|
本帖最后由 zhr_19970822 于 2021-1-22 19:09 编辑
LSTM是打算从pytorch转到onnx再转到RKNNpytorch中的代码:self.rnn = nn.LSTM(nIn, nHidden, num_layers=1, bidirectional=True) (pytorch转onnx是成功的了)
onnx转RKNN的时候报错:
--> config model
done
--> Loading model
W Call onnx.optimizer.optimize fail, skip optimize
E Calc node LSTM : LSTM_323 output shape fail
W ----------------Warning(1)----------------
E Catch exception when loading onnx model: test.onnx!
E Traceback (most recent call last):
E File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/shape_inference.py", line 57, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.shape_inference.infer_shape
E File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/ops/LSTM.py", line 122, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.ops.LSTM.LSTM
E File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/ops/LSTM.py", line 48, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.ops.LSTM.lstm_helper.__init__
E NotImplementedError
E During handling of the above exception, another exception occurred:
E Traceback (most recent call last):
E File "rknn/api/rknn_base.py", line 264, in rknn.api.rknn_base.RKNNBase.load_onnx
E File "rknn/base/RKNNlib/RK_nn.py", line 135, in rknn.base.RKNNlib.RK_nn.RKnn.load_onnx
E File "rknn/base/RKNNlib/app/importer/import_onnx.py", line 111, in rknn.base.RKNNlib.app.importer.import_onnx.Importonnx.run
E File "rknn/base/RKNNlib/converter/convert_onnx.py", line 85, in rknn.base.RKNNlib.converter.convert_onnx.convert_onnx.__init__
E File "rknn/base/RKNNlib/converter/convert_onnx.py", line 434, in rknn.base.RKNNlib.converter.convert_onnx.convert_onnx._shape_inference
E File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/shape_inference.py", line 62, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.shape_inference.infer_shape
E File "rknn/api/rknn_log.py", line 312, in rknn.api.rknn_log.RKNNLog.e
E ValueError: Calc node LSTM : LSTM_323 output shape fail
done
--> Building model
Traceback (most recent call last):
File "test.py", line 26, in <module>
ret = rknn.build(do_quantization=True, dataset='./dataset.txt')
File "/usr/local/lib/python3.6/dist-packages/rknn/api/rknn.py", line 228, in build
inputs = self.rknn_base.net.get_input_layers()
AttributeError: 'NoneType' object has no attribute 'get_input_layers'
我通过Netron查看了onnx中该层的参数,
![](http://data:image/png;base64,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)
发现这里面的参数除了sequence_lens没有,其他都是有定义的。
是这个sequence_lens没有的问题嘛?
如果是的话 要怎么修改呢? 如果不是的话 问题出在哪里?
|
|