K230模型转换

可以使用YOLO的脚本进行转换，参考这个文档 https://www.kendryte.com/k230_canmv/zh/main/zh/example/ai/YOLO%E5%A4%A7%E4%BD%9C%E6%88%98.html#

(base) C:\Users\lenovo>cd .

(base) C:\Users\lenovo>cd ..

(base) C:\Users>cd ..

(base) C:>D:

(base) D:>cd D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>pip install nncase_kpu-2.8.3-py2.py3-none-win_amd64.whl
Processing .\nncase_kpu-2.8.3-py2.py3-none-win_amd64.whl
Requirement already satisfied: nncase>=2.8.3 in D:\Users\miniconda3\Lib\site-packages (from nncase-kpu==2.8.3) (2.11.0)
Requirement already satisfied: numpy in D:\Users\miniconda3\Lib\site-packages (from nncase>=2.8.3->nncase-kpu==2.8.3) (2.4.4)
Installing collected packages: nncase-kpu
Attempting uninstall: nncase-kpu
Found existing installation: nncase-kpu 2.9.0
Uninstalling nncase-kpu-2.9.0:
Successfully uninstalled nncase-kpu-2.9.0
Successfully installed nncase-kpu-2.8.3

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>pip install onnx
Requirement already satisfied: onnx in D:\Users\miniconda3\Lib\site-packages (1.21.0)
Requirement already satisfied: numpy>=1.23.2 in D:\Users\miniconda3\Lib\site-packages (from onnx) (2.4.4)
Requirement already satisfied: protobuf>=4.25.1 in D:\Users\miniconda3\Lib\site-packages (from onnx) (7.34.1)
Requirement already satisfied: typing_extensions>=4.7.1 in D:\Users\miniconda3\Lib\site-packages (from onnx) (4.15.0)
Requirement already satisfied: ml_dtypes>=0.5.0 in D:\Users\miniconda3\Lib\site-packages (from onnx) (0.5.4)

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>pip install onnxruntime
Requirement already satisfied: onnxruntime in D:\Users\miniconda3\Lib\site-packages (1.24.4)
Requirement already satisfied: flatbuffers in D:\Users\miniconda3\Lib\site-packages (from onnxruntime) (25.12.19)
Requirement already satisfied: numpy>=1.21.6 in D:\Users\miniconda3\Lib\site-packages (from onnxruntime) (2.4.4)
Requirement already satisfied: packaging in D:\Users\miniconda3\Lib\site-packages (from onnxruntime) (25.0)
Requirement already satisfied: protobuf in D:\Users\miniconda3\Lib\site-packages (from onnxruntime) (7.34.1)
Requirement already satisfied: sympy in D:\Users\miniconda3\Lib\site-packages (from onnxruntime) (1.14.0)
Requirement already satisfied: mpmath<1.4,>=1.1.0 in D:\Users\miniconda3\Lib\site-packages (from sympy->onnxruntime) (1.3.0)

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>pip install onnxsim
Requirement already satisfied: onnxsim in D:\Users\miniconda3\Lib\site-packages (0.6.2)
Requirement already satisfied: onnx in D:\Users\miniconda3\Lib\site-packages (from onnxsim) (1.21.0)
Requirement already satisfied: rich in D:\Users\miniconda3\Lib\site-packages (from onnxsim) (14.2.0)
Requirement already satisfied: numpy>=1.23.2 in D:\Users\miniconda3\Lib\site-packages (from onnx->onnxsim) (2.4.4)
Requirement already satisfied: protobuf>=4.25.1 in D:\Users\miniconda3\Lib\site-packages (from onnx->onnxsim) (7.34.1)
Requirement already satisfied: typing_extensions>=4.7.1 in D:\Users\miniconda3\Lib\site-packages (from onnx->onnxsim) (4.15.0)
Requirement already satisfied: ml_dtypes>=0.5.0 in D:\Users\miniconda3\Lib\site-packages (from onnx->onnxsim) (0.5.4)
Requirement already satisfied: markdown-it-py>=2.2.0 in D:\Users\miniconda3\Lib\site-packages (from rich->onnxsim) (4.0.0)
Requirement already satisfied: pygments<3.0.0,>=2.13.0 in D:\Users\miniconda3\Lib\site-packages (from rich->onnxsim) (2.19.2)
Requirement already satisfied: mdurl~=0.1 in D:\Users\miniconda3\Lib\site-packages (from markdown-it-py>=2.2.0->rich->onnxsim) (0.1.2)

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>yolo export model=D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.pt format=onnx imgsz=640
'yolo' 不是内部或外部命令，也不是可运行的程序
或批处理文件。

(base) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>conda activate yolov8_k230_env

(yolov8_k230_env) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>yolo export model=D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.pt format=onnx imgsz=640
Ultralytics 8.4.33 Python-3.8.20 torch-2.4.1+cpu CPU (13th Gen Intel Core i7-13620H)
Model summary (fused): 73 layers, 3,007,598 parameters, 0 gradients, 8.1 GFLOPs

PyTorch: starting from 'D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.pt' with input shape (1, 3, 640, 640) BCHW and output shape(s) (1, 14, 8400) (6.0 MB)

ONNX: starting export with onnx 1.17.0 opset 19...
ONNX: slimming with onnxslim 0.1.90...
ONNX: export success 1.7s, saved as 'D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.onnx' (11.7 MB)

Export complete (2.2s)
Results saved to D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230
Predict: yolo predict task=detect model=D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.onnx imgsz=640
Validate: yolo val task=detect model=D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.onnx imgsz=640 data=D:/moxingxunlian/2-L+R/dataset.yaml
Visualize: https://netron.app
Learn more at https://docs.ultralytics.com/modes/export

(yolov8_k230_env) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>python to_kmodel.py --target k230 --model ../../D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.onnx --dataset ../test --input_width 640 --input_height 480 --ptq_option 0
python: can't open file 'to_kmodel.py': [Errno 2] No such file or directory

(yolov8_k230_env) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230>cd test_yolov8/classify

(yolov8_k230_env) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\test_yolov8\classify>python to_kmodel.py --target k230 --model ../../D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\best.onnx --dataset ../test --input_width 640 --input_height 480 --ptq_option 0
Traceback (most recent call last):
File "to_kmodel.py", line 6, in
import nncase
File "D:\Users\miniconda3\envs\yolov8_k230_env\lib\site-packages\nncase_init_.py", line 34, in
import _nncase
ImportError: DLL load failed while importing _nncase: 找不到指定的模块。

(yolov8_k230_env) D:\moxingxunlian\2-L+R\YOLO_Models_cunfang\K230\test_yolov8\classify>
试了，输入最后一步后就一直提示说少东西，没看懂到底缺什么，AI只是不断让重新下载

可参考这个博主，带有文件下载连接，，
【nncase插件使用问题 - CSDN App】https://blog.csdn.net/Lieber_0402/article/details/132021911?sharetype=blog&shareId=132021911&sharerefer=APP&sharesource=cvge2&sharefrom=link