v0.46.1

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.46.1 for changelog.

Files changed (5) hide show

README.md +73 -221
Sequencer2D_float.dlc +0 -3
Sequencer2D_float.onnx.zip +0 -3
Sequencer2D_float.tflite +0 -3
tool-versions.yaml +0 -4

README.md CHANGED Viewed

@@ -9,239 +9,91 @@ pipeline_tag: image-classification
 ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/web-assets/model_demo.png)
-# Sequencer2D: Optimized for Mobile Deployment
-## Imagenet classifier and general purpose backbone
 Sequencer2D is a vision transformer model that can classify images from the Imagenet dataset.
-This model is an implementation of Sequencer2D found [here](https://github.com/okojoalg/sequencer).
-This repository provides scripts to run Sequencer2D on Qualcomm® devices.
-More details on model performance across various devices, can be found
-[here](https://aihub.qualcomm.com/models/sequencer2d).
-### Model Details
-- **Model Type:** Model_use_case.image_classification
-- **Model Stats:**
-  - Model checkpoint: sequencer2d_s
-  - Input resolution: 224x224
-  - Number of parameters: 27.6M
-  - Model size (float): 106 MB
-  - Model size (w8a8): 69.1 MB
-| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
-|---|---|---|---|---|---|---|---|---|
-| Sequencer2D | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 36.957 ms | 0 - 735 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 52.6 ms | 1 - 939 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 21.357 ms | 0 - 630 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 26.43 ms | 1 - 864 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 16.809 ms | 0 - 12 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 20.682 ms | 1 - 3 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 45.919 ms | 0 - 83 MB | NPU | [Sequencer2D.onnx.zip](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.onnx.zip) |
-| Sequencer2D | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 19.466 ms | 0 - 731 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 24.943 ms | 1 - 895 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 12.066 ms | 0 - 882 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 14.631 ms | 0 - 2010 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 33.094 ms | 1 - 1349 MB | NPU | [Sequencer2D.onnx.zip](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.onnx.zip) |
-| Sequencer2D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 9.455 ms | 0 - 792 MB | NPU | [Sequencer2D.tflite](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.tflite) |
-| Sequencer2D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 12.411 ms | 0 - 982 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 23.242 ms | 1 - 767 MB | NPU | [Sequencer2D.onnx.zip](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.onnx.zip) |
-| Sequencer2D | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_DLC | 9.278 ms | 1 - 1181 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | ONNX | 17.922 ms | 1 - 807 MB | NPU | [Sequencer2D.onnx.zip](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.onnx.zip) |
-| Sequencer2D | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 21.518 ms | 1 - 1 MB | NPU | [Sequencer2D.dlc](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.dlc) |
-| Sequencer2D | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 48.381 ms | 66 - 66 MB | NPU | [Sequencer2D.onnx.zip](https://huggingface.co/qualcomm/Sequencer2D/blob/main/Sequencer2D.onnx.zip) |
-## Installation
-Install the package via pip:
-```bash
-# NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
-pip install "qai-hub-models[sequencer2d]"
-```
-## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
-Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
-Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
-With this API token, you can configure your client to run models on the cloud
-hosted devices.
-```bash
-qai-hub configure --api_token API_TOKEN
-```
-Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
-## Demo off target
-The package contains a simple end-to-end demo that downloads pre-trained
-weights and runs this model on a sample input.
-```bash
-python -m qai_hub_models.models.sequencer2d.demo
-```
-The above demo runs a reference implementation of pre-processing, model
-inference, and post processing.
-**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
-environment, please add the following to your cell (instead of the above).
-```
-%run -m qai_hub_models.models.sequencer2d.demo
-```
-### Run model on a cloud-hosted device
-In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
-device. This script does the following:
-* Performance check on-device on a cloud-hosted device
-* Downloads compiled assets that can be deployed on-device for Android.
-* Accuracy check between PyTorch and on-device outputs.
-```bash
-python -m qai_hub_models.models.sequencer2d.export
-```
-## How does this work?
-This [export script](https://aihub.qualcomm.com/models/sequencer2d/qai_hub_models/models/Sequencer2D/export.py)
-leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
-on-device. Lets go through each step below in detail:
-Step 1: **Compile model for on-device deployment**
-To compile a PyTorch model for on-device deployment, we first trace the model
-in memory using the `jit.trace` and then call the `submit_compile_job` API.
-```python
-import torch
-import qai_hub as hub
-from qai_hub_models.models.sequencer2d import Model
-# Load the model
-torch_model = Model.from_pretrained()
-# Device
-device = hub.Device("Samsung Galaxy S25")
-# Trace model
-input_shape = torch_model.get_input_spec()
-sample_inputs = torch_model.sample_inputs()
-pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
-# Compile model on a specific device
-compile_job = hub.submit_compile_job(
-    model=pt_model,
-    device=device,
-    input_specs=torch_model.get_input_spec(),
-)
-# Get target model to run on-device
-target_model = compile_job.get_target_model()
-```
-Step 2: **Performance profiling on cloud-hosted device**
-After compiling models from step 1. Models can be profiled model on-device using the
-`target_model`. Note that this scripts runs the model on a device automatically
-provisioned in the cloud.  Once the job is submitted, you can navigate to a
-provided job URL to view a variety of on-device performance metrics.
-```python
-profile_job = hub.submit_profile_job(
-    model=target_model,
-    device=device,
-)
-```
-Step 3: **Verify on-device accuracy**
-To verify the accuracy of the model on-device, you can run on-device inference
-on sample input data on the same cloud hosted device.
-```python
-input_data = torch_model.sample_inputs()
-inference_job = hub.submit_inference_job(
-    model=target_model,
-    device=device,
-    inputs=input_data,
-)
-    on_device_output = inference_job.download_output_data()
-```
-With the output of the model, you can compute like PSNR, relative errors or
-spot check the output with expected output.
-**Note**: This on-device profiling and inference requires access to Qualcomm®
-AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
-## Run demo on a cloud-hosted device
-You can also run the demo on-device.
-```bash
-python -m qai_hub_models.models.sequencer2d.demo --eval-mode on-device
-```
-**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
-environment, please add the following to your cell (instead of the above).
-```
-%run -m qai_hub_models.models.sequencer2d.demo -- --eval-mode on-device
-```
-## Deploying compiled model to Android
-The models can be deployed using multiple runtimes:
-- TensorFlow Lite (`.tflite` export): [This
-  tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
-  guide to deploy the .tflite model in an Android application.
-- QNN (`.so` export ): This [sample
-  app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
-provides instructions on how to use the `.so` shared library  in an Android application.
-## View on Qualcomm® AI Hub
-Get more details on Sequencer2D's performance across various devices [here](https://aihub.qualcomm.com/models/sequencer2d).
-Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
 ## License
 * The license for the original implementation of Sequencer2D can be found
   [here](https://github.com/okojoalg/sequencer/blob/main/LICENSE).
 ## References
 * [Sequencer: Deep LSTM for Image Classification](https://arxiv.org/abs/2205.01972)
 * [Source Model Implementation](https://github.com/okojoalg/sequencer)
 ## 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).

 ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/web-assets/model_demo.png)
+# Sequencer2D: Optimized for Qualcomm Devices
 Sequencer2D is a vision transformer model that can classify images from the Imagenet dataset.
+This is based on the implementation of Sequencer2D found [here](https://github.com/okojoalg/sequencer).
+This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/sequencer2d) 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 2.37, ONNX Runtime 1.23.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/releases/v0.46.1/sequencer2d-onnx-float.zip)
+| ONNX | w8a16 | Universal | QAIRT 2.37, ONNX Runtime 1.23.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/releases/v0.46.1/sequencer2d-onnx-w8a16.zip)
+| QNN_DLC | float | Universal | QAIRT 2.42 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/releases/v0.46.1/sequencer2d-qnn_dlc-float.zip)
+| TFLITE | float | Universal | QAIRT 2.42, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/sequencer2d/releases/v0.46.1/sequencer2d-tflite-float.zip)
+For more device-specific assets and performance metrics, visit **[Sequencer2D on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/sequencer2d)**.
+### Option 2: Export with Custom Configurations
+Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/sequencer2d) 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 [Sequencer2D on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/sequencer2d) for usage instructions.
+## Model Details
+**Model Type:** Model_use_case.image_classification
+**Model Stats:**
+- Model checkpoint: sequencer2d_s
+- Input resolution: 224x224
+- Number of parameters: 27.6M
+- Model size (float): 106 MB
+- Model size (w8a8): 69.1 MB
+## Performance Summary
+| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
+|---|---|---|---|---|---|---
+| Sequencer2D | ONNX | float | Snapdragon® X Elite | 48.401 ms | 66 - 66 MB | NPU
+| Sequencer2D | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 33.466 ms | 1 - 1349 MB | NPU
+| Sequencer2D | ONNX | float | Qualcomm® QCS8550 (Proxy) | 46.558 ms | 0 - 83 MB | NPU
+| Sequencer2D | ONNX | float | Qualcomm® QCS9075 | 57.399 ms | 0 - 4 MB | NPU
+| Sequencer2D | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 23.058 ms | 1 - 766 MB | NPU
+| Sequencer2D | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 17.991 ms | 1 - 807 MB | NPU
+| Sequencer2D | ONNX | w8a16 | Snapdragon® X Elite | 123.498 ms | 132 - 132 MB | NPU
+| Sequencer2D | ONNX | w8a16 | Qualcomm® QCS6490 | 647.033 ms | 46 - 56 MB | CPU
+| Sequencer2D | ONNX | w8a16 | Qualcomm® QCS9075 | 165.125 ms | 107 - 110 MB | NPU
+| Sequencer2D | ONNX | w8a16 | Qualcomm® QCM6690 | 274.09 ms | 47 - 63 MB | CPU
+| Sequencer2D | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 105.431 ms | 106 - 488 MB | NPU
+| Sequencer2D | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 260.838 ms | 26 - 37 MB | CPU
+| Sequencer2D | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 102.361 ms | 50 - 436 MB | NPU
+| Sequencer2D | QNN_DLC | float | Snapdragon® X Elite | 21.419 ms | 1 - 1 MB | NPU
+| Sequencer2D | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 14.292 ms | 0 - 2279 MB | NPU
+| Sequencer2D | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 20.806 ms | 1 - 565 MB | NPU
+| Sequencer2D | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 26.272 ms | 0 - 846 MB | NPU
+| Sequencer2D | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 12.521 ms | 1 - 1065 MB | NPU
+| Sequencer2D | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.34 ms | 0 - 1225 MB | NPU
+| Sequencer2D | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 11.933 ms | 0 - 917 MB | NPU
+| Sequencer2D | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 37.369 ms | 0 - 810 MB | NPU
+| Sequencer2D | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 17.044 ms | 0 - 12 MB | NPU
+| Sequencer2D | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 21.299 ms | 0 - 724 MB | NPU
+| Sequencer2D | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 9.406 ms | 0 - 744 MB | NPU
+| Sequencer2D | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 7.34 ms | 0 - 1082 MB | NPU
 ## License
 * The license for the original implementation of Sequencer2D can be found
   [here](https://github.com/okojoalg/sequencer/blob/main/LICENSE).
 ## References
 * [Sequencer: Deep LSTM for Image Classification](https://arxiv.org/abs/2205.01972)
 * [Source Model Implementation](https://github.com/okojoalg/sequencer)
 ## 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).

Sequencer2D_float.dlc DELETED Viewed

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d06ccb3a689682c1d37163aed322fdd7be4de21cb67a78d38ab219cb562bf0f6
-size 123861260

Sequencer2D_float.onnx.zip DELETED Viewed

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:91447c9fb86fcb3bf1d99ec17934e23c2fb584a0f895702a2ee42d229c314a57
-size 102959885

Sequencer2D_float.tflite DELETED Viewed

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0f1dccc18a5341ce10321f8a6515ca6f6860a33e37f0d7d80d227cb4ea207cb5
-size 114050616

tool-versions.yaml DELETED Viewed

@@ -1,4 +0,0 @@
-tool_versions:
-  onnx:
-    qairt: 2.37.1.250807093845_124904
-    onnx_runtime: 1.23.0