目录
sam部署教程：
segment-anyghing地址：
sam onnx地址：
报错：
encode继续报错：
sam部署教程：
Ascend/ModelZoo-PyTorch - Gitee.com
segment-anyghing地址：
https://github.com/visheratin/segment-anything
https://github.com/OroChippw/segment-anything
ais_bench
sam onnx地址：
https://github.com/OroChippw/SegmentAnything-OnnxRunner?tab=readme-ov-file
报错：
import auto_optimizer
解决方法：安装msit
msit: 统一推理工具链入口，提供客户一体化开发工具，支持一站式调试调优 - Gitee.com
encode继续报错：
Traceback (most recent call last):
File "/root/lbg/project/ACL_PyTorch/built-in/cv/SAM/encoder_onnx_modify.py", line 158, in <module>
graph = fa_adapter.adapt()
File "/root/lbg/project/ACL_PyTorch/built-in/cv/SAM/encoder_onnx_modify.py", line 29, in adapt
self.__adapt_layer(softmax)
File "/root/lbg/project/ACL_PyTorch/built-in/cv/SAM/encoder_onnx_modify.py", line 72, in __adapt_layer
target_shape = self.__infer_target_shape(layer_name_prefix, q)
File "/root/lbg/project/ACL_PyTorch/built-in/cv/SAM/encoder_onnx_modify.py", line 110, in __infer_target_shape
shape_q = self.__graph.add_node(
File "/usr/local/lib/python3.10/dist-packages/auto_optimizer/graph_refactor/onnx/graph.py", line 402, in add_node
return self._add_node(node)
File "/usr/local/lib/python3.10/dist-packages/auto_optimizer/graph_refactor/interface/base_graph.py", line 569, in _add_node
raise ValueError("node name '{}' already exists!".format(node.name))
ValueError: node name 'Shape_q' already exists!
