Bureaucrats, dave_user, Administrators
4,650
edits
Changes
→Results
== Results ==
{| class="wikitable"
|+
Inference times
!Model
!Threads parameter
[ms]
|-
| rowspan="3" |'''Floating-point'''
|unspecified
|220
|-
|'''Half-quantized'''1
|
|-
|2
|
|-
| rowspan="2" |'''Half-quantized'''
|unspecified
|330
|-
|
|
|-
| rowspan="2" |'''Fully-quantized'''
The total prediction time takes into account the time needed to fill the input tensor with the image and the average inference time over three predictions.
The same tests were repeated using an ext4 a network file system stored on a microSD card(NFS) over an Ethernet connection, too. No significant variations in the prediction times were observed.
