Skip to content

Latest commit

 

History

History
48 lines (29 loc) · 2.36 KB

README.md

File metadata and controls

48 lines (29 loc) · 2.36 KB
Raw

drawing Blender ControlNet Plugin

Blender ControlNet plugin runs on Qualcomm X-Elite NPU using Qualcomm AI Runtime (QAIRT) SDK. The plugin uses QAIRT python APIs to run context binaries (.bin) generated by QAIRT SDK. We provide optimal performance by using QAIRT to run AI models on Qualcomm X-Elite NPU. These models are hosted on Qualcomm AI hub.

controlnet This clip is at 2x speed

Supported features

  • ControlNet model, 512x512 as input image, 512x512 or 2048x2048 as output image
  • Sampling method: DPM++ 2M
  • Upscaling method: ESRGAN-x4

Installation Instructions:

Step 1: Install Dependencies

Download and install Python 3.10.6

NOTE: The program is tested to work on Python 3.10.6. Don't use other versions.

Step 2: Download Blender Arm64

Download Blender Arm64 4.3 Alpha version from Daily build. The Stable version of Arm64 is expected to be releasing soon.

Step 3: Download Blender ControlNet Plugin

Download blender-controlnet-plugin-v1.0.zip from the latest release. Unzip and place the SnapdragonImageGeneration folder under C:\Users\<user>\AppData\Roaming\Blender Foundation\Blender\4.3\scripts\addons

NOTE: The path C:\Users\<user>\AppData\Roaming\Blender Foundation\Blender\4.3 won't be created until the Blender application is opened at least once.

NOTE: Create the scripts\addons folder if not already present.

Step 4: Install Plugin

Run below command to install the plugin. This installs QAIRT SDK and models required for the plugin. This script also creates a python virtual environment for the plugin.

.\install.bat

Step 5: Enable Plugin

Run Blender Arm64 4.3 and go to Edit->Preferences->Add-ons and search for Snapdragon Image Generation and enable it. drawing

Troubleshooting

  • To reinstall from scratch, delete directories: venv, qnn_assets