.

utils.py

@@ -84,6 +84,8 @@ def sample_frame(model: LatentDiffusion, prompt: str, image_sequence: torch.Tens
         #                                 unconditional_guidance_scale=5.0,
         #                                 unconditional_conditioning=uc,
         #                                 eta=0)
+
+        print ('dfsf1')
         if False and DEBUG:
             print ('samples_ddim.shape', samples_ddim.shape)
             x_samples_ddim = samples_ddim[:, :3]
@@ -93,12 +95,14 @@ def sample_frame(model: LatentDiffusion, prompt: str, image_sequence: torch.Tens
             #x_samples_ddim = torch.zeros((1, 3, 384, 512))
             #x_samples_ddim[:, :, 128:128+48, 160:160+64] = samples_ddim[:, :3]
         else:
+            print ('dfsf2')
             data_mean = -0.54
             data_std = 6.78
             data_min = -27.681446075439453
             data_max = 30.854148864746094
             x_samples_ddim = samples_ddim * data_std + data_mean
             x_samples_ddim = model.decode_first_stage(x_samples_ddim)
+        print ('dfsf3')
         #x_samples_ddim = pos_map.to(c['c_concat'].device).unsqueeze(0).expand(-1, 3, -1, -1)
         #x_samples_ddim = model.decode_first_stage(x_samples_ddim)
         #x_samples_ddim = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0)