You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A test started to fail in optimum-benchmark. I tested a minimal script and it also fails:
fromoptimum.intel.openvinoimportOVStableDiffusionPipelinemodel="hf-internal-testing/tiny-stable-diffusion-torch"# or a runway modelpipe=OVStableDiffusionPipeline.from_pretrained(model, device="CPU", export=True)
pipe(prompt="prompt", num_inference_steps=1)
The error is AttributeError: 'OVStableDiffusionPipeline' object has no attribute 'image_processor'
The problem comes from the new VaeImageProcessor which was added in optimum but only for ORTStableDiffusionPipelineBase StableDiffusionPipelineMixin that's used by OVStableDiffusionPipeline supposes its existence without instantiating it.
Should it be added in each sub-package ?
The text was updated successfully, but these errors were encountered:
Using main branches of
optimum
andoptimum-intel
.A test started to fail in
optimum-benchmark
. I tested a minimal script and it also fails:The error is
AttributeError: 'OVStableDiffusionPipeline' object has no attribute 'image_processor'
The problem comes from the new
VaeImageProcessor
which was added inoptimum
but only forORTStableDiffusionPipelineBase
StableDiffusionPipelineMixin
that's used byOVStableDiffusionPipeline
supposes its existence without instantiating it.Should it be added in each sub-package ?
The text was updated successfully, but these errors were encountered: