v0.31.0
Browse filesSee https://github.com/quic/ai-hub-models/releases/v0.31.0 for changelog.
- .gitattributes +2 -0
- README.md +49 -35
- Swin-Base.dlc +3 -0
- Swin-Base_w8a16.dlc +3 -0
.gitattributes
CHANGED
@@ -34,3 +34,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
Swin-Base.so filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
Swin-Base.so filter=lfs diff=lfs merge=lfs -text
|
37 |
+
Swin-Base.dlc filter=lfs diff=lfs merge=lfs -text
|
38 |
+
Swin-Base_w8a16.dlc filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
@@ -35,39 +35,53 @@ More details on model performance across various devices, can be found
|
|
35 |
|
36 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
-
| Swin-Base | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE |
|
39 |
-
| Swin-Base | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) |
|
40 |
-
| Swin-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 28.
|
41 |
-
| Swin-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) |
|
42 |
-
| Swin-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 22.
|
43 |
-
| Swin-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) |
|
44 |
-
| Swin-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 25.
|
45 |
-
| Swin-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) |
|
46 |
-
| Swin-Base | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE |
|
47 |
-
| Swin-Base | float | SA7255P ADP | Qualcomm® SA7255P |
|
48 |
-
| Swin-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 22.
|
49 |
-
| Swin-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) |
|
50 |
-
| Swin-Base | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE |
|
51 |
-
| Swin-Base | float | SA8295P ADP | Qualcomm® SA8295P |
|
52 |
-
| Swin-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 22.
|
53 |
-
| Swin-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) |
|
54 |
-
| Swin-Base | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 25.
|
55 |
-
| Swin-Base | float | SA8775P ADP | Qualcomm® SA8775P |
|
56 |
-
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 22.
|
57 |
-
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile |
|
58 |
-
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 19.
|
59 |
-
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 15.
|
60 |
-
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile |
|
61 |
-
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 13.
|
62 |
-
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 14.
|
63 |
-
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile |
|
64 |
-
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 10.
|
65 |
-
| Swin-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite |
|
66 |
-
| Swin-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 22.
|
67 |
-
| Swin-Base | w8a16 |
|
68 |
-
| Swin-Base | w8a16 |
|
69 |
-
| Swin-Base | w8a16 |
|
70 |
-
| Swin-Base | w8a16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
71 |
|
72 |
|
73 |
|
@@ -131,8 +145,8 @@ Profiling Results
|
|
131 |
Swin-Base
|
132 |
Device : cs_8275 (ANDROID 14)
|
133 |
Runtime : TFLITE
|
134 |
-
Estimated inference time (ms) :
|
135 |
-
Estimated peak memory usage (MB): [
|
136 |
Total # Ops : 1568
|
137 |
Compute Unit(s) : npu (1568 ops) gpu (0 ops) cpu (0 ops)
|
138 |
```
|
|
|
35 |
|
36 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
+
| Swin-Base | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 61.812 ms | 116 - 520 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
39 |
+
| Swin-Base | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 55.545 ms | 24 - 381 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
40 |
+
| Swin-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 28.857 ms | 0 - 395 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
41 |
+
| Swin-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 28.781 ms | 1 - 379 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
42 |
+
| Swin-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 22.217 ms | 0 - 38 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
43 |
+
| Swin-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 19.954 ms | 0 - 70 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
44 |
+
| Swin-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 25.745 ms | 0 - 404 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
45 |
+
| Swin-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 22.962 ms | 0 - 357 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
46 |
+
| Swin-Base | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 61.812 ms | 116 - 520 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
47 |
+
| Swin-Base | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 55.545 ms | 24 - 381 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
48 |
+
| Swin-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 22.1 ms | 0 - 41 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
49 |
+
| Swin-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 19.98 ms | 0 - 71 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
50 |
+
| Swin-Base | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 32.003 ms | 0 - 392 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
51 |
+
| Swin-Base | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 28.705 ms | 1 - 349 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
52 |
+
| Swin-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 22.133 ms | 0 - 36 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
53 |
+
| Swin-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 20.04 ms | 0 - 66 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
54 |
+
| Swin-Base | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 25.745 ms | 0 - 404 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
55 |
+
| Swin-Base | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 22.962 ms | 0 - 357 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
56 |
+
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 22.293 ms | 0 - 33 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
57 |
+
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 20.111 ms | 0 - 59 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
58 |
+
| Swin-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 19.643 ms | 0 - 428 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.onnx) |
|
59 |
+
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 15.358 ms | 0 - 403 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
60 |
+
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 13.609 ms | 1 - 388 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
61 |
+
| Swin-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 13.264 ms | 1 - 393 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.onnx) |
|
62 |
+
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 14.538 ms | 0 - 394 MB | NPU | [Swin-Base.tflite](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.tflite) |
|
63 |
+
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 10.108 ms | 1 - 357 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
64 |
+
| Swin-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 10.151 ms | 1 - 355 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.onnx) |
|
65 |
+
| Swin-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 23.11 ms | 1155 - 1155 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.dlc) |
|
66 |
+
| Swin-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 22.329 ms | 175 - 175 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base.onnx) |
|
67 |
+
| Swin-Base | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 38.573 ms | 0 - 284 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
68 |
+
| Swin-Base | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 27.011 ms | 0 - 344 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
69 |
+
| Swin-Base | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 20.663 ms | 0 - 58 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
70 |
+
| Swin-Base | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 21.136 ms | 0 - 287 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
71 |
+
| Swin-Base | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 85.9 ms | 0 - 1212 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
72 |
+
| Swin-Base | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 38.573 ms | 0 - 284 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
73 |
+
| Swin-Base | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 20.769 ms | 0 - 69 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
74 |
+
| Swin-Base | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 25.612 ms | 0 - 298 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
75 |
+
| Swin-Base | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 20.822 ms | 0 - 55 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
76 |
+
| Swin-Base | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 21.136 ms | 0 - 287 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
77 |
+
| Swin-Base | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 20.818 ms | 0 - 51 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
78 |
+
| Swin-Base | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 146.316 ms | 539 - 823 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.onnx) |
|
79 |
+
| Swin-Base | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 13.998 ms | 0 - 288 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
80 |
+
| Swin-Base | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 135.409 ms | 647 - 965 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.onnx) |
|
81 |
+
| Swin-Base | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 13.039 ms | 0 - 281 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
82 |
+
| Swin-Base | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 106.076 ms | 623 - 949 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.onnx) |
|
83 |
+
| Swin-Base | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 24.823 ms | 325 - 325 MB | NPU | [Swin-Base.dlc](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.dlc) |
|
84 |
+
| Swin-Base | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 157.748 ms | 921 - 921 MB | NPU | [Swin-Base.onnx](https://huggingface.co/qualcomm/Swin-Base/blob/main/Swin-Base_w8a16.onnx) |
|
85 |
|
86 |
|
87 |
|
|
|
145 |
Swin-Base
|
146 |
Device : cs_8275 (ANDROID 14)
|
147 |
Runtime : TFLITE
|
148 |
+
Estimated inference time (ms) : 61.8
|
149 |
+
Estimated peak memory usage (MB): [116, 520]
|
150 |
Total # Ops : 1568
|
151 |
Compute Unit(s) : npu (1568 ops) gpu (0 ops) cpu (0 ops)
|
152 |
```
|
Swin-Base.dlc
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:154508635450a3ba8769929a44b823539bcb124d7327ef776b2b5610bc181b86
|
3 |
+
size 356143203
|
Swin-Base_w8a16.dlc
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:13977e8298c09afe641b9cf1969210889b648b2c29f6f9010860da97b97997d1
|
3 |
+
size 94603094
|