Toybrick

标题: 模型转换load onnx报错 [打印本页]

作者: jasonlyo    时间: 2025-3-19 17:24
标题: 模型转换load onnx报错
onnx模型转rknn,报错如下,请问是模型有啥问题吗?RKNN-toolkit版本是1.7.1。详细的日志,可以看附件。模型由于大小限制暂时传不上来。

--> Loading model
W Call onnx.optimizer.optimize fail, skip optimize
/mnt/python36new/lib/python3.6/site-packages/rknn/api/rknn.py:150: RuntimeWarning: overflow encountered in exp
  ret = self.rknn_base.load_onnx(model, inputs, input_size_list, outputs)
E Calc node Slice : /model.22/Slice output shape fail
E Catch exception when loading onnx model: yolov8n_opset11_imgsz2560_fp32_simplifyFalse_dynamicFalse_optimizeFalse.onnx!
E Traceback (most recent call last):
E   File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/shape_inference.py", line 65, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.shape_inference.infer_shape
E   File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/ops/slice.py", line 41, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.ops.slice.Slice
E   File "<string>", line 1, in <module>
E TypeError: slice indices must be integers or None or have an __index__ method
E During handling of the above exception, another exception occurred:
E Traceback (most recent call last):
E   File "rknn/base/RKNNlib/app/importer/import_onnx.py", line 118, in rknn.base.RKNNlib.app.importer.import_onnx.Importonnx.run
E   File "rknn/base/RKNNlib/converter/convert_onnx.py", line 112, in rknn.base.RKNNlib.converter.convert_onnx.convert_onnx.__init__
E   File "rknn/base/RKNNlib/converter/convert_onnx.py", line 1062, in rknn.base.RKNNlib.converter.convert_onnx.convert_onnx._shape_inference
E   File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/shape_inference.py", line 70, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.shape_inference.infer_shape
E   File "rknn/api/rknn_log.py", line 323, in rknn.api.rknn_log.RKNNLog.e
E ValueError: Calc node Slice : /model.22/Slice output shape fail
E Please feedback the detailed log file <log_feedback_to_the_rknn_toolkit_dev_team.log> to the RKNN Toolkit development team.
E You can also check github issues: https://github.com/rockchip-linux/rknn-toolkit/issues




作者: andornot    时间: 2025-3-20 15:28
yolov8官方模型转rknn时,output部分有算子不支持,
你看下rknn_model_zoo中yolov8的example,有对output作修改优化,
文档地址: https://github.com/airockchip/rk ... ain/examples/yolov8
作者: andornot    时间: 2025-3-20 15:31
yolov8官方模型,output部分有算子不支持,
你看下rknn_model_zoo中有 yolov8的example,有对output做了修改优化,
文档地址: https://github.com/airockchip/rk ... ain/examples/yolov8




欢迎光临 Toybrick (https://t.rock-chips.com/) Powered by Discuz! X3.3