【STM32N6570-DK评测】五、一个AI新手为N6构建软件环境
本帖最后由 地瓜patch 于 2025-5-11 22:50 编辑一、初级阶段
查看代码，编辑代码用IAR或keil或StmCubeIDE或CubeMX是足够的
下载程序，用StmCubeProgrammer.
在之前的帖子里，用脚本下载多个程序文件，也是在脚本中调用StmCubeProgrammer，在一下帖子中有说明。
【STM32N6570-DK评测】三、初识AI及下载多个程序文件的快捷方法
https://bbs.21ic.com/icview-3452 ... 9C%B0%E7%93%9Cpatch
有了以上几个软件，下载ST官方给出的例程足够了。
二、进阶
除去上述ST官方给出的例程，还给出了一些更其他的例程，用stm32ai-modelzoo训练完成的一些例程
比如stm32ai-modelzoo-main\object_detection\ssd_mobilenet_v2_fpnlite\ST_pretrainedmodel_public_dataset\coco_2017_person\ssd_mobilenet_v2_fpnlite_035_192\ssd_mobilenet_v2_fpnlite_035_192_int8.tflite
比如stm32ai-modelzoo-main\object_detection\ssd_mobilenet_v2_fpnlite\ST_pretrainedmodel_public_dataset\coco_2017_person\ssd_mobilenet_v2_fpnlite_035_192\ssd_mobilenet_v2_fpnlite_035_192_int8.tflite
这些都是modelzoo中用tersorflow训练完成的一些模型，要调用这些模型需要进行编译后下载
学习众多高手的对STM32N6570 AI的测试文章、帖子、讨论等等，自己也在学习尝试编译这些模型。
在学习过程中遇到了重困难。
利用pip install -r requirements.txt命令通过python安装所有工具软件插件。
requirements.txt文件就是一个软件列表，包含使用版本
mlflow==2.3.0
hydra-core==1.3.2
omegaconf==2.3.0
numpy==1.23.4
pandas==1.5.3
wget==3.2
munch==2.5.0
seaborn==0.12.2
scikit-learn==1.2.2
tqdm==4.65.0
skl2onnx==1.14.0
onnx==1.12.0
onnxconverter-common==1.13.0
onnxruntime==1.15.1
optuna==3.1.1
imgaug==0.4.0
matplotlib==3.6.2
opencv_python==4.6.0.66
joblib==1.2.0
librosa==0.10.0.post2
mako==1.2.4
pyserial==3.5
colorama==0.4.5
larq==0.13.3
Requests==2.28.2
urllib3==1.26.13
xmlrunner==1.7.7
marshmallow==3.20.1
tf2onnx==1.14.0
clearml==1.16.5
pip-system-certs==4.0
执行pip install -r requirements.txt命令时遇到了不少错误提示
比如
ModuleNotFoundError: No module named 'tensorflow'；
No matching distribution found for tensorflow==2.8.3；
ModuleNotFoundError: No module named 'mlflow'；
ModuleNotFoundError: No module named 'hydra'；
Running `icl ""` gave " 系统找不到指定的文件。"；
Running `cl /?` gave " 系统找不到指定的文件。"；
Running `cc --version` gave " 系统找不到指定的文件。"；
Running `gcc --version` gave " 系统找不到指定的文件。"；
Running `clang --version` gave " 系统找不到指定的文件。"；
Running `clang-cl /?` gave " 系统找不到指定的文件。"；
Running `pgcc --version` gave " 系统找不到指定的文件。"；
这些错误通过执行pip install ...命令将软件安装后，基本都能解决这些错误。
No matching distribution found for tensorflow==2.8.3；的错误试过很多种方法总是不能解决。
最后通过deepseek找到了一种可能，
于是重新下载python3.8后重新安装，然后再次执行，终于可以下载tensorflow2.8.3了
开始逐个下载 requirements.txt中的软件，插件或工具
安装完成tenserflow以后，又出现新问题
PASS: 0%| | 0/92
PASS:12%|█▏ | 11/92
PASS:13%|█▎ | 12/92
PASS:20%|█▉ | 18/92
PASS:24%|██▍ | 22/92
PASS:28%|██▊ | 26/92
PASS:36%|███▌ | 33/92
PASS:41%|████▏ | 38/92
PASS:46%|████▌ | 42/92
PASS:50%|█████ | 46/92
PASS:54%|█████▍ | 50/92
PASS:59%|█████▊ | 54/92
PASS:64%|██████▍ | 59/92
PASS:68%|██████▊ | 63/92
PASS:73%|███████▎| 67/92
PASS:77%|███████▋| 71/92
PASS:82%|████████▏ | 75/92
PASS:87%|████████▋ | 80/92
PASS:91%|█████████▏| 84/92
>>>> EXECUTING NEURAL ART COMPILER
PASS:91%|█████████▏| 84/92
E:/ProgramFiles/STEdgeAI/2.1/Utilities/windows/atonn.exe -i "G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/mobilenet_v2_0.35_128_fft_int8_OE_3_2_0.onnx" --json-quant-file "G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/mobilenet_v2_0.35_128_fft_int8_OE_3_2_0_Q.json" -g "network.c" --load-mdesc "E:/ProgramFiles/STEdgeAI/2.1/Utilities/configs/stm32n6.mdesc" --load-mpool "G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/application_code/image_classification/STM32N6/Model/my_mpools/stm32n6-app2.mpool" --save-mpool-file "G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/neural_art__network/stm32n6-app2.mpool" --out-dir-prefix "G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/neural_art__network/" --all-buffers-info --no-hw-sw-parallelism --cache-maintenance --enable-virtual-mem-pools --native-float --optimization 3 --Os --Omax-ca-pipe 4 --Ocache-opt --enable-epoch-controller --output-info-file "c_info.json"
PASS:91%|█████████▏| 84/92
<<<< DONE EXECUTING NEURAL ART COMPILER
PASS:91%|█████████▏| 84/92
LOAD ERROR: WindowsPath('G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/network.h') and WindowsPath('G:/STCL562F429N6570/STM32N6570DK/stm32ai-modelzoo-services-main/image_classification/src/experiments_outputs/2025_05_11_21_56_42/network.h') are the same file
Error executing job with overrides: []
Traceback (most recent call last):
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\clearml\binding\hydra_bind.py", line 230, in _patched_task_function
return task_function(a_config, *a_args, **a_kwargs)
File "stm32ai_main.py", line 407, in main
process_mode(mode=mode, configs=cfg, train_ds=train_ds, valid_ds=valid_ds, quantization_ds=quantization_ds,
File "stm32ai_main.py", line 302, in process_mode
deploy(cfg=configs)
File "../deployment\deploy.py", line 90, in deploy
stm32ai_deploy_stm32n6(target=board, stlink_serial_number=stlink_serial_number, stm32ai_version=stm32ai_version, c_project_path=c_project_path,
File "../../common/deployment\common_deploy.py", line 469, in stm32ai_deploy_stm32n6
stmaic_local_call(session)
File "../../common/deployment\common_deploy.py", line 443, in stmaic_local_call
stmaic.compile(session=session, options=opt, target=session._board_config)
File "../../common\stm32ai_local\compile.py", line 208, in cmd_compile
raise Exception('Error during compilation')
Exception: Error during compilation
Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
支持一下。。。。 环境搭建还是必要的 大佬都玩单片机的AI了啊，牛啊。 收一块STM32N6开发板。
页:
[1]