Update app.py

app.py

@@ -91,7 +91,6 @@ def generate(segment, image, prompt, size, guidance_scale, num_inference_steps,
         prompt_embeds, prompt_attention_mask, negative_prompt_embeds, negative_attention_mask, pooled_prompt_embeds, negative_pooled_prompt_embeds = pipe.encode_prompt(
         prompt=prompt, do_classifier_free_guidance=True, negative_prompt=negative_prompt, device=device
         )
-        
         transformer_dtype = pipe.transformer.dtype
         prompt_embeds = prompt_embeds.to(transformer_dtype)
         prompt_attention_mask = prompt_attention_mask.to(transformer_dtype)
@@ -102,7 +101,6 @@ def generate(segment, image, prompt, size, guidance_scale, num_inference_steps,
         prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
         prompt_attention_mask = torch.cat([negative_attention_mask, prompt_attention_mask])
         pooled_prompt_embeds = torch.cat([negative_pooled_prompt_embeds, pooled_prompt_embeds])
-                
         pipe.scheduler.set_timesteps(num_inference_steps, device=device)
         timesteps = pipe.scheduler.timesteps
         all_timesteps_cpu = timesteps.cpu()
@@ -110,6 +108,8 @@ def generate(segment, image, prompt, size, guidance_scale, num_inference_steps,
         segment_timesteps = torch.from_numpy(timesteps_split_np[0]).to("cuda")
         num_channels_latents = pipe.transformer.config.in_channels
         num_channels_latents = int(num_channels_latents / 2)
+        image = Image.open(image.name).convert('RGB')
+        image.resize((size,size), Image.LANCZOS)
         image = pipe.video_processor.preprocess(image, height=size, width=size).to(
                 device, dtype=prompt_embeds.dtype
         )