Hugging Face's logo

qualcomm
/
First-Order-Motion-Model

Image-to-Video
PyTorch
LiteRT
android
Model card Files
xet
Community
qaihm-bot commited on
Commit
39308bb
·
verified ·
1 Parent(s): 63eb70b

v0.42.0

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.42.0 for changelog.

Files changed (5) hide show
  1. First-Order-Motion-Model_FOMMDetector_float.onnx.zip +1 -1
  2. First-Order-Motion-Model_FOMMGenerator_float.onnx.zip +2 -2
  3. First-Order-Motion-Model_FOMMGenerator_float.tflite +2 -2
  4. README.md +18 -18
  5. tool-versions.yaml +1 -1
First-Order-Motion-Model_FOMMDetector_float.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:49b256e4d4c3e3b24e008f49b8ae954a8206c69d93824255fdf0efd4f890c544
3
  size 52977831
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c415156a4751da68b7633db1958d714745b752d19b85a0dc89c3f8cd030475e1
3
  size 52977831
First-Order-Motion-Model_FOMMGenerator_float.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:611a7f33df762a4be6078cf76088667ecbcfa28cfa17741657ba449a70914440
3
- size 169943553
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fa422c19dbd29263f4f5ea04d3dcdee5f526ab9dd918b3e259f1817b09b6012
3
+ size 170025547
First-Order-Motion-Model_FOMMGenerator_float.tflite CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2dbfa2b1b5675059dc3c61a9f079fa03b92e6b8ed4454ad40fc5c15e87956f8d
3
- size 182855312
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6017ea1026ab44af52ec47b1e95b5fdee4358a01e5844a1cad8f8fabe0e79285
3
+ size 183183072
README.md CHANGED
@@ -35,20 +35,20 @@ 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
- | FOMMDetector | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 4.614 ms | 0 - 86 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
39
- | FOMMDetector | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.423 ms | 0 - 23 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
40
- | FOMMDetector | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 3.038 ms | 0 - 24 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
41
- | FOMMDetector | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 3.031 ms | 0 - 23 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
42
- | FOMMDetector | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 2.885 ms | 0 - 23 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
43
- | FOMMDetector | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 2.909 ms | 1 - 22 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
44
- | FOMMDetector | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.67 ms | 28 - 28 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
45
- | FOMMGenerator | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 21.912 ms | 0 - 106 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
46
- | FOMMGenerator | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 15.867 ms | 18 - 87 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
47
- | FOMMGenerator | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 29.908 ms | 1 - 730 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
48
- | FOMMGenerator | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 13.837 ms | 16 - 72 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
49
- | FOMMGenerator | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 25.79 ms | 1 - 780 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
50
- | FOMMGenerator | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 10.707 ms | 18 - 88 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
51
- | FOMMGenerator | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 21.585 ms | 88 - 88 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
52
 
53
 
54
 
@@ -63,9 +63,9 @@ pip install "qai-hub-models[fomm]"
63
  ```
64
 
65
 
66
- ## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
67
 
68
- Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
69
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
70
 
71
  With this API token, you can configure your client to run models on the cloud
@@ -73,7 +73,7 @@ hosted devices.
73
  ```bash
74
  qai-hub configure --api_token API_TOKEN
75
  ```
76
- Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.
77
 
78
 
79
 
@@ -184,7 +184,7 @@ With the output of the model, you can compute like PSNR, relative errors or
184
  spot check the output with expected output.
185
 
186
  **Note**: This on-device profiling and inference requires access to Qualcomm®
187
- AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
188
 
189
 
190
 
 
35
 
36
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | FOMMDetector | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 4.608 ms | 0 - 91 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
39
+ | FOMMDetector | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.43 ms | 0 - 24 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
40
+ | FOMMDetector | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 3.043 ms | 0 - 26 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
41
+ | FOMMDetector | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 3.033 ms | 0 - 23 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
42
+ | FOMMDetector | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 2.888 ms | 0 - 25 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
43
+ | FOMMDetector | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 2.907 ms | 1 - 19 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
44
+ | FOMMDetector | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.692 ms | 28 - 28 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
45
+ | FOMMGenerator | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 23.17 ms | 0 - 104 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
46
+ | FOMMGenerator | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 16.557 ms | 18 - 80 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
47
+ | FOMMGenerator | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 31.33 ms | 0 - 762 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
48
+ | FOMMGenerator | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 13.956 ms | 6 - 63 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
49
+ | FOMMGenerator | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 25.442 ms | 1 - 701 MB | NPU | [First-Order-Motion-Model.tflite](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.tflite) |
50
+ | FOMMGenerator | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 11.012 ms | 19 - 86 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
51
+ | FOMMGenerator | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 23.04 ms | 88 - 88 MB | NPU | [First-Order-Motion-Model.onnx.zip](https://huggingface.co/qualcomm/First-Order-Motion-Model/blob/main/First-Order-Motion-Model.onnx.zip) |
52
 
53
 
54
 
 
63
  ```
64
 
65
 
66
+ ## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
67
 
68
+ Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
69
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
70
 
71
  With this API token, you can configure your client to run models on the cloud
 
73
  ```bash
74
  qai-hub configure --api_token API_TOKEN
75
  ```
76
+ Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
77
 
78
 
79
 
 
184
  spot check the output with expected output.
185
 
186
  **Note**: This on-device profiling and inference requires access to Qualcomm®
187
+ AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
188
 
189
 
190
 
tool-versions.yaml CHANGED
@@ -1,4 +1,4 @@
1
  tool_versions:
2
  tflite:
3
- qairt: 2.39.0.250925215840_163802
4
  tflite: 2.17.0
 
1
  tool_versions:
2
  tflite:
3
+ qairt: 2.40.0.251030114326_189385
4
  tflite: 2.17.0