search

intel / openvino-ai-plugins-gimp

GIMP AI plugins with OpenVINO Backend
Apache License 2.0
447 stars 49 forks source link

Load SD model failed #50

Closed makejiang closed 12 months ago

makejiang commented 1 year ago

My test environment: i7-12700M+A370M, Ubuntu 22.04.2, Kernel 5.19.0.32-generic, Python 3.9.17

/home/asd/Downloads/gimpenv3/lib/python3.9/site-packages/gimpopenvino/tools/openvino_common/models_ov/stable_diffusion_engine_NEW.py:22: FutureWarning: Importing DiffusionPipeline or ImagePipelineOutput from diffusers.pipeline_utils is deprecated. Please import from diffusers.pipelines.pipeline_utils instead. from diffusers.pipeline_utils import DiffusionPipeline [ INFO ] Model Name: SD_1.5_square [ INFO ] Initializing Inference Engine... [ INFO ] Model Path: /home/asd/openvino-ai-plugins-gimp/weights/stable-diffusion-ov/stable-diffusion-1.5/square [ INFO ] device_name: ['GPU.1', 'GPU.1', 'GPU.1'] Looking for GIMP process Text Device: GPU.1 Done looking for GIMP process no gimp process found! unet Device: GPU.1 Vae Device: GPU.1 Exception in thread Thread-1: Traceback (most recent call last): File "/usr/lib/python3.9/threading.py", line 980, in _bootstrap_inner self.run() File "/usr/lib/python3.9/threading.py", line 917, in run self._target(*self._args, **self._kwargs) File "/home/asd/Downloads/gimpenv3/lib/python3.9/site-packages/gimpopenvino/plugins/stable-diffusion-ov/../../tools/stable-diffusion-ov-server.py", line 102, in run engine = StableDiffusionEngine( File "/home/asd/Downloads/gimpenv3/lib/python3.9/site-packages/gimpopenvino/tools/openvino_common/models_ov/stable_diffusion_engine_NEW.py", line 127, in init self.vae_encoder = self.core.compile_model(os.path.join(model, "vae_encoder.xml"), device[2]) File "/home/asd/Downloads/gimpenv3/lib/python3.9/site-packages/openvino/runtime/ie_api.py", line 399, in compile_model super().compile_model(model, device_name, {} if config is None else config), RuntimeError: cldnn program build failed! Kernel mvn_gpu_bfyx_opt_248_0_15 not found in the kernel cache!

Other features like SR, StypeTransfer and semantic segmentation is good on the dGPU. And SD works on iGPU and CPU. Is A370M not good enough to support SD?

makejiang commented 12 months ago

set VAE to CPU or iGPU can bypass this issue.