Toybrick

标题: rknn-toolkit转换onnx模型出错 [打印本页]

---

作者: wujialiang    时间: 2020-2-21 16:26
标题: rknn-toolkit转换onnx模型出错
D calc tensor shape of Relu_1808ut0 [[0, 256, 100, 152]]
D calc tensor shape of Relu_1218ut0 [[0, 256, 100, 152]]
D calc tensor shape of Relu_1454ut0 [[1, 256, 25, 38]]
D calc tensor shape of Relu_1506ut0 [[1, 256, 25, 38]]
D calc tensor shape of Relu_1388ut0 [[0, 256, 50, 76]]
D calc tensor shape of Relu_1572ut0 [[1, 256, 13, 19]]
D calc tensor shape of Relu_1742ut0 [[1, 256, 7, 10]]
D calc tensor shape of Relu_1690ut0 [[1, 256, 7, 10]]
D calc tensor shape of Relu_1270ut0 [[0, 256, 100, 152]]
E Catch exception when loading onnx model: G:\phnix\AI\embed_mask\embedmask.onnx!
E Traceback (most recent call last):
E   File "rknn\api\rknn_base.py", line 469, in rknn.api.rknn_base.RKNNBase.load_onnx
E   File "rknn\base\RKNNlib\converter\convert_onnx.py", line 514, in rknn.base.RKNNlib.converter.convert_onnx.convert_onnx.__init__
E   File "rknn\base\RKNNlib\converter\convert_onnx.py", line 225, in rknn.base.RKNNlib.converter.convert_onnx.onnx_shape_infer_engine.infer_shape
E   File "C:\ProgramData\Anaconda3\envs\pytorch_gpu\lib\site-packages\onnx\shape_inference.py", line 36, in infer_shapes
E     inferred_model_str = C.infer_shapes(model_str)
E RuntimeError: Inferred shape and existing shape differ in dimension 0: (0) vs (1)
--> Load onnx model failed!



请问各位大神，是什么错误呢？求赐教，并且如何解决？

---

作者: jefferyzhang    时间: 2020-2-22 16:15
转换之前请先确认这个onnx是可以被加载和推理的。
这里的log是在onnx库里报出来的
\site-packages\onnx\shape_inference.py", line 36, in infer_shapes

---

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

Powered by Discuz! X3.3