- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi MCC,
You are receiving a lot of "Shape and content size of tensor don't match" warning.
You may need to specify the input size as it seems that the propagated size is not correct.
Regards,
Rizal
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Rizal,
I provided "--input" with "--input_shape" and "-b 1" for mo.py, and both way got the same error.
Regards,
MCC
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Rizal,
I fixed it. Just need to add "--static_shape".
Thanks,
MCC
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi MCC,
Congratulation on being able to fix it.
Can you share what is different compared to your other custom models for other readers that may encounter the same issue as you?
Is there anything else you would like to ask?
Regards,
Rizal
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Rizal,
I guess the issue is kernel size between layers. For some reasons, this model did not use resnet like architecture.
Regards,
MCC
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi MCC,
Thank you for the information.
Intel will no longer monitor this thread. If you need any additional information from Intel, please submit a new question.
Regards,
Rizal
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page