- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Can I change movidius model graph's input data from float16 to float32, how to do it?
Thanks.
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@miily At the moment, the current NCSDK (1.12) requires the input image to be float16. So when you call LoadTensor() it expects a float16 tensor. You may get unexpected results or errors using float32.
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page