Hugging Face's logo

aplux
/
ResNet-50

ONNX
Model card Files Community
qc903113684 commited on
Commit
ed03456
·
verified ·
1 Parent(s): f2d4cda

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +29 -8
README.md CHANGED
@@ -12,15 +12,36 @@ It introduces the concept of residual learning, protects the integrity of inform
12
 
13
  The model can be found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py)
14
 
 
15
 
16
  |Device|SoC|Runtime|Model|Size (pixels)|Inference Time (ms)|Precision|Compute Unit|Model Download|
17
  |:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|
18
- |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|3.6|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS6490/resnet50_int8.qnn.serialized.bin)|
19
- |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|5.9|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS6490/resnet50_int16.qnn.serialized.bin)|
20
- |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|3.4|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS6490/resnet50_int8_htp_snpe2.dlc)|
21
- |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|4.8|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS6490/resnet50_int16_htp_snpe2.dlc)|
 
 
 
 
22
  |AidBox GS865|QCS8250|SNPE|ResNet-50|224|9|INT8|NPU|[model download]()|
23
- |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|2.5|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS8550/resnet50_int8.qnn.serialized.bin)|
24
- |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|3.5|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS8550/resnet50_int16.qnn.serialized.bin)|
25
- |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|1|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS8550/resnet50_int8_htp_snpe2.dlc)|
26
- |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|1.4|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/QCS8550/resnet50_int16_htp_snpe2.dlc)|
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
 
13
  The model can be found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py)
14
 
15
+ **Performance on devices**
16
 
17
  |Device|SoC|Runtime|Model|Size (pixels)|Inference Time (ms)|Precision|Compute Unit|Model Download|
18
  |:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|
19
+ |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|3.6|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS6490/resnet50_int8.qnn.serialized.bin)|
20
+ |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|5.9|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS6490/resnet50_int16.qnn.serialized.bin)|
21
+ |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|3.4|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS6490/resnet50_int8_htp_snpe2.dlc)|
22
+ |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|4.8|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS6490/resnet50_int16_htp_snpe2.dlc)|
23
+ |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|2.5|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS8550/resnet50_int8.qnn.serialized.bin)|
24
+ |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|3.5|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS8550/resnet50_int16.qnn.serialized.bin)|
25
+ |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|1|INT8|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS8550/resnet50_int8_htp_snpe2.dlc)|
26
+ |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|1.4|INT16|NPU|[model download](https://huggingface.co/aidlux/ResNet-50/blob/main/Models/QCS8550/resnet50_int16_htp_snpe2.dlc)|
27
  |AidBox GS865|QCS8250|SNPE|ResNet-50|224|9|INT8|NPU|[model download]()|
28
+
29
+ **Demo models conversion**
30
+
31
+ Demo models converted from [**AIMO(AI Model Optimizier)**](https://aidlux.com/en/product/aimo).
32
+
33
+ The source model **ResNet-50.onnx** can be found [here](https://huggingface.co/aplux/ResNet-50/blob/main/resnet50.onnx).
34
+
35
+ The demo model conversion step on AIMO can be found blow:
36
+
37
+ |Device|SoC|Runtime|Model|Size (pixels)|Precision|Compute Unit|AIMO Conversion Steps|
38
+ |:----:|:----:|:----:|:----:|:----:|:----:|:----:|:----:|
39
+ |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|INT8|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS6490/aimo_resnet50_qnn_int8.png)|
40
+ |AidBox QCS6490|QCS6490|QNN|ResNet-50|224|INT16|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS6490/aimo_resnet50_qnn_int16.png)|
41
+ |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|INT8|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS6490/aimo_resnet50_snpe_int8.png)|
42
+ |AidBox QCS6490|QCS6490|SNPE|ResNet-50|224|INT16|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS6490/aimo_resnet50_snpe_int16.png)|
43
+ |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|INT8|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS8550/aimo_resnet50_qnn_int8.png)|
44
+ |APLUX QCS8550|QCS8550|QNN|ResNet-50|224|INT16|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS8550/aimo_resnet50_qnn_int16.png)|
45
+ |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|INT8|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS8550/aimo_resnet50_snpe_int8.png)|
46
+ |APLUX QCS8550|QCS8550|SNPE|ResNet-50|224|INT16|NPU|[View Steps](https://huggingface.co/aplux/ResNet-50/blob/main/AIMO/QCS8550/aimo_resnet50_snpe_int16.png)|
47
+ |AidBox GS865|QCS8250|SNPE|ResNet-50|224|INT8|NPU|[View Steps]()|