--- library_name: pytorch license: other tags: - generative_ai - android pipeline_tag: unconditional-image-generation --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/stable_diffusion_v1_5/web-assets/model_demo.png) # Stable-Diffusion-v1.5: Optimized for Mobile Deployment ## State-of-the-art generative AI model used to generate detailed images conditioned on text descriptions Generates high resolution images from text prompts using a latent diffusion model. This model uses CLIP ViT-L/14 as text encoder, U-Net based latent denoising, and VAE based decoder to generate the final image. This model is an implementation of Stable-Diffusion-v1.5 found [here](https://github.com/CompVis/stable-diffusion/tree/main). This repository provides scripts to run Stable-Diffusion-v1.5 on Qualcomm® devices. More details on model performance across various devices, can be found [here](https://aihub.qualcomm.com/models/stable_diffusion_v1_5). ### Model Details - **Model Type:** Model_use_case.image_generation - **Model Stats:** - Input: Text prompt to generate image - Text Encoder Number of parameters: 340M - UNet Number of parameters: 865M - VAE Decoder Number of parameters: 83M - Model size: 1GB | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model |---|---|---|---|---|---|---|---|---| | text_encoder | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 11.111 ms | 0 - 9 MB | NPU | Use Export Script | | text_encoder | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 5.363 ms | 0 - 2 MB | NPU | Use Export Script | | text_encoder | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 5.933 ms | 0 - 10 MB | NPU | Use Export Script | | text_encoder | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN | 11.111 ms | 0 - 9 MB | NPU | Use Export Script | | text_encoder | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 5.396 ms | 0 - 3 MB | NPU | Use Export Script | | text_encoder | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 5.398 ms | 0 - 3 MB | NPU | Use Export Script | | text_encoder | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN | 5.933 ms | 0 - 10 MB | NPU | Use Export Script | | text_encoder | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 5.413 ms | 0 - 2 MB | NPU | Use Export Script | | text_encoder | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 5.773 ms | 0 - 162 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | text_encoder | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 3.909 ms | 0 - 18 MB | NPU | Use Export Script | | text_encoder | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 4.097 ms | 0 - 19 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | text_encoder | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 3.501 ms | 0 - 14 MB | NPU | Use Export Script | | text_encoder | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 3.754 ms | 0 - 14 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | text_encoder | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 5.836 ms | 0 - 0 MB | NPU | Use Export Script | | text_encoder | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 5.977 ms | 157 - 157 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | unet | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 263.602 ms | 0 - 7 MB | NPU | Use Export Script | | unet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 110.81 ms | 0 - 2 MB | NPU | Use Export Script | | unet | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 105.273 ms | 0 - 8 MB | NPU | Use Export Script | | unet | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN | 263.602 ms | 0 - 7 MB | NPU | Use Export Script | | unet | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 111.393 ms | 0 - 7 MB | NPU | Use Export Script | | unet | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 111.635 ms | 0 - 3 MB | NPU | Use Export Script | | unet | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN | 105.273 ms | 0 - 8 MB | NPU | Use Export Script | | unet | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 111.871 ms | 0 - 3 MB | NPU | Use Export Script | | unet | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 111.763 ms | 0 - 898 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | unet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 78.019 ms | 0 - 18 MB | NPU | Use Export Script | | unet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 80.045 ms | 0 - 15 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | unet | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 67.577 ms | 0 - 14 MB | NPU | Use Export Script | | unet | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 67.421 ms | 0 - 15 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | unet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 113.221 ms | 0 - 0 MB | NPU | Use Export Script | | unet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 114.344 ms | 842 - 842 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | vae | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 718.731 ms | 0 - 9 MB | NPU | Use Export Script | | vae | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 270.917 ms | 0 - 4 MB | NPU | Use Export Script | | vae | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 248.917 ms | 0 - 10 MB | NPU | Use Export Script | | vae | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN | 718.731 ms | 0 - 9 MB | NPU | Use Export Script | | vae | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 270.775 ms | 0 - 3 MB | NPU | Use Export Script | | vae | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 270.064 ms | 0 - 3 MB | NPU | Use Export Script | | vae | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN | 248.917 ms | 0 - 10 MB | NPU | Use Export Script | | vae | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 271.182 ms | 0 - 2 MB | NPU | Use Export Script | | vae | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 269.668 ms | 0 - 66 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | vae | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 207.636 ms | 0 - 18 MB | NPU | Use Export Script | | vae | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 207.55 ms | 3 - 22 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | vae | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 175.815 ms | 0 - 14 MB | NPU | Use Export Script | | vae | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 194.632 ms | 3 - 17 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | | vae | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 265.394 ms | 0 - 0 MB | NPU | Use Export Script | | vae | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 265.427 ms | 63 - 63 MB | NPU | [Stable-Diffusion-v1.5.onnx](https://huggingface.co/qualcomm/Stable-Diffusion-v1.5/blob/main/Stable-Diffusion-v1.5_w8a16.onnx) | ## Deploy to Snapdragon X Elite NPU Please follow the [Stable Diffusion Windows App](https://github.com/quic/ai-hub-apps/tree/main/apps/windows/python/StableDiffusion) tutorial to quantize model with custom weights. ## Quantize and Deploy Your Own Fine-Tuned Stable Diffusion Please follow the [Quantize Stable Diffusion]({REPOSITORY_URL}/tutorials/stable_diffusion/quantize_stable_diffusion.md) tutorial to quantize model with custom weights. ## Installation Install the package via pip: ```bash pip install "qai-hub-models[stable-diffusion-v1-5]" ``` ## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device Sign-in to [Qualcomm® AI Hub](https://app.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://app.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.stable_diffusion_v1_5.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.stable_diffusion_v1_5.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.stable_diffusion_v1_5.export ``` ``` Profiling Results ------------------------------------------------------------ text_encoder Device : cs_8275 (ANDROID 14) Runtime : QNN Estimated inference time (ms) : 11.1 Estimated peak memory usage (MB): [0, 9] Total # Ops : 437 Compute Unit(s) : npu (437 ops) gpu (0 ops) cpu (0 ops) ------------------------------------------------------------ unet Device : cs_8275 (ANDROID 14) Runtime : QNN Estimated inference time (ms) : 263.6 Estimated peak memory usage (MB): [0, 7] Total # Ops : 4041 Compute Unit(s) : npu (4041 ops) gpu (0 ops) cpu (0 ops) ------------------------------------------------------------ vae Device : cs_8275 (ANDROID 14) Runtime : QNN Estimated inference time (ms) : 718.7 Estimated peak memory usage (MB): [0, 9] Total # Ops : 173 Compute Unit(s) : npu (173 ops) gpu (0 ops) cpu (0 ops) ``` ## 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 Stable-Diffusion-v1.5's performance across various devices [here](https://aihub.qualcomm.com/models/stable_diffusion_v1_5). Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/) ## License * The license for the original implementation of Stable-Diffusion-v1.5 can be found [here](https://github.com/CompVis/stable-diffusion/blob/main/LICENSE). * The license for the compiled assets for on-device deployment can be found [here](https://github.com/CompVis/stable-diffusion/blob/main/LICENSE) ## References * [High-Resolution Image Synthesis with Latent Diffusion Models](https://arxiv.org/abs/2112.10752) * [Source Model Implementation](https://github.com/CompVis/stable-diffusion/tree/main) ## 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).