 时间: 2019-11-11 14:25
标题: RKNN模型转换问题
您好，目前遇到转换模型的问题：tf 1.14
--> Loading model
E Unknow layer "tensorarraygatherv3"
E Catch exception when loading tensorflow model: /Users/xupeng/models-master/research/object_detection/result/frozen_inference_graph_1104.pb!
T Traceback (most recent call last):
T File "rknn/api/rknn_base.py", line 188, in rknn.api.rknn_base.RKNNBase.load_tensorflow
T File "rknn/base/RKNNlib/converter/convert_tf.py", line 594, in rknn.base.RKNNlib.converter.convert_tf.convert_tf.match_paragraph_and_param
T File "rknn/base/RKNNlib/RKNNnet.py", line 171, in rknn.base.RKNNlib.RKNNnet.RKNNNet.new_layer
T File "/Users/xupeng/py36/lib/python3.6/site-packages/rknn/base/RKNNlib/RKNNlog.py", line 267, in e
T raise ValueError(msg)
T ValueError:Unknow layer "tensorarraygatherv3"
done
这个问题有解决方法吗？谢谢

作者: xupeng 时间: 2019-11-11 14:30
tf 1.14, 这个算子是不支持吗？训练和转换都是在同一台机器的

作者: anmyxuxuxu 时间: 2019-11-12 15:56
求大神帮忙解答下
xah@xah-XPS-15-9570:~/RKNN/rknn-toolkit/example/rknn_convert$ python3.6 rknn_convert.py ./models/caffe/mobilenet_v2/ ./out_rknn/ false
--> Loading model...
into caffe
caffe end
done
--> Build RKNN model...
Segmentation fault (core dumped)
转换caffe mobilenet_v2模型出错

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