--> Config model
W config: 'target_platform' is None, use rk3566 as default, Please set according to the actual platform!
done
--> Loading model
E load_pytorch: The following operators are not implemented: ['aten::clamp_min']
W load_pytorch: ===================== WARN(2) =====================
E rknn-toolkit2 version: 1.3.0-11912b58
W load_pytorch: Catch exception when loading pytorch model: ValueError("The following operators are not implemented: ['aten::clamp_min']",)
E load_pytorch: Traceback (most recent call last):
E load_pytorch: File "rknn/api/rknn_base.py", line 1310, in rknn.api.rknn_base.RKNNBase.load_pytorch
E load_pytorch: File "rknn/base/convertor/torch2onnx/pytorch.py", line 4209, in rknn.base.convertor.torch2onnx.pytorch.from_pytorch
E load_pytorch: File "rknn/base/convertor/torch2onnx/pytorch.py", line 3341, in rknn.base.convertor.torch2onnx.pytorch.PyTorchOpConverter.report_missing_conversion
E load_pytorch: File "rknn/api/rknn_log.py", line 113, in rknn.api.rknn_log.RKNNLog.e
E load_pytorch: ValueError: The following operators are not implemented: ['aten::clamp_min']
W load_pytorch: Load_pytorch exception! Try another methed!
done
--> Building model
Build model failed!作者: jefferyzhang 时间: 2022-7-15 11:16
是的,从log来看就是aten::clamp_min这个算子不支持。
那就有3种做法:
要么你自定义op,
要么改模型改掉这个op
要么如果这个op在头或者尾,就拉出来自己实现,不用放rknn里