File size: 5,912 Bytes
213d473 b2b8506 213d473 b2b8506 213d473 b2b8506 213d473 b2b8506 213d473 b2b8506 213d473 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 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 89 90 91 92 93 94 95 96 97 98 99 100 101 102 | ---
library_name: pytorch
license: other
tags:
- android
pipeline_tag: robotics
---
# ACT: Optimized for Qualcomm Devices
ACT (Action Chunking with Transformers) is a robotic policy model that is trained to predict the next chunk of actions that the robotic hand is expected to perform.
This is based on the implementation of ACT found [here](https://github.com/tonyzhaozh/act).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/act) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
## Getting Started
There are two ways to deploy this model on your device:
### Option 1: Download Pre-Exported Models
Below are pre-exported model assets ready for deployment.
| Runtime | Precision | Chipset | SDK Versions | Download |
|---|---|---|---|---|
| ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/act/releases/v0.48.0/act-onnx-float.zip)
| QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/act/releases/v0.48.0/act-qnn_dlc-float.zip)
| TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/act/releases/v0.48.0/act-tflite-float.zip)
For more device-specific assets and performance metrics, visit **[ACT on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/act)**.
### Option 2: Export with Custom Configurations
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/act) Python library to compile and export the model with your own:
- Custom weights (e.g., fine-tuned checkpoints)
- Custom input shapes
- Target device and runtime configurations
This option is ideal if you need to customize the model beyond the default configuration provided here.
See our repository for [ACT on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/act) for usage instructions.
## Model Details
**Model Type:** Model_use_case.robotics
**Model Stats:**
- Model checkpoint: act
- Input resolution: 480x640
- Number of parameters: 83.92M
- Model size (float): 255M
## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| ACT | ONNX | float | Snapdragon® X2 Elite | 6.165 ms | 63 - 63 MB | NPU
| ACT | ONNX | float | Snapdragon® X Elite | 11.872 ms | 62 - 62 MB | NPU
| ACT | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 8.103 ms | 3 - 385 MB | NPU
| ACT | ONNX | float | Qualcomm® QCS8550 (Proxy) | 11.214 ms | 0 - 81 MB | NPU
| ACT | ONNX | float | Qualcomm® QCS9075 | 18.835 ms | 4 - 10 MB | NPU
| ACT | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 6.573 ms | 3 - 331 MB | NPU
| ACT | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.496 ms | 4 - 334 MB | NPU
| ACT | QNN_DLC | float | Snapdragon® X2 Elite | 4.932 ms | 4 - 4 MB | NPU
| ACT | QNN_DLC | float | Snapdragon® X Elite | 8.926 ms | 4 - 4 MB | NPU
| ACT | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 5.974 ms | 1 - 344 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 43.942 ms | 1 - 308 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 8.288 ms | 4 - 7 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® SA8775P | 13.325 ms | 1 - 294 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® QCS9075 | 15.977 ms | 4 - 9 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 16.41 ms | 2 - 263 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® SA7255P | 43.942 ms | 1 - 308 MB | NPU
| ACT | QNN_DLC | float | Qualcomm® SA8295P | 15.517 ms | 0 - 230 MB | NPU
| ACT | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.768 ms | 4 - 299 MB | NPU
| ACT | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.739 ms | 4 - 321 MB | NPU
| ACT | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.981 ms | 0 - 361 MB | NPU
| ACT | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 44.134 ms | 0 - 318 MB | NPU
| ACT | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 8.318 ms | 0 - 3 MB | NPU
| ACT | TFLITE | float | Qualcomm® SA8775P | 13.44 ms | 0 - 304 MB | NPU
| ACT | TFLITE | float | Qualcomm® QCS9075 | 16.034 ms | 0 - 71 MB | NPU
| ACT | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 16.392 ms | 0 - 276 MB | NPU
| ACT | TFLITE | float | Qualcomm® SA7255P | 44.134 ms | 0 - 318 MB | NPU
| ACT | TFLITE | float | Qualcomm® SA8295P | 15.711 ms | 0 - 233 MB | NPU
| ACT | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.816 ms | 0 - 310 MB | NPU
| ACT | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.799 ms | 0 - 327 MB | NPU
## License
* The license for the original implementation of ACT can be found
[here](https://github.com/tonyzhaozh/act/blob/main/LICENSE).
## References
* [Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware](https://arxiv.org/abs/2304.13705)
* [Source Model Implementation](https://github.com/tonyzhaozh/act)
## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).
|