Description
我先用命令：
paddle2onnx --model_dir ./ --model_filename model.pdmodel --params_filename model.pdiparams --save_file ./model.onnx --enable_onnx_checker True
转换paddle模型：https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_l_36e_pipeline.zip 到onnx，转换成功。
然后用命令：
x2paddle --framework=onnx --model=model.onnx --save_dir=pd_model
转换onnx模型到paddle模型，转换失败。
paddle2onnx的输出：
paddle2onnx --model_dir ./
--model_filename model.pdmodel
--params_filename model.pdiparams
--save_file ./model.onnx
--enable_onnx_checker True
[Paddle2ONNX] Start to parse PaddlePaddle model...
[Paddle2ONNX] Model file path: ./model.pdmodel
[Paddle2ONNX] Parameters file path: ./model.pdiparams
[Paddle2ONNX] Start to parsing Paddle model...
[Paddle2ONNX] [reduce_mean: mean_0.tmp_0] Requires the minimal opset version of 11.
[Paddle2ONNX] [reduce_mean: mean_1.tmp_0] Requires the minimal opset version of 11.
[Paddle2ONNX] [reduce_mean: mean_2.tmp_0] Requires the minimal opset version of 11.
[Paddle2ONNX] [reduce_mean: mean_3.tmp_0] Requires the minimal opset version of 11.
[Paddle2ONNX] [multiclass_nms3: multiclass_nms3_0.tmp_1] Requires the minimal opset version of 10.
[Paddle2ONNX] Due to the operator: reduce_mean, requires opset_version >= 11.
[Paddle2ONNX] Opset version will change to 11 from 9
[Paddle2ONNX] Use opset_version = 11 for ONNX export.
[Paddle2ONNX] PaddlePaddle model is exported as ONNX format now.
x2paddle的输出：
x2paddle --framework=onnx --model=model.onnx --save_dir=pd_model
model ir_version: 6, op version: 11
Shape inferencing ...
[WARNING] Incomplete symbolic shape inference
Shape inferenced.
Traceback (most recent call last):
File "/opt/anaconda3/envs/paddle/bin/x2paddle", line 8, in
sys.exit(main())
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/x2paddle/convert.py", line 501, in main
onnx2paddle(args.model,
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/x2paddle/convert.py", line 315, in onnx2paddle
model = ONNXDecoder(model_path, enable_onnx_checker, input_shape_dict)
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/x2paddle/decoder/onnx_decoder.py", line 415, in init
self.graph = ONNXGraph(onnx_model, input_shape_dict)
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/x2paddle/decoder/onnx_decoder.py", line 201, in init
self.collect_value_infos()
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/x2paddle/decoder/onnx_decoder.py", line 374, in collect_value_infos
TENSOR_TYPE_TO_NP_TYPE[item.type.tensor_type.elem_type],
File "/opt/anaconda3/envs/paddle/lib/python3.10/site-packages/onnx/mapping.py", line 132, in getitem
return super().getitem(key)
KeyError: 0
python包版本信息：
Package Version
anyio 4.6.2.post1
astor 0.8.1
certifi 2024.8.30
charset-normalizer 3.4.0
decorator 5.1.1
exceptiongroup 1.2.2
h11 0.14.0
httpcore 1.0.6
httpx 0.27.2
idna 3.10
mpmath 1.3.0
numpy 2.1.2
onnx 1.17.0
opt-einsum 3.3.0
packaging 24.1
paddle2onnx 1.2.11
paddlepaddle 2.6.2
pillow 11.0.0
pip 24.2
protobuf 5.28.2
requests 2.32.3
setuptools 75.2.0
six 1.16.0
sniffio 1.3.1
sympy 1.13.3
typing_extensions 4.12.2
urllib3 2.2.3
wheel 0.44.0
x2paddle 1.5.0
系统版本：
os ubuntu 22.04 Linux version 6.8.0-47-generic
email: [email protected]