Upload Date | May 09 2024 08:12 AM |
Views | 1 |
System Information | |
---|---|
Operating System | Android 14 |
Model | brcm Raspberry Pi 5 |
Model ID | brcm Raspberry Pi 5 |
Motherboard | rpi5 |
Governor | ondemand |
CPU Information | |
---|---|
Name | ARM ARMv8 |
Topology | 1 Processor, 4 Cores |
Identifier | ARM implementer 65 architecture 8 variant 4 part 3339 revision 1 |
Base Frequency | 2.40 GHz |
Cluster 1 | 4 Cores @ 2.40 GHz |
Memory Information | |
---|---|
Size | 7.75 GB |
Inference Information | |
---|---|
Framework | TensorFlow Lite |
Backend | GPU |
Device | Default |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (F32)
|
100% |
40
7.43 IPS |
|
Image Classification (F16)
|
100% |
48
8.98 IPS |
|
Image Classification (I8)
|
97% |
47
8.74 IPS |
|
Image Segmentation (F32)
|
100% |
54
0.90 IPS |
|
Image Segmentation (F16)
|
100% |
67
1.11 IPS |
|
Image Segmentation (I8)
|
98% |
65
1.09 IPS |
|
Pose Estimation (F32)
|
100% |
75
0.09 IPS |
|
Pose Estimation (F16)
|
100% |
92
0.11 IPS |
|
Pose Estimation (I8)
|
100% |
92
0.11 IPS |
|
Object Detection (F32)
|
100% |
34
2.51 IPS |
|
Object Detection (F16)
|
100% |
40
2.99 IPS |
|
Object Detection (I8)
|
65% |
39
2.91 IPS |
|
Face Detection (F32)
|
100% |
102
1.21 IPS |
|
Face Detection (F16)
|
98% |
118
1.40 IPS |
|
Face Detection (I8)
|
86% |
112
1.34 IPS |
|
Depth Estimation (F32)
|
100% |
91
0.71 IPS |
|
Depth Estimation (F16)
|
100% |
105
0.82 IPS |
|
Depth Estimation (I8)
|
94% |
104
0.81 IPS |
|
Style Transfer (F32)
|
100% |
223
0.29 IPS |
|
Style Transfer (F16)
|
100% |
268
0.35 IPS |
|
Style Transfer (I8)
|
98% |
268
0.35 IPS |
|
Image Super-Resolution (F32)
|
100% |
54
1.91 IPS |
|
Image Super-Resolution (F16)
|
100% |
60
2.13 IPS |
|
Image Super-Resolution (I8)
|
98% |
59
2.11 IPS |
|
Text Classification (F32)
|
100% |
139
199.3 IPS |
|
Text Classification (F16)
|
100% |
142
203.9 IPS |
|
Text Classification (I8)
|
92% |
267
383.4 IPS |
|
Machine Translation (F32)
|
100% |
202
3.72 IPS |
|
Machine Translation (F16)
|
100% |
196
3.61 IPS |
|
Machine Translation (I8)
|
64% |
210
3.86 IPS |