From b2c3389e9c6375d9081625e75a99de98395f8e77 Mon Sep 17 00:00:00 2001 From: Volpeon Date: Tue, 1 Nov 2022 16:19:01 +0100 Subject: Update --- pipelines/stable_diffusion/vlpn_stable_diffusion.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'pipelines/stable_diffusion') diff --git a/pipelines/stable_diffusion/vlpn_stable_diffusion.py b/pipelines/stable_diffusion/vlpn_stable_diffusion.py index fc12355..cd5ae7e 100644 --- a/pipelines/stable_diffusion/vlpn_stable_diffusion.py +++ b/pipelines/stable_diffusion/vlpn_stable_diffusion.py @@ -203,6 +203,7 @@ class VlpnStableDiffusion(DiffusionPipeline): # However this currently doesn't work in `mps`. latents_dtype = text_embeddings.dtype latents_shape = (batch_size, self.unet.in_channels, height // 8, width // 8) + if latents is None: if self.device.type == "mps": # randn does not exist on mps @@ -264,7 +265,7 @@ class VlpnStableDiffusion(DiffusionPipeline): for i, t in enumerate(self.progress_bar(timesteps_tensor)): # expand the latents if we are doing classifier free guidance latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents - latent_model_input = self.scheduler.scale_model_input(latent_model_input, t, i) + latent_model_input = self.scheduler.scale_model_input(latent_model_input, t) # predict the noise residual noise_pred = self.unet(latent_model_input, t, encoder_hidden_states=text_embeddings).sample @@ -275,7 +276,7 @@ class VlpnStableDiffusion(DiffusionPipeline): noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond) # compute the previous noisy sample x_t -> x_t-1 - latents = self.scheduler.step(noise_pred, t, i, latents, **extra_step_kwargs).prev_sample + latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs).prev_sample # scale and decode the image latents with vae latents = 1 / 0.18215 * latents -- cgit v1.2.3-54-g00ecf