Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -36,38 +36,38 @@ More details on model performance across various devices, can be found
|
|
36 |
|
37 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
38 |
|---|---|---|---|---|---|---|---|---|
|
39 |
-
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.
|
40 |
-
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 1.
|
41 |
-
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 2.
|
42 |
-
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.
|
43 |
-
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.
|
44 |
-
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX |
|
45 |
-
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.
|
46 |
-
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.
|
47 |
-
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX |
|
48 |
-
| WideResNet50-Quantized | SA7255P ADP | SA7255P | TFLITE | 22.
|
49 |
-
| WideResNet50-Quantized | SA7255P ADP | SA7255P | QNN | 23.
|
50 |
-
| WideResNet50-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.
|
51 |
-
| WideResNet50-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 1.
|
52 |
-
| WideResNet50-Quantized | SA8295P ADP | SA8295P | TFLITE | 2.
|
53 |
-
| WideResNet50-Quantized | SA8295P ADP | SA8295P | QNN |
|
54 |
-
| WideResNet50-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.
|
55 |
-
| WideResNet50-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 1.
|
56 |
-
| WideResNet50-Quantized | SA8775P ADP | SA8775P | TFLITE | 2.
|
57 |
-
| WideResNet50-Quantized | SA8775P ADP | SA8775P | QNN | 2.
|
58 |
-
| WideResNet50-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 7.
|
59 |
-
| WideResNet50-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 9.
|
60 |
-
| WideResNet50-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 24.
|
61 |
-
| WideResNet50-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 22.
|
62 |
-
| WideResNet50-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 23.
|
63 |
-
| WideResNet50-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.
|
64 |
-
| WideResNet50-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 1.
|
65 |
-
| WideResNet50-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 2.
|
66 |
-
| WideResNet50-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 2.
|
67 |
-
| WideResNet50-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.
|
68 |
-
| WideResNet50-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 2.
|
69 |
-
| WideResNet50-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.
|
70 |
-
| WideResNet50-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.
|
71 |
|
72 |
|
73 |
|
@@ -131,8 +131,8 @@ Profiling Results
|
|
131 |
WideResNet50-Quantized
|
132 |
Device : Samsung Galaxy S23 (13)
|
133 |
Runtime : TFLITE
|
134 |
-
Estimated inference time (ms) : 1.
|
135 |
-
Estimated peak memory usage (MB): [0,
|
136 |
Total # Ops : 82
|
137 |
Compute Unit(s) : NPU (82 ops)
|
138 |
```
|
|
|
36 |
|
37 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
38 |
|---|---|---|---|---|---|---|---|---|
|
39 |
+
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.748 ms | 0 - 347 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
40 |
+
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 1.892 ms | 0 - 2 MB | INT8 | NPU | [WideResNet50-Quantized.so](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.so) |
|
41 |
+
| WideResNet50-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 2.923 ms | 0 - 209 MB | INT8 | NPU | [WideResNet50-Quantized.onnx](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.onnx) |
|
42 |
+
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.327 ms | 0 - 104 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
43 |
+
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.455 ms | 0 - 20 MB | INT8 | NPU | [WideResNet50-Quantized.so](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.so) |
|
44 |
+
| WideResNet50-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.13 ms | 0 - 118 MB | INT8 | NPU | [WideResNet50-Quantized.onnx](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.onnx) |
|
45 |
+
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.205 ms | 0 - 43 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
46 |
+
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.222 ms | 0 - 44 MB | INT8 | NPU | Use Export Script |
|
47 |
+
| WideResNet50-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.205 ms | 0 - 59 MB | INT8 | NPU | [WideResNet50-Quantized.onnx](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.onnx) |
|
48 |
+
| WideResNet50-Quantized | SA7255P ADP | SA7255P | TFLITE | 22.219 ms | 0 - 36 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
49 |
+
| WideResNet50-Quantized | SA7255P ADP | SA7255P | QNN | 23.427 ms | 0 - 8 MB | INT8 | NPU | Use Export Script |
|
50 |
+
| WideResNet50-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.745 ms | 0 - 344 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
51 |
+
| WideResNet50-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 1.917 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
|
52 |
+
| WideResNet50-Quantized | SA8295P ADP | SA8295P | TFLITE | 2.721 ms | 0 - 39 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
53 |
+
| WideResNet50-Quantized | SA8295P ADP | SA8295P | QNN | 2.803 ms | 0 - 14 MB | INT8 | NPU | Use Export Script |
|
54 |
+
| WideResNet50-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.745 ms | 0 - 338 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
55 |
+
| WideResNet50-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 1.892 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
|
56 |
+
| WideResNet50-Quantized | SA8775P ADP | SA8775P | TFLITE | 2.052 ms | 0 - 37 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
57 |
+
| WideResNet50-Quantized | SA8775P ADP | SA8775P | QNN | 2.252 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
|
58 |
+
| WideResNet50-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 7.917 ms | 0 - 92 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
59 |
+
| WideResNet50-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 9.607 ms | 0 - 12 MB | INT8 | NPU | Use Export Script |
|
60 |
+
| WideResNet50-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 24.195 ms | 0 - 2 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
61 |
+
| WideResNet50-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 22.219 ms | 0 - 36 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
62 |
+
| WideResNet50-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 23.427 ms | 0 - 8 MB | INT8 | NPU | Use Export Script |
|
63 |
+
| WideResNet50-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.745 ms | 0 - 347 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
64 |
+
| WideResNet50-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 1.874 ms | 0 - 4 MB | INT8 | NPU | Use Export Script |
|
65 |
+
| WideResNet50-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 2.052 ms | 0 - 37 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
66 |
+
| WideResNet50-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 2.252 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
|
67 |
+
| WideResNet50-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.182 ms | 0 - 98 MB | INT8 | NPU | [WideResNet50-Quantized.tflite](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.tflite) |
|
68 |
+
| WideResNet50-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 2.27 ms | 0 - 98 MB | INT8 | NPU | Use Export Script |
|
69 |
+
| WideResNet50-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.797 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
|
70 |
+
| WideResNet50-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.79 ms | 70 - 70 MB | INT8 | NPU | [WideResNet50-Quantized.onnx](https://huggingface.co/qualcomm/WideResNet50-Quantized/blob/main/WideResNet50-Quantized.onnx) |
|
71 |
|
72 |
|
73 |
|
|
|
131 |
WideResNet50-Quantized
|
132 |
Device : Samsung Galaxy S23 (13)
|
133 |
Runtime : TFLITE
|
134 |
+
Estimated inference time (ms) : 1.7
|
135 |
+
Estimated peak memory usage (MB): [0, 347]
|
136 |
Total # Ops : 82
|
137 |
Compute Unit(s) : NPU (82 ops)
|
138 |
```
|