Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -35,32 +35,34 @@ 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 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.
|
39 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.
|
40 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 5.
|
41 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.
|
42 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.
|
43 |
-
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 4.
|
44 |
-
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.
|
45 |
-
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.
|
46 |
-
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 4.
|
47 |
-
| QuickSRNetMedium-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE |
|
48 |
-
| QuickSRNetMedium-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 1.
|
49 |
-
| QuickSRNetMedium-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 12.
|
50 |
-
| QuickSRNetMedium-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.
|
51 |
-
| QuickSRNetMedium-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.
|
52 |
-
| QuickSRNetMedium-Quantized |
|
53 |
-
| QuickSRNetMedium-Quantized |
|
54 |
-
| QuickSRNetMedium-Quantized |
|
55 |
-
| QuickSRNetMedium-Quantized |
|
56 |
-
| QuickSRNetMedium-Quantized |
|
57 |
-
| QuickSRNetMedium-Quantized |
|
58 |
-
| QuickSRNetMedium-Quantized |
|
59 |
-
| QuickSRNetMedium-Quantized |
|
60 |
-
| QuickSRNetMedium-Quantized |
|
61 |
-
| QuickSRNetMedium-Quantized |
|
62 |
-
| QuickSRNetMedium-Quantized |
|
63 |
-
| QuickSRNetMedium-Quantized |
|
|
|
|
|
64 |
|
65 |
|
66 |
|
@@ -125,7 +127,7 @@ Profiling Results
|
|
125 |
QuickSRNetMedium-Quantized
|
126 |
Device : Samsung Galaxy S23 (13)
|
127 |
Runtime : TFLITE
|
128 |
-
Estimated inference time (ms) : 1.
|
129 |
Estimated peak memory usage (MB): [0, 6]
|
130 |
Total # Ops : 19
|
131 |
Compute Unit(s) : NPU (16 ops) CPU (3 ops)
|
@@ -147,13 +149,29 @@ in memory using the `jit.trace` and then call the `submit_compile_job` API.
|
|
147 |
import torch
|
148 |
|
149 |
import qai_hub as hub
|
150 |
-
from qai_hub_models.models.quicksrnetmedium_quantized import
|
151 |
|
152 |
# Load the model
|
|
|
153 |
|
154 |
# Device
|
155 |
device = hub.Device("Samsung Galaxy S23")
|
156 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
157 |
|
158 |
```
|
159 |
|
|
|
35 |
|
36 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
37 |
|---|---|---|---|---|---|---|---|---|
|
38 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.197 ms | 0 - 6 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
39 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.517 ms | 0 - 10 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.so) |
|
40 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 5.435 ms | 12 - 14 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.onnx) |
|
41 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.909 ms | 0 - 12 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
42 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.353 ms | 0 - 11 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.so](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.so) |
|
43 |
+
| QuickSRNetMedium-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 4.52 ms | 13 - 46 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.onnx) |
|
44 |
+
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.396 ms | 0 - 9 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
45 |
+
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.296 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
|
46 |
+
| QuickSRNetMedium-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 4.203 ms | 12 - 33 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.onnx) |
|
47 |
+
| QuickSRNetMedium-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 2.337 ms | 0 - 11 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
48 |
+
| QuickSRNetMedium-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 1.046 ms | 0 - 7 MB | INT8 | NPU | Use Export Script |
|
49 |
+
| QuickSRNetMedium-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 12.837 ms | 2 - 9 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
50 |
+
| QuickSRNetMedium-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.125 ms | 0 - 2 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
51 |
+
| QuickSRNetMedium-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.41 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
|
52 |
+
| QuickSRNetMedium-Quantized | SA7255P ADP | SA7255P | TFLITE | 4.58 ms | 2 - 11 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
53 |
+
| QuickSRNetMedium-Quantized | SA7255P ADP | SA7255P | QNN | 2.764 ms | 0 - 5 MB | INT8 | NPU | Use Export Script |
|
54 |
+
| QuickSRNetMedium-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.122 ms | 0 - 3 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
55 |
+
| QuickSRNetMedium-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.414 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
|
56 |
+
| QuickSRNetMedium-Quantized | SA8295P ADP | SA8295P | TFLITE | 1.909 ms | 0 - 8 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
57 |
+
| QuickSRNetMedium-Quantized | SA8295P ADP | SA8295P | QNN | 0.897 ms | 0 - 6 MB | INT8 | NPU | Use Export Script |
|
58 |
+
| QuickSRNetMedium-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.111 ms | 0 - 8 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
59 |
+
| QuickSRNetMedium-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.441 ms | 0 - 1 MB | INT8 | NPU | Use Export Script |
|
60 |
+
| QuickSRNetMedium-Quantized | SA8775P ADP | SA8775P | TFLITE | 2.117 ms | 1 - 9 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
61 |
+
| QuickSRNetMedium-Quantized | SA8775P ADP | SA8775P | QNN | 0.964 ms | 0 - 5 MB | INT8 | NPU | Use Export Script |
|
62 |
+
| QuickSRNetMedium-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.051 ms | 2 - 14 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.tflite) |
|
63 |
+
| QuickSRNetMedium-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.582 ms | 0 - 12 MB | INT8 | NPU | Use Export Script |
|
64 |
+
| QuickSRNetMedium-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.502 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
|
65 |
+
| QuickSRNetMedium-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 6.063 ms | 15 - 15 MB | INT8 | NPU | [QuickSRNetMedium-Quantized.onnx](https://huggingface.co/qualcomm/QuickSRNetMedium-Quantized/blob/main/QuickSRNetMedium-Quantized.onnx) |
|
66 |
|
67 |
|
68 |
|
|
|
127 |
QuickSRNetMedium-Quantized
|
128 |
Device : Samsung Galaxy S23 (13)
|
129 |
Runtime : TFLITE
|
130 |
+
Estimated inference time (ms) : 1.2
|
131 |
Estimated peak memory usage (MB): [0, 6]
|
132 |
Total # Ops : 19
|
133 |
Compute Unit(s) : NPU (16 ops) CPU (3 ops)
|
|
|
149 |
import torch
|
150 |
|
151 |
import qai_hub as hub
|
152 |
+
from qai_hub_models.models.quicksrnetmedium_quantized import Model
|
153 |
|
154 |
# Load the model
|
155 |
+
torch_model = Model.from_pretrained()
|
156 |
|
157 |
# Device
|
158 |
device = hub.Device("Samsung Galaxy S23")
|
159 |
|
160 |
+
# Trace model
|
161 |
+
input_shape = torch_model.get_input_spec()
|
162 |
+
sample_inputs = torch_model.sample_inputs()
|
163 |
+
|
164 |
+
pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
|
165 |
+
|
166 |
+
# Compile model on a specific device
|
167 |
+
compile_job = hub.submit_compile_job(
|
168 |
+
model=pt_model,
|
169 |
+
device=device,
|
170 |
+
input_specs=torch_model.get_input_spec(),
|
171 |
+
)
|
172 |
+
|
173 |
+
# Get target model to run on-device
|
174 |
+
target_model = compile_job.get_target_model()
|
175 |
|
176 |
```
|
177 |
|