[BUG] PaddleOCR OnnxSlim failure track

[inisis]

Steps to reproduce

git clone https://github.com/PaddlePaddle/PaddleOCR
git checkout e621d034b584fae03c22cd51c26f6a52d62417b6
cd PaddleOCR

mkdir -p inference
wget https://paddleocr.bj.bcebos.com/openatom/ch_ppstructure_openatom_SLANetv2_infer.tar -P inference
tar xvf inference/ch_ppstructure_openatom_SLANetv2_infer.tar -C inference

cd ppstructure/
python table/predict_structure.py \
    --table_model_dir=../inference/ch_ppstructure_openatom_SLANetv2_infer \
    --table_char_dict_path=../ppocr/utils/dict/table_structure_dict.txt \
    --image_dir=docs/table/table.jpg \
    --output=../output/table_slanet_lcnetv2 \
    --use_gpu=False \
    --table_max_len=512

cd ../
paddle2onnx --model_dir inference/ch_ppstructure_openatom_SLANetv2_infer \
--model_filename inference.pdmodel \
--params_filename inference.pdiparams \
--save_file ./inference/ch_ppstructure_openatom_SLANetv2_infer/model.onnx \
--enable_onnx_checker True

onnxslim inference/ch_ppstructure_openatom_SLANetv2_infer/model.onnx inference/ch_ppstructure_openatom_SLANetv2_infer/model.onnx

cd ppstructure/
python table/predict_structure.py \
    --table_model_dir=../inference/ch_ppstructure_openatom_SLANetv2_infer/model.onnx \
    --table_char_dict_path=../ppocr/utils/dict/table_structure_dict.txt \
    --image_dir=docs/table/table.jpg \
    --output=../output/table_slanet_lcnetv2 \
    --use_gpu=False \
    --use_onnx=True \
    --table_max_len=512

[inisis]

(paddle) (base) root@ubuntu20:~/OnnxSlim# python
Python 3.10.15 (main, Oct  3 2024, 07:27:34) [GCC 11.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import onnxruntime as rt
>>> sess = rt.InferenceSession('slim.onnx')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/root/miniconda3/envs/paddle/lib/python3.10/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 419, in __init__
    self._create_inference_session(providers, provider_options, disabled_optimizers)
  File "/root/miniconda3/envs/paddle/lib/python3.10/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 480, in _create_inference_session
    sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model)
onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Load model from slim.onnx failed:Node (Loop.0) Op (Loop) [TypeInferenceError] Graph attribute inferencing failed: Size mismatch validating subgraph inputs. Got 29 inputs but subgraph has 25 inputs and requires 25 inputs. Either provide all subgraph inputs, or just the required inputs.

[inisis]

When a model contains subgraph, it's usually a good practice to not remove unused inputs, because for subgraph, unused inputs works like Identity operator to pass information down

[inisis]

fixed f6c77ac