Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -35,45 +35,44 @@ More details on model performance across various devices, can be found
|
|
35 |
|
36 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
-
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 7.
|
39 |
-
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.
|
40 |
-
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 9.
|
41 |
-
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.
|
42 |
-
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 6.
|
43 |
-
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 6.
|
44 |
-
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE |
|
45 |
-
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN |
|
46 |
-
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX |
|
47 |
-
| FastSam-S | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 7.
|
48 |
-
| FastSam-S | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 7.
|
49 |
-
| FastSam-S | SA7255P ADP | SA7255P | TFLITE |
|
50 |
-
| FastSam-S | SA7255P ADP | SA7255P | QNN | 257.
|
51 |
-
| FastSam-S | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 7.
|
52 |
-
| FastSam-S | SA8255 (Proxy) | SA8255P Proxy | QNN | 7.
|
53 |
-
| FastSam-S | SA8295P ADP | SA8295P | TFLITE | 15.
|
54 |
-
| FastSam-S | SA8295P ADP | SA8295P | QNN | 14.
|
55 |
-
| FastSam-S | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 7.
|
56 |
-
| FastSam-S | SA8650 (Proxy) | SA8650P Proxy | QNN | 7.
|
57 |
-
| FastSam-S | SA8775P ADP | SA8775P | TFLITE |
|
58 |
-
| FastSam-S | SA8775P ADP | SA8775P | QNN | 14.
|
59 |
-
| FastSam-S | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 13.
|
60 |
-
| FastSam-S | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 13.
|
61 |
-
| FastSam-S | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.
|
62 |
-
| FastSam-S | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 9.
|
63 |
|
64 |
|
65 |
|
66 |
|
67 |
## Installation
|
68 |
|
69 |
-
This model can be installed as a Python package via pip.
|
70 |
|
|
|
71 |
```bash
|
72 |
-
pip install "qai-hub-models[
|
73 |
```
|
74 |
|
75 |
|
76 |
-
|
77 |
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
78 |
|
79 |
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
|
@@ -124,8 +123,8 @@ Profiling Results
|
|
124 |
FastSam-S
|
125 |
Device : Samsung Galaxy S23 (13)
|
126 |
Runtime : TFLITE
|
127 |
-
Estimated inference time (ms) : 7.
|
128 |
-
Estimated peak memory usage (MB): [4,
|
129 |
Total # Ops : 287
|
130 |
Compute Unit(s) : NPU (287 ops)
|
131 |
```
|
@@ -152,7 +151,7 @@ from qai_hub_models.models.fastsam_s import Model
|
|
152 |
torch_model = Model.from_pretrained()
|
153 |
|
154 |
# Device
|
155 |
-
device = hub.Device("Samsung Galaxy
|
156 |
|
157 |
# Trace model
|
158 |
input_shape = torch_model.get_input_spec()
|
@@ -244,7 +243,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
|
|
244 |
|
245 |
|
246 |
## License
|
247 |
-
* The license for the original implementation of FastSam-S can be found
|
|
|
248 |
* The license for the compiled assets for on-device deployment can be found [here](https://github.com/CASIA-IVA-Lab/FastSAM/blob/main/LICENSE)
|
249 |
|
250 |
|
|
|
35 |
|
36 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
+
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 7.672 ms | 4 - 33 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
39 |
+
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.099 ms | 5 - 19 MB | FP16 | NPU | [FastSam-S.so](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.so) |
|
40 |
+
| FastSam-S | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 9.47 ms | 3 - 68 MB | FP16 | NPU | [FastSam-S.onnx](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx) |
|
41 |
+
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.65 ms | 2 - 37 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
42 |
+
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 6.02 ms | 5 - 41 MB | FP16 | NPU | [FastSam-S.so](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.so) |
|
43 |
+
| FastSam-S | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 6.587 ms | 17 - 57 MB | FP16 | NPU | [FastSam-S.onnx](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx) |
|
44 |
+
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 4.394 ms | 0 - 36 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
45 |
+
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 5.555 ms | 5 - 38 MB | FP16 | NPU | Use Export Script |
|
46 |
+
| FastSam-S | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 6.312 ms | 13 - 48 MB | FP16 | NPU | [FastSam-S.onnx](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx) |
|
47 |
+
| FastSam-S | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 7.768 ms | 4 - 32 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
48 |
+
| FastSam-S | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 7.644 ms | 5 - 8 MB | FP16 | NPU | Use Export Script |
|
49 |
+
| FastSam-S | SA7255P ADP | SA7255P | TFLITE | 256.975 ms | 4 - 35 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
50 |
+
| FastSam-S | SA7255P ADP | SA7255P | QNN | 257.258 ms | 5 - 14 MB | FP16 | NPU | Use Export Script |
|
51 |
+
| FastSam-S | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 7.71 ms | 4 - 26 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
52 |
+
| FastSam-S | SA8255 (Proxy) | SA8255P Proxy | QNN | 7.615 ms | 5 - 8 MB | FP16 | NPU | Use Export Script |
|
53 |
+
| FastSam-S | SA8295P ADP | SA8295P | TFLITE | 15.291 ms | 4 - 37 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
54 |
+
| FastSam-S | SA8295P ADP | SA8295P | QNN | 14.221 ms | 0 - 14 MB | FP16 | NPU | Use Export Script |
|
55 |
+
| FastSam-S | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 7.701 ms | 4 - 31 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
56 |
+
| FastSam-S | SA8650 (Proxy) | SA8650P Proxy | QNN | 7.721 ms | 5 - 8 MB | FP16 | NPU | Use Export Script |
|
57 |
+
| FastSam-S | SA8775P ADP | SA8775P | TFLITE | 14.057 ms | 4 - 35 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
58 |
+
| FastSam-S | SA8775P ADP | SA8775P | QNN | 14.157 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
|
59 |
+
| FastSam-S | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 13.845 ms | 4 - 41 MB | FP16 | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) |
|
60 |
+
| FastSam-S | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 13.721 ms | 5 - 40 MB | FP16 | NPU | Use Export Script |
|
61 |
+
| FastSam-S | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.186 ms | 5 - 5 MB | FP16 | NPU | Use Export Script |
|
62 |
+
| FastSam-S | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 9.934 ms | 22 - 22 MB | FP16 | NPU | [FastSam-S.onnx](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx) |
|
63 |
|
64 |
|
65 |
|
66 |
|
67 |
## Installation
|
68 |
|
|
|
69 |
|
70 |
+
Install the package via pip:
|
71 |
```bash
|
72 |
+
pip install "qai-hub-models[fastsam-s]"
|
73 |
```
|
74 |
|
75 |
|
|
|
76 |
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
77 |
|
78 |
Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
|
|
|
123 |
FastSam-S
|
124 |
Device : Samsung Galaxy S23 (13)
|
125 |
Runtime : TFLITE
|
126 |
+
Estimated inference time (ms) : 7.7
|
127 |
+
Estimated peak memory usage (MB): [4, 33]
|
128 |
Total # Ops : 287
|
129 |
Compute Unit(s) : NPU (287 ops)
|
130 |
```
|
|
|
151 |
torch_model = Model.from_pretrained()
|
152 |
|
153 |
# Device
|
154 |
+
device = hub.Device("Samsung Galaxy S24")
|
155 |
|
156 |
# Trace model
|
157 |
input_shape = torch_model.get_input_spec()
|
|
|
243 |
|
244 |
|
245 |
## License
|
246 |
+
* The license for the original implementation of FastSam-S can be found
|
247 |
+
[here](https://github.com/CASIA-IVA-Lab/FastSAM/blob/main/LICENSE).
|
248 |
* The license for the compiled assets for on-device deployment can be found [here](https://github.com/CASIA-IVA-Lab/FastSAM/blob/main/LICENSE)
|
249 |
|
250 |
|