mindspore-ai / community Goto Github PK

Dear valued developers,

MindSpore is our versatile deep learning framework, and we're constantly looking at new ways to make AI modeling and execution more efficient, and as well as cover more scenarios. With 2023 just around the corner, the MindSpore community would like to wish you a happy holiday – we hope you'll give us some feedback, so that we can create a community that best serves your needs.

We'd appreciate it if you could take a few minutes to answer the three questions below – please write your comments on this page. Thank you for your support!

1. What changes would you like to see to MindSpore in 2023? Feel free to list a few different suggestions.

2. Are you interested in MindSpore meetups, courses, tutorials, and conferences? If so, which interests you the most? What do you hope to gain from these activities?

3. Would you consider using MindSpore in your Google Summer of Code (GSoC) projects in 2023? If yes, please give us more details.

The MindSpore community highly values your feedback. We have prepared special Christmas prizes for a few lucky winners. Leave a comment below for the chance to be picked; all winners will be contacted via GitHub. Merry Christmas and a happy new year!

For more info, follow our social media accounts.
Twitter: https://twitter.com/MindSporeOSS
LinkedIn: https://www.linkedin.com/company/80968760/
YouTube: https://www.youtube.com/@MindsporeOSS
Reddit: https://new.reddit.com/r/MindSporeOSS/

I am using VOC dataset to train Fast-RCNN model using Mindspore, refering to the offciel link:
both_Errors.docx

https://gitee.com/mindspore/mindspore/tree/r1.3/model_zoo/official/cv/faster_rcnn

When I try to run the train.py file using Mindspore 1.3 Iget the follwing error:

RuntimeError: mindspore\ccsrc\backend\kernel_compiler\cpu\cpu_kernel_factory.cc:157 GetSupportedKernelAttrList] Not registered CPU kernel: op[CheckValid]!
# 
WARNING: Logging before InitGoogleLogging() is written to STDERR
[ERROR] KERNEL(6860,41095,?):2021-9-28 14:20:41 [mindspore\ccsrc\backend\kernel_compiler\cpu\cpu_kernel_factory.cc:157] GetSupportedKernelAttrList] Not registered CPU kernel: op[CheckValid]!

I got an Advice from the community to use Mindspore 1.5, and I did but this time I got a very complicated error which I think because I am using FastRcnn for the version 1.3 and I could not find the model in version 1.5.

Following is a snapshot of the error I got when I use Mindspore 1.5:

[WARNING] OPTIMIZER(20896,1,?):2021-9-28 14:43:54 [mindspore\ccsrc\frontend\optimizer\ad\kprim.cc:287] GetBprop] Fail to find bprop function for BoundingBoxEncode. fn: None

TypeError: mindspore\ccsrc\runtime\device\cpu\kernel_select_cpu.cc:219 KernelNotSupportException] Operator[StridedSlice]  input(kNumberTypeUInt8) output(kNumberTypeUInt8) is not support. This error means the current input type is not supported, please refer to the MindSpore doc for supported types.

[WARNING] PIPELINE(20896,1,?):2021-9-28 14:44:6 [mindspore\ccsrc\pipeline\jit\init.cc:310] operator()] Start releasing dataset handles...
[WARNING] PIPELINE(20896,1,?):2021-9-28 14:44:6 [mindspore\ccsrc\pipeline\jit\init.cc:313] operator()] End release dataset handles.

knowing that my os is windows 10, and I am using CPU as soon as it's the only device compatible with mindpore when it comes to windows. I attached the full traceback for both errors as doc files.

please provide me with solution and let me know the reason behins this errors, is it a hardware problem?

Dear valued developers,

MindSpore is our versatile deep learning framework, and we're constantly looking at new ways to make AI modeling and execution more efficient, and as well as cover more scenarios. With 2023 just around the corner, the MindSpore community would like to wish you a happy holiday – we hope you'll give us some feedback, so that we can create a community that best serves your needs.

We'd appreciate it if you could take a few minutes to answer the three questions below – please write your comments on this page. Thank you for your support!

1. What changes would you like to see to MindSpore in 2023? Feel free to list a few different suggestions.

2. Are you interested in MindSpore meetups, courses, tutorials, and conferences? If so, which interests you the most? What do you hope to gain from these activities?

3. Would you consider using MindSpore in your Google Summer of Code (GSoC) projects in 2023? If yes, please give us more details.

The MindSpore community highly values your feedback. We have prepared special Christmas prizes for a few lucky winners. Leave a comment below for the chance to be picked; all winners will be contacted via GitHub. Merry Christmas and a happy new year!

For more info, follow our social media accounts.
Twitter: https://twitter.com/MindSporeOSS
LinkedIn: https://www.linkedin.com/company/80968760/
YouTube: https://www.youtube.com/@MindsporeOSS
Reddit: https://new.reddit.com/r/MindSporeOSS/