|
Download README.md from qualcomm/SigLIP2: direct link, hf CLI and curl.
- Browser
- Download file 17.9 kB
-
https://huggingface.co/qualcomm/SigLIP2/resolve/main/README.md
- Command line
-
hf download hf://qualcomm/SigLIP2/README.md
-
curl -L -o README.md https://huggingface.co/qualcomm/SigLIP2/resolve/main/README.md
17.9 kB
| library_name: pytorch | |
| license: apache-2.0 | |
| tags: | |
| - foundation | |
| - android | |
| pipeline_tag: image-classification | |
|  | |
| # SigLIP2: Optimized for Qualcomm Devices | |
| SigLIP2 (Sigmoid Loss for Language-Image Pre-training 2) is a vision-language model from Google that computes cosine-similarity scores between images and text prompts. It can be used for zero-shot image classification, image search, and content moderation without any task-specific fine-tuning. | |
| This is based on the implementation of SigLIP2 found [here](https://huggingface.co/google/siglip2-base-patch16-224). | |
| 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/v0.64.0/src/qai_hub_models/models/siglip2) 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 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/siglip2/releases/v0.64.0/siglip2-onnx-float.zip) | |
| | ONNX | w8a16 | Universal | QAIRT 2.50, ONNX Runtime 1.30.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/siglip2/releases/v0.64.0/siglip2-onnx-w8a16.zip) | |
| | QNN_DLC | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/siglip2/releases/v0.64.0/siglip2-qnn_dlc-float.zip) | |
| | QNN_DLC | w8a16 | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/siglip2/releases/v0.64.0/siglip2-qnn_dlc-w8a16.zip) | |
| | TFLITE | float | Universal | QAIRT 2.50 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/siglip2/releases/v0.64.0/siglip2-tflite-float.zip) | |
| For more device-specific assets and performance metrics, visit **[SigLIP2 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/siglip2)**. | |
| ### Option 2: Export with Custom Configurations | |
| Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.64.0/src/qai_hub_models/models/siglip2) 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 [SigLIP2 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.64.0/src/qai_hub_models/models/siglip2) for usage instructions. | |
| ## Model Details | |
| **Model Type:** Model_use_case.image_classification | |
| **Model Stats:** | |
| - Image input resolution: 224x224 | |
| - Model checkpoint: google/siglip2-base-patch16-224 | |
| - Model size (image_encoder) (float): 352 MB | |
| - Model size (image_encoder) (w8a16): 92.8 MB | |
| - Model size (text_encoder) (float): 1.05 GB | |
| - Model size (text_encoder) (w8a16): 461 MB | |
| - Text sequence length: 64 | |
| ## Performance Summary | |
| | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | |
| |---|---|---|---|---|---|--- | |
| | image_encoder | ONNX | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 3.341 ms | 0 - 168 MB | NPU | |
| | image_encoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.188 ms | 0 - 167 MB | NPU | |
| | image_encoder | ONNX | float | Snapdragon® X2 Elite | 3.518 ms | 2 - 2 MB | NPU | |
| | image_encoder | ONNX | float | Snapdragon® X Elite | 8.692 ms | 182 - 182 MB | NPU | |
| | image_encoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 5.805 ms | 0 - 302 MB | NPU | |
| | image_encoder | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 12.863 ms | 1 - 294 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 11.913 ms | 0 - 5 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 8.358 ms | 0 - 572 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® QCS8450 | 12.863 ms | 1 - 294 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 11.329 ms | 1 - 4 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 8.692 ms | 182 - 182 MB | NPU | |
| | image_encoder | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 4.188 ms | 0 - 167 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 2.124 ms | 0 - 194 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 2.852 ms | 0 - 189 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® X2 Elite | 2.354 ms | 1 - 1 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® X Elite | 5.558 ms | 91 - 91 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 3.672 ms | 0 - 295 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® 8 Gen 1 Mobile | 6.995 ms | 0 - 284 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS6490 | 18.544 ms | 3 - 6 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 5.099 ms | 0 - 5 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 5.301 ms | 0 - 98 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® QCS8450 | 6.995 ms | 0 - 284 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 5.264 ms | 0 - 4 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 5.558 ms | 91 - 91 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 53.881 ms | 0 - 374 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 7.025 ms | 0 - 259 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 2.852 ms | 0 - 189 MB | NPU | |
| | image_encoder | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 7.025 ms | 0 - 259 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 3.42 ms | 1 - 164 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.228 ms | 0 - 162 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® X2 Elite | 3.962 ms | 1 - 1 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® X Elite | 9.208 ms | 1 - 1 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 5.963 ms | 1 - 253 MB | NPU | |
| | image_encoder | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 13.297 ms | 1 - 246 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-8275 | 12.005 ms | 1 - 4 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 8.492 ms | 1 - 3 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® SA8775P | 11.317 ms | 1 - 163 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® SA8650P | 11.317 ms | 1 - 163 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® SA8255P | 11.317 ms | 1 - 163 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® QCS8450 | 13.297 ms | 1 - 246 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 11.146 ms | 3 - 5 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 9.208 ms | 1 - 1 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 4.228 ms | 0 - 162 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® SA7255P | 36.956 ms | 1 - 162 MB | NPU | |
| | image_encoder | QNN_DLC | float | Qualcomm® SA8295P | 13.957 ms | 1 - 154 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 2.466 ms | 0 - 184 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 3.319 ms | 0 - 180 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 2.977 ms | 0 - 0 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® X Elite | 6.858 ms | 0 - 0 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 4.273 ms | 0 - 285 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 5.82 ms | 0 - 4 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 6.307 ms | 0 - 2 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® SA8775P | 6.419 ms | 0 - 231 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® SA8650P | 6.419 ms | 0 - 231 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® SA8255P | 6.419 ms | 0 - 231 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 6.226 ms | 2 - 5 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 6.858 ms | 0 - 0 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 66.246 ms | 0 - 361 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 7.733 ms | 0 - 264 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 3.319 ms | 0 - 180 MB | NPU | |
| | image_encoder | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 7.733 ms | 0 - 264 MB | NPU | |
| | image_encoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 3.313 ms | 0 - 165 MB | NPU | |
| | image_encoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.127 ms | 0 - 165 MB | NPU | |
| | image_encoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.876 ms | 0 - 252 MB | NPU | |
| | image_encoder | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 12.969 ms | 0 - 245 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 11.943 ms | 0 - 185 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 8.347 ms | 0 - 2 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® QCS8450 | 12.969 ms | 0 - 245 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 11.308 ms | 0 - 184 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 4.127 ms | 0 - 165 MB | NPU | |
| | image_encoder | TFLITE | float | Qualcomm® SA8295P | 13.73 ms | 0 - 151 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 2.405 ms | 0 - 143 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.454 ms | 0 - 142 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® X2 Elite | 1.351 ms | 1 - 1 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® X Elite | 2.887 ms | 542 - 542 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 2.728 ms | 0 - 256 MB | NPU | |
| | text_encoder | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 4.087 ms | 0 - 249 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 3.776 ms | 0 - 4 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 3.523 ms | 0 - 643 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® QCS8450 | 4.087 ms | 0 - 249 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 3.514 ms | 0 - 3 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 2.887 ms | 542 - 542 MB | NPU | |
| | text_encoder | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 2.454 ms | 0 - 142 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 1.235 ms | 0 - 191 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 1.244 ms | 0 - 190 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® X2 Elite | 0.747 ms | 1 - 1 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® X Elite | 1.69 ms | 461 - 461 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 1.464 ms | 0 - 276 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® 8 Gen 1 Mobile | 2.238 ms | 0 - 279 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS6490 | 6.331 ms | 0 - 3 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 2.074 ms | 0 - 4 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.846 ms | 0 - 515 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® QCS8450 | 2.238 ms | 0 - 279 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 2.192 ms | 0 - 3 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.69 ms | 461 - 461 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 13.182 ms | 0 - 247 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 2.852 ms | 0 - 242 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 1.244 ms | 0 - 190 MB | NPU | |
| | text_encoder | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 2.852 ms | 0 - 242 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 2.395 ms | 0 - 140 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.478 ms | 0 - 139 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® X2 Elite | 1.705 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® X Elite | 3.053 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 2.757 ms | 0 - 253 MB | NPU | |
| | text_encoder | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 4.065 ms | 0 - 243 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-8275 | 3.74 ms | 0 - 3 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 3.516 ms | 0 - 627 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® SA8775P | 3.642 ms | 0 - 138 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® SA8650P | 3.642 ms | 0 - 138 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® SA8255P | 3.642 ms | 0 - 138 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® QCS8450 | 4.065 ms | 0 - 243 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 3.53 ms | 2 - 4 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 3.053 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 2.478 ms | 0 - 139 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® SA7255P | 9.998 ms | 0 - 137 MB | NPU | |
| | text_encoder | QNN_DLC | float | Qualcomm® SA8295P | 4.506 ms | 0 - 128 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 1.243 ms | 0 - 188 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 1.25 ms | 0 - 186 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 0.975 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® X Elite | 1.972 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 1.452 ms | 0 - 269 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-8275 | 2.059 ms | 0 - 3 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.895 ms | 0 - 82 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® SA8775P | 2.198 ms | 0 - 232 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® SA8650P | 2.198 ms | 0 - 232 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® SA8255P | 2.198 ms | 0 - 232 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-9075 | 2.047 ms | 2 - 4 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ IQ-X7181 | 1.972 ms | 0 - 0 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-6690 | 13.538 ms | 0 - 247 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-7790 | 2.96 ms | 0 - 241 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Qualcomm® Dragonwing™ Q-8750 | 1.25 ms | 0 - 186 MB | NPU | |
| | text_encoder | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 2.96 ms | 0 - 241 MB | NPU | |
| | text_encoder | TFLITE | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 2.408 ms | 0 - 151 MB | NPU | |
| | text_encoder | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.417 ms | 0 - 145 MB | NPU | |
| | text_encoder | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 2.743 ms | 0 - 260 MB | NPU | |
| | text_encoder | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 4.082 ms | 0 - 245 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 3.776 ms | 0 - 544 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 3.536 ms | 0 - 2 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® QCS8450 | 4.082 ms | 0 - 245 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 3.537 ms | 0 - 543 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 2.417 ms | 0 - 145 MB | NPU | |
| | text_encoder | TFLITE | float | Qualcomm® SA8295P | 4.547 ms | 0 - 131 MB | NPU | |
| ## License | |
| * The license for the original implementation of SigLIP2 can be found | |
| [here](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/apache-2.0.md). | |
| ## References | |
| * [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://arxiv.org/abs/2502.14786) | |
| * [Source Model Implementation](https://huggingface.co/google/siglip2-base-patch16-224) | |
| ## 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). | |