Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -36,56 +36,55 @@ 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 |
-
|
|
| 40 |
-
|
|
| 41 |
-
|
|
| 42 |
-
|
|
| 43 |
-
|
|
| 44 |
-
|
|
| 45 |
-
|
|
| 46 |
-
|
|
| 47 |
-
|
|
| 48 |
-
|
|
| 49 |
-
|
|
| 50 |
-
|
|
| 51 |
-
|
|
| 52 |
-
|
|
| 53 |
-
|
|
| 54 |
-
|
|
| 55 |
-
|
|
| 56 |
-
|
|
| 57 |
-
|
|
| 58 |
-
|
|
| 59 |
-
|
|
| 60 |
-
|
|
| 61 |
-
|
|
| 62 |
-
|
|
| 63 |
-
|
|
| 64 |
-
|
|
| 65 |
-
|
|
| 66 |
-
|
|
| 67 |
-
|
|
| 68 |
-
|
|
| 69 |
-
|
|
| 70 |
-
|
|
| 71 |
-
|
|
| 72 |
-
|
|
| 73 |
-
|
|
| 74 |
-
|
|
| 75 |
-
|
|
| 76 |
-
|
|
| 77 |
-
|
|
| 78 |
-
|
|
| 79 |
-
|
|
| 80 |
-
|
|
| 81 |
-
|
|
| 82 |
-
|
|
| 83 |
-
|
|
| 84 |
-
|
|
| 85 |
-
|
|
| 86 |
-
|
|
| 87 |
-
|
|
| 88 |
-
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.74 ms | 68 - 68 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
|
| 89 |
|
| 90 |
|
| 91 |
|
|
@@ -146,23 +145,23 @@ python -m qai_hub_models.models.trocr.export
|
|
| 146 |
```
|
| 147 |
```
|
| 148 |
Profiling Results
|
| 149 |
-
------------------------------------------------------------
|
| 150 |
-
TrOCREncoder
|
| 151 |
-
Device : Samsung Galaxy S23 (13)
|
| 152 |
-
Runtime : TFLITE
|
| 153 |
-
Estimated inference time (ms) : 50.1
|
| 154 |
-
Estimated peak memory usage (MB): [7, 31]
|
| 155 |
-
Total # Ops : 591
|
| 156 |
-
Compute Unit(s) : NPU (591 ops)
|
| 157 |
-
|
| 158 |
------------------------------------------------------------
|
| 159 |
TrOCRDecoder
|
| 160 |
Device : Samsung Galaxy S23 (13)
|
| 161 |
Runtime : TFLITE
|
| 162 |
Estimated inference time (ms) : 2.2
|
| 163 |
-
Estimated peak memory usage (MB): [0,
|
| 164 |
Total # Ops : 399
|
| 165 |
Compute Unit(s) : NPU (399 ops)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 166 |
```
|
| 167 |
|
| 168 |
|
|
@@ -185,42 +184,42 @@ from qai_hub_models.models.trocr import Model
|
|
| 185 |
|
| 186 |
# Load the model
|
| 187 |
model = Model.from_pretrained()
|
| 188 |
-
encoder_model = model.encoder
|
| 189 |
decoder_model = model.decoder
|
|
|
|
| 190 |
|
| 191 |
# Device
|
| 192 |
device = hub.Device("Samsung Galaxy S23")
|
| 193 |
|
| 194 |
# Trace model
|
| 195 |
-
|
| 196 |
-
|
| 197 |
|
| 198 |
-
|
| 199 |
|
| 200 |
# Compile model on a specific device
|
| 201 |
-
|
| 202 |
-
model=
|
| 203 |
device=device,
|
| 204 |
-
input_specs=
|
| 205 |
)
|
| 206 |
|
| 207 |
# Get target model to run on-device
|
| 208 |
-
|
| 209 |
# Trace model
|
| 210 |
-
|
| 211 |
-
|
| 212 |
|
| 213 |
-
|
| 214 |
|
| 215 |
# Compile model on a specific device
|
| 216 |
-
|
| 217 |
-
model=
|
| 218 |
device=device,
|
| 219 |
-
input_specs=
|
| 220 |
)
|
| 221 |
|
| 222 |
# Get target model to run on-device
|
| 223 |
-
|
| 224 |
|
| 225 |
```
|
| 226 |
|
|
@@ -232,14 +231,14 @@ After compiling models from step 1. Models can be profiled model on-device using
|
|
| 232 |
provisioned in the cloud. Once the job is submitted, you can navigate to a
|
| 233 |
provided job URL to view a variety of on-device performance metrics.
|
| 234 |
```python
|
| 235 |
-
encoder_profile_job = hub.submit_profile_job(
|
| 236 |
-
model=encoder_target_model,
|
| 237 |
-
device=device,
|
| 238 |
-
)
|
| 239 |
decoder_profile_job = hub.submit_profile_job(
|
| 240 |
model=decoder_target_model,
|
| 241 |
device=device,
|
| 242 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
|
| 244 |
```
|
| 245 |
|
|
@@ -248,13 +247,6 @@ Step 3: **Verify on-device accuracy**
|
|
| 248 |
To verify the accuracy of the model on-device, you can run on-device inference
|
| 249 |
on sample input data on the same cloud hosted device.
|
| 250 |
```python
|
| 251 |
-
encoder_input_data = encoder_model.sample_inputs()
|
| 252 |
-
encoder_inference_job = hub.submit_inference_job(
|
| 253 |
-
model=encoder_target_model,
|
| 254 |
-
device=device,
|
| 255 |
-
inputs=encoder_input_data,
|
| 256 |
-
)
|
| 257 |
-
encoder_inference_job.download_output_data()
|
| 258 |
decoder_input_data = decoder_model.sample_inputs()
|
| 259 |
decoder_inference_job = hub.submit_inference_job(
|
| 260 |
model=decoder_target_model,
|
|
@@ -262,6 +254,13 @@ decoder_inference_job = hub.submit_inference_job(
|
|
| 262 |
inputs=decoder_input_data,
|
| 263 |
)
|
| 264 |
decoder_inference_job.download_output_data()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 265 |
|
| 266 |
```
|
| 267 |
With the output of the model, you can compute like PSNR, relative errors or
|
|
|
|
| 36 |
|
| 37 |
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|
| 38 |
|---|---|---|---|---|---|---|---|---|
|
| 39 |
+
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.203 ms | 0 - 143 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 40 |
+
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 2.367 ms | 2 - 355 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) |
|
| 41 |
+
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 2.762 ms | 1 - 3 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
|
| 42 |
+
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.561 ms | 0 - 48 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 43 |
+
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.729 ms | 0 - 51 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) |
|
| 44 |
+
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.1 ms | 0 - 174 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
|
| 45 |
+
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.451 ms | 0 - 46 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 46 |
+
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.545 ms | 0 - 46 MB | FP16 | NPU | Use Export Script |
|
| 47 |
+
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.88 ms | 0 - 132 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
|
| 48 |
+
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.205 ms | 0 - 364 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 49 |
+
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.25 ms | 0 - 1 MB | FP16 | NPU | Use Export Script |
|
| 50 |
+
| TrOCRDecoder | SA7255P ADP | SA7255P | TFLITE | 12.302 ms | 0 - 44 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 51 |
+
| TrOCRDecoder | SA7255P ADP | SA7255P | QNN | 12.414 ms | 7 - 17 MB | FP16 | NPU | Use Export Script |
|
| 52 |
+
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 2.21 ms | 0 - 87 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 53 |
+
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 2.316 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
|
| 54 |
+
| TrOCRDecoder | SA8295P ADP | SA8295P | TFLITE | 3.067 ms | 0 - 42 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 55 |
+
| TrOCRDecoder | SA8295P ADP | SA8295P | QNN | 3.74 ms | 7 - 13 MB | FP16 | NPU | Use Export Script |
|
| 56 |
+
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 2.27 ms | 0 - 346 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 57 |
+
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 2.35 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
|
| 58 |
+
| TrOCRDecoder | SA8775P ADP | SA8775P | TFLITE | 3.341 ms | 0 - 45 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 59 |
+
| TrOCRDecoder | SA8775P ADP | SA8775P | QNN | 3.578 ms | 7 - 13 MB | FP16 | NPU | Use Export Script |
|
| 60 |
+
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.671 ms | 0 - 48 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
|
| 61 |
+
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 2.74 ms | 4 - 56 MB | FP16 | NPU | Use Export Script |
|
| 62 |
+
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.444 ms | 7 - 7 MB | FP16 | NPU | Use Export Script |
|
| 63 |
+
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.741 ms | 69 - 69 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
|
| 64 |
+
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 50.015 ms | 7 - 34 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 65 |
+
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 53.281 ms | 2 - 22 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) |
|
| 66 |
+
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 38.056 ms | 0 - 57 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
|
| 67 |
+
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 39.322 ms | 5 - 67 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 68 |
+
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 41.406 ms | 2 - 61 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) |
|
| 69 |
+
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 31.095 ms | 0 - 259 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
|
| 70 |
+
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 36.222 ms | 5 - 68 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 71 |
+
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 33.821 ms | 2 - 66 MB | FP16 | NPU | Use Export Script |
|
| 72 |
+
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 24.497 ms | 16 - 138 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
|
| 73 |
+
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 49.833 ms | 7 - 32 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 74 |
+
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 36.818 ms | 2 - 8 MB | FP16 | NPU | Use Export Script |
|
| 75 |
+
| TrOCREncoder | SA7255P ADP | SA7255P | TFLITE | 266.53 ms | 7 - 69 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 76 |
+
| TrOCREncoder | SA7255P ADP | SA7255P | QNN | 247.644 ms | 2 - 12 MB | FP16 | NPU | Use Export Script |
|
| 77 |
+
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 50.253 ms | 7 - 30 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 78 |
+
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 37.723 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
|
| 79 |
+
| TrOCREncoder | SA8295P ADP | SA8295P | QNN | 50.866 ms | 4 - 10 MB | FP16 | NPU | Use Export Script |
|
| 80 |
+
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 50.307 ms | 7 - 34 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 81 |
+
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 37.01 ms | 2 - 3 MB | FP16 | NPU | Use Export Script |
|
| 82 |
+
| TrOCREncoder | SA8775P ADP | SA8775P | TFLITE | 59.803 ms | 7 - 69 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 83 |
+
| TrOCREncoder | SA8775P ADP | SA8775P | QNN | 42.412 ms | 2 - 8 MB | FP16 | NPU | Use Export Script |
|
| 84 |
+
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 60.304 ms | 7 - 69 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
|
| 85 |
+
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 63.0 ms | 0 - 64 MB | FP16 | NPU | Use Export Script |
|
| 86 |
+
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 34.029 ms | 2 - 2 MB | FP16 | NPU | Use Export Script |
|
| 87 |
+
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 36.913 ms | 49 - 49 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
|
|
|
|
| 88 |
|
| 89 |
|
| 90 |
|
|
|
|
| 145 |
```
|
| 146 |
```
|
| 147 |
Profiling Results
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 148 |
------------------------------------------------------------
|
| 149 |
TrOCRDecoder
|
| 150 |
Device : Samsung Galaxy S23 (13)
|
| 151 |
Runtime : TFLITE
|
| 152 |
Estimated inference time (ms) : 2.2
|
| 153 |
+
Estimated peak memory usage (MB): [0, 143]
|
| 154 |
Total # Ops : 399
|
| 155 |
Compute Unit(s) : NPU (399 ops)
|
| 156 |
+
|
| 157 |
+
------------------------------------------------------------
|
| 158 |
+
TrOCREncoder
|
| 159 |
+
Device : Samsung Galaxy S23 (13)
|
| 160 |
+
Runtime : TFLITE
|
| 161 |
+
Estimated inference time (ms) : 50.0
|
| 162 |
+
Estimated peak memory usage (MB): [7, 34]
|
| 163 |
+
Total # Ops : 591
|
| 164 |
+
Compute Unit(s) : NPU (591 ops)
|
| 165 |
```
|
| 166 |
|
| 167 |
|
|
|
|
| 184 |
|
| 185 |
# Load the model
|
| 186 |
model = Model.from_pretrained()
|
|
|
|
| 187 |
decoder_model = model.decoder
|
| 188 |
+
encoder_model = model.encoder
|
| 189 |
|
| 190 |
# Device
|
| 191 |
device = hub.Device("Samsung Galaxy S23")
|
| 192 |
|
| 193 |
# Trace model
|
| 194 |
+
decoder_input_shape = decoder_model.get_input_spec()
|
| 195 |
+
decoder_sample_inputs = decoder_model.sample_inputs()
|
| 196 |
|
| 197 |
+
traced_decoder_model = torch.jit.trace(decoder_model, [torch.tensor(data[0]) for _, data in decoder_sample_inputs.items()])
|
| 198 |
|
| 199 |
# Compile model on a specific device
|
| 200 |
+
decoder_compile_job = hub.submit_compile_job(
|
| 201 |
+
model=traced_decoder_model ,
|
| 202 |
device=device,
|
| 203 |
+
input_specs=decoder_model.get_input_spec(),
|
| 204 |
)
|
| 205 |
|
| 206 |
# Get target model to run on-device
|
| 207 |
+
decoder_target_model = decoder_compile_job.get_target_model()
|
| 208 |
# Trace model
|
| 209 |
+
encoder_input_shape = encoder_model.get_input_spec()
|
| 210 |
+
encoder_sample_inputs = encoder_model.sample_inputs()
|
| 211 |
|
| 212 |
+
traced_encoder_model = torch.jit.trace(encoder_model, [torch.tensor(data[0]) for _, data in encoder_sample_inputs.items()])
|
| 213 |
|
| 214 |
# Compile model on a specific device
|
| 215 |
+
encoder_compile_job = hub.submit_compile_job(
|
| 216 |
+
model=traced_encoder_model ,
|
| 217 |
device=device,
|
| 218 |
+
input_specs=encoder_model.get_input_spec(),
|
| 219 |
)
|
| 220 |
|
| 221 |
# Get target model to run on-device
|
| 222 |
+
encoder_target_model = encoder_compile_job.get_target_model()
|
| 223 |
|
| 224 |
```
|
| 225 |
|
|
|
|
| 231 |
provisioned in the cloud. Once the job is submitted, you can navigate to a
|
| 232 |
provided job URL to view a variety of on-device performance metrics.
|
| 233 |
```python
|
|
|
|
|
|
|
|
|
|
|
|
|
| 234 |
decoder_profile_job = hub.submit_profile_job(
|
| 235 |
model=decoder_target_model,
|
| 236 |
device=device,
|
| 237 |
)
|
| 238 |
+
encoder_profile_job = hub.submit_profile_job(
|
| 239 |
+
model=encoder_target_model,
|
| 240 |
+
device=device,
|
| 241 |
+
)
|
| 242 |
|
| 243 |
```
|
| 244 |
|
|
|
|
| 247 |
To verify the accuracy of the model on-device, you can run on-device inference
|
| 248 |
on sample input data on the same cloud hosted device.
|
| 249 |
```python
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 250 |
decoder_input_data = decoder_model.sample_inputs()
|
| 251 |
decoder_inference_job = hub.submit_inference_job(
|
| 252 |
model=decoder_target_model,
|
|
|
|
| 254 |
inputs=decoder_input_data,
|
| 255 |
)
|
| 256 |
decoder_inference_job.download_output_data()
|
| 257 |
+
encoder_input_data = encoder_model.sample_inputs()
|
| 258 |
+
encoder_inference_job = hub.submit_inference_job(
|
| 259 |
+
model=encoder_target_model,
|
| 260 |
+
device=device,
|
| 261 |
+
inputs=encoder_input_data,
|
| 262 |
+
)
|
| 263 |
+
encoder_inference_job.download_output_data()
|
| 264 |
|
| 265 |
```
|
| 266 |
With the output of the model, you can compute like PSNR, relative errors or
|