In LatentDiffusionInferer (generative > inferers > LatentDiffusionInferer) there is a bug in the sample function, lines 354 and 361: `image = autoencoder_model.decode_stage_2_outputs(latent) * self.scale_factor` should change to: `image = autoencoder_model.decode_stage_2_outputs(latent * self.scale_factor)`
In LatentDiffusionInferer (generative > inferers > LatentDiffusionInferer) there is a bug in the sample function, lines 354 and 361:
image = autoencoder_model.decode_stage_2_outputs(latent) * self.scale_factorshould change to:
image = autoencoder_model.decode_stage_2_outputs(latent * self.scale_factor)