J5 Open Explorer hb_mapper checker报错

已解决

Darth Nihilus2026-05-14

报错如下：sudo docker run -it --rm -v /open_explorer -v ./dataset:/data/horizon_j5/data openexplorer/ai_toolchain_ubuntu_20_j5_cpu:v1.1.77-py38

[sudo] password for openexplorer:

root@72665e89a381:/open_explorer# cd ..

root@72665e89a381:/# hb_mapper checker --model-type onnx --model /data/horizon_j5/data/yolov5n.onnx --march bayes

/usr/local/lib/python3.8/dist-packages/paramiko/pkey.py:100: CryptographyDeprecationWarning: TripleDES has been moved to cryptography.hazmat.decrepit.ciphers.algorithms.TripleDES and will be removed from this module in 48.0.0.

"cipher": algorithms.TripleDES,

/usr/local/lib/python3.8/dist-packages/paramiko/transport.py:259: CryptographyDeprecationWarning: TripleDES has been moved to cryptography.hazmat.decrepit.ciphers.algorithms.TripleDES and will be removed from this module in 48.0.0.

"class": algorithms.TripleDES,

2026-05-14 14:42:01,498 INFO log will be stored in //hb_mapper_checker.log

2026-05-14 14:42:01,498 INFO Start hb_mapper....

2026-05-14 14:42:01,498 INFO hbdk version 3.49.13

2026-05-14 14:42:01,498 INFO horizon_nn version 1.0.9

2026-05-14 14:42:01,498 INFO hb_mapper version 1.24.1

2026-05-14 14:42:01,518 INFO Model type: onnx

2026-05-14 14:42:01,519 INFO input names []

2026-05-14 14:42:01,519 INFO input shapes {}

2026-05-14 14:42:01,519 INFO Begin model checking....

2026-05-14 14:42:01,530 INFO Start to Horizon NN Model Convert.

2026-05-14 14:42:01,561 INFO Loading horizon_nn debug methods:[]

2026-05-14 14:42:01,562 INFO The specified model compilation architecture: bayes.

2026-05-14 14:42:01,563 INFO The specified model compilation optimization parameters: [].

2026-05-14 14:42:01,563 INFO Start to prepare the onnx model.

2026-05-14 14:42:01,598 INFO Input ONNX Model Information:

ONNX IR version: 8

Opset version: ['ai.onnx v17', 'horizon v1']

Producer: pytorch v1.14.0

Domain: None

Model version: None

Graph input:

images: shape=[1, 3, 640, 640], dtype=FLOAT16

Graph output:

output0: shape=[1, 25200, 85], dtype=FLOAT16

2026-05-14 14:42:01,601 ERROR *** ERROR-OCCUR-DURING {horizon_nn.build_onnx} ***, error message: The opset version of the model is 17, the maximum supported version is 11.

2026-05-14 14:42:01,601 INFO End to prepare the onnx model.

2026-05-14 14:42:01,601 INFO End to Horizon NN Model Convert.

YOLOv5模型下载如下：https://github.com/ultralytics/yolov5/releases/download/v7.0/yolov5n.onnx

算法工具链

社区活动征程5

0/600

YCJ
Lv.4
问题出在“2026-05-14 14:42:01,601 ERROR *** ERROR-OCCUR-DURING {horizon_nn.build_onnx} ***, error message: The opset version of the model is 17, the maximum supported version is 11.”，J5 量化编译只支持opset版本为11及以下，建议你导出onnx时候加上opset版本参数为11
2026-05-14
0
10
- Darth Nihilus回复YCJ:
  我是直接下载的官方onnx文件。目前我没有找到支持11版本的yolo的onnx。机器性能可能不大支撑我自己导出。我问了咱们天工平台，给出的修改opset的方法好像都不大行。
  2026-05-14
  0
  回复
- YCJ回复Darth Nihilus:
  需要的话我可以转一个给您
  2026-05-14
  0
  回复
- Darth Nihilus回复YCJ:
  您转我一个吧~感谢感谢
  2026-05-14
  0
  回复
- YCJ回复Darth Nihilus:
  通过网盘分享的文件：yolov5n.onnx
  链接: https://pan.baidu.com/s/1nhtf_pmOY5PAC6z9ZvJiXw?pwd=mgma 提取码: mgma
  
  我用yolov5n.pt转的（v7.0）
  2026-05-14
  0
  回复
- Darth Nihilus回复YCJ:
  好滴好滴，感谢老师~另外，我做模型转化的时候运行与报错：
  2026-05-15 09:49:15,806 ERROR Parse 'calibration_parameters' failed!
  2026-05-15 09:49:15,806 ERROR yaml file parse failed. Please double check your config file inputs
  2026-05-15 09:49:15,808 ERROR Wrong cal_dir num given, cal_dir num 0 is not equal to input num 1这个是为什么呢？想请教下~感谢啦
  2026-05-15
  0
  回复
- Darth Nihilus回复YCJ:
  好滴好滴，感谢老师~另外，我做模型转化的时候运行与报错：
  2026-05-15 09:49:15,806 ERROR Parse 'calibration_parameters' failed!
  2026-05-15 09:49:15,806 ERROR yaml file parse failed. Please double check your config file inputs
  2026-05-15 09:49:15,808 ERROR Wrong cal_dir num given, cal_dir num 0 is not equal to input num 1这个是为什么呢？想请教下~感谢啦
  2026-05-15
  0
  回复
- Darth Nihilus回复YCJ:
  好滴好滴，感谢老师~另外，我做模型转化的时候运行与报错：
  2026-05-15 09:49:15,806 ERROR Parse 'calibration_parameters' failed!
  2026-05-15 09:49:15,806 ERROR yaml file parse failed. Please double check your config file inputs
  2026-05-15 09:49:15,808 ERROR Wrong cal_dir num given, cal_dir num 0 is not equal to input num 1这个是为什么呢？想请教下~感谢啦
  2026-05-15
  0
  回复
- Darth Nihilus回复YCJ:
  yaml文件如下：
  model_parameters:
  onnx_model: '/data/horizon_j5/data/yolov5n.onnx'
  march: "bayes"
  layer_out_dump: False
  working_dir: '../model_output'
  output_model_file_prefix: 'yolov5n-det'
  remove_node_type: "Dequantize"
  node_info: {"/model.10/m/m.0/attn/Softmax": {'ON': 'BPU','InputType': 'int8','OutputType': 'int8'}}
  input_parameters:
  input_name: 'images'
  input_shape: '1x3x640x640'
  input_type_rt: 'nv12'
  input_type_train: 'rgb'
  input_layout_train: 'NCHW'
  mean_value: '0.0 0.0 0.0'
  scale_value: '0.00392157 0.00392157 0.00392157'
  calibration_parameters:
  cal_dir: '/data/horizon_j5/data/calib_data'
  compiler_parameters:
  compile_mode: 'latency'
  core_num: 1
  optimize_level: 'O2'
  jobs: 8
  2026-05-15
  0
  回复
- YCJ回复Darth Nihilus:
  cal_data_dir参数名写错了，工具链没识别到校准集
  2026-05-15
  0
  回复
- Darth Nihilus回复YCJ:
  解决啦~感谢啦~
  2026-05-20
  1
  回复