Hugging Face's logo

qualcomm
/
FFNet-78S-LowRes

Image Segmentation
PyTorch
LiteRT
real_time
android
Model card Files
xet
Community
bhushans commited on
Commit
a1eb53d
·
verified ·
1 Parent(s): 8d26cb3

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +43 -25
README.md CHANGED
@@ -36,29 +36,31 @@ 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
- | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 8.264 ms | 0 - 2 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
40
- | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.274 ms | 6 - 32 MB | FP16 | NPU | [FFNet-78S-LowRes.so](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.so) |
41
- | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 8.079 ms | 0 - 51 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
42
- | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.721 ms | 1 - 62 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
43
- | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 5.849 ms | 6 - 31 MB | FP16 | NPU | [FFNet-78S-LowRes.so](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.so) |
44
- | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 5.848 ms | 7 - 88 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
45
- | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 4.707 ms | 0 - 29 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
46
- | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 4.754 ms | 0 - 21 MB | FP16 | NPU | Use Export Script |
47
- | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 5.862 ms | 3 - 46 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
48
- | FFNet-78S-LowRes | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 8.143 ms | 1 - 2 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
49
- | FFNet-78S-LowRes | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 7.645 ms | 6 - 7 MB | FP16 | NPU | Use Export Script |
50
- | FFNet-78S-LowRes | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 8.25 ms | 1 - 2 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
51
- | FFNet-78S-LowRes | SA8255 (Proxy) | SA8255P Proxy | QNN | 7.736 ms | 6 - 8 MB | FP16 | NPU | Use Export Script |
52
- | FFNet-78S-LowRes | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 8.276 ms | 0 - 2 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
53
- | FFNet-78S-LowRes | SA8775 (Proxy) | SA8775P Proxy | QNN | 7.737 ms | 6 - 8 MB | FP16 | NPU | Use Export Script |
54
- | FFNet-78S-LowRes | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 8.265 ms | 0 - 24 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
55
- | FFNet-78S-LowRes | SA8650 (Proxy) | SA8650P Proxy | QNN | 7.624 ms | 6 - 7 MB | FP16 | NPU | Use Export Script |
56
- | FFNet-78S-LowRes | SA8295P ADP | SA8295P | TFLITE | 14.374 ms | 1 - 25 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
57
- | FFNet-78S-LowRes | SA8295P ADP | SA8295P | QNN | 13.596 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
58
- | FFNet-78S-LowRes | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 12.361 ms | 1 - 51 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
59
- | FFNet-78S-LowRes | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 12.443 ms | 6 - 26 MB | FP16 | NPU | Use Export Script |
60
- | FFNet-78S-LowRes | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.123 ms | 6 - 6 MB | FP16 | NPU | Use Export Script |
61
- | FFNet-78S-LowRes | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 8.829 ms | 47 - 47 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
 
 
62
 
63
 
64
 
@@ -124,7 +126,7 @@ FFNet-78S-LowRes
124
  Device : Samsung Galaxy S23 (13)
125
  Runtime : TFLITE
126
  Estimated inference time (ms) : 8.3
127
- Estimated peak memory usage (MB): [0, 2]
128
  Total # Ops : 149
129
  Compute Unit(s) : NPU (149 ops)
130
  ```
@@ -145,13 +147,29 @@ in memory using the `jit.trace` and then call the `submit_compile_job` API.
145
  import torch
146
 
147
  import qai_hub as hub
148
- from qai_hub_models.models.ffnet_78s_lowres import
149
 
150
  # Load the model
 
151
 
152
  # Device
153
  device = hub.Device("Samsung Galaxy S23")
154
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
155
 
156
  ```
157
 
 
36
 
37
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
38
  |---|---|---|---|---|---|---|---|---|
39
+ | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 8.256 ms | 0 - 51 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
40
+ | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.341 ms | 6 - 30 MB | FP16 | NPU | [FFNet-78S-LowRes.so](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.so) |
41
+ | FFNet-78S-LowRes | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 7.912 ms | 0 - 51 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
42
+ | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.748 ms | 1 - 26 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
43
+ | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 5.793 ms | 6 - 32 MB | FP16 | NPU | [FFNet-78S-LowRes.so](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.so) |
44
+ | FFNet-78S-LowRes | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 5.817 ms | 7 - 91 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
45
+ | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 5.64 ms | 0 - 22 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
46
+ | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 4.777 ms | 6 - 26 MB | FP16 | NPU | Use Export Script |
47
+ | FFNet-78S-LowRes | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 5.824 ms | 2 - 46 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
48
+ | FFNet-78S-LowRes | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 8.172 ms | 0 - 42 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
49
+ | FFNet-78S-LowRes | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 7.604 ms | 6 - 7 MB | FP16 | NPU | Use Export Script |
50
+ | FFNet-78S-LowRes | SA7255P ADP | SA7255P | TFLITE | 289.313 ms | 1 - 22 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
51
+ | FFNet-78S-LowRes | SA7255P ADP | SA7255P | QNN | 288.598 ms | 4 - 9 MB | FP16 | NPU | Use Export Script |
52
+ | FFNet-78S-LowRes | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 8.357 ms | 1 - 24 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
53
+ | FFNet-78S-LowRes | SA8255 (Proxy) | SA8255P Proxy | QNN | 7.706 ms | 6 - 7 MB | FP16 | NPU | Use Export Script |
54
+ | FFNet-78S-LowRes | SA8295P ADP | SA8295P | TFLITE | 13.989 ms | 1 - 17 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
55
+ | FFNet-78S-LowRes | SA8295P ADP | SA8295P | QNN | 13.538 ms | 0 - 6 MB | FP16 | NPU | Use Export Script |
56
+ | FFNet-78S-LowRes | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 8.152 ms | 0 - 62 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
57
+ | FFNet-78S-LowRes | SA8650 (Proxy) | SA8650P Proxy | QNN | 7.592 ms | 6 - 7 MB | FP16 | NPU | Use Export Script |
58
+ | FFNet-78S-LowRes | SA8775P ADP | SA8775P | TFLITE | 15.754 ms | 1 - 21 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
59
+ | FFNet-78S-LowRes | SA8775P ADP | SA8775P | QNN | 15.156 ms | 1 - 7 MB | FP16 | NPU | Use Export Script |
60
+ | FFNet-78S-LowRes | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 12.133 ms | 1 - 25 MB | FP16 | NPU | [FFNet-78S-LowRes.tflite](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.tflite) |
61
+ | FFNet-78S-LowRes | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 12.541 ms | 6 - 26 MB | FP16 | NPU | Use Export Script |
62
+ | FFNet-78S-LowRes | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.199 ms | 6 - 6 MB | FP16 | NPU | Use Export Script |
63
+ | FFNet-78S-LowRes | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 8.874 ms | 47 - 47 MB | FP16 | NPU | [FFNet-78S-LowRes.onnx](https://huggingface.co/qualcomm/FFNet-78S-LowRes/blob/main/FFNet-78S-LowRes.onnx) |
64
 
65
 
66
 
 
126
  Device : Samsung Galaxy S23 (13)
127
  Runtime : TFLITE
128
  Estimated inference time (ms) : 8.3
129
+ Estimated peak memory usage (MB): [0, 51]
130
  Total # Ops : 149
131
  Compute Unit(s) : NPU (149 ops)
132
  ```
 
147
  import torch
148
 
149
  import qai_hub as hub
150
+ from qai_hub_models.models.ffnet_78s_lowres import Model
151
 
152
  # Load the model
153
+ torch_model = Model.from_pretrained()
154
 
155
  # Device
156
  device = hub.Device("Samsung Galaxy S23")
157
 
158
+ # Trace model
159
+ input_shape = torch_model.get_input_spec()
160
+ sample_inputs = torch_model.sample_inputs()
161
+
162
+ pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
163
+
164
+ # Compile model on a specific device
165
+ compile_job = hub.submit_compile_job(
166
+ model=pt_model,
167
+ device=device,
168
+ input_specs=torch_model.get_input_spec(),
169
+ )
170
+
171
+ # Get target model to run on-device
172
+ target_model = compile_job.get_target_model()
173
 
174
  ```
175