update code

app.py

@@ -433,15 +433,10 @@ def model_define():
     # Load average latent vector
     ws_avg = torch.load(default_config.ws_avg_pkl).to(device)[0]
 
-    # Set up face verse for amimation
-    base_coff = np.load(
-        'pretrained_model/temp.npy').astype(
-        np.float32)
-    base_coff = torch.from_numpy(base_coff).float()
-    Faceverse = Faceverse_manager(device=device, base_coeff=base_coff)
+
 
     return motion_aware_render_model, sample_steps, DiT_model, \
-        vae_triplane, image_encoder, dinov2, dino_img_processor, clip_image_processor, triplane_std, triplane_mean, ws_avg, Faceverse, device, input_process_model
+        vae_triplane, image_encoder, dinov2, dino_img_processor, clip_image_processor, triplane_std, triplane_mean, ws_avg,  device, input_process_model
 
 
 def duplicate_batch(tensor, batch_size=2):
@@ -482,9 +477,15 @@ def avatar_generation(items, save_path_base, video_path_input, source_type, is_s
     image_encoder.to(device)
     vae_triplane.to(device)
     dinov2.to(device)
-    Faceverse.to(device)
     ws_avg = ws_avg.to(device)
     DiT_model = DiT_model.to(device)
+    # Set up face verse for amimation
+    base_coff = np.load(
+        'pretrained_model/temp.npy').astype(
+        np.float32)
+    base_coff = torch.from_numpy(base_coff).float()
+    Faceverse = Faceverse_manager(device=device, base_coeff=base_coff)
+
     if source_type == 'example':
         input_img_fvid = './demo_data/source_img/img_generate_different_domain/coeffs/demo_imgs'
         input_img_motion = './demo_data/source_img/img_generate_different_domain/motions/demo_imgs'
@@ -999,7 +1000,7 @@ if __name__ == '__main__':
     image_folder = "./demo_data/source_img/img_generate_different_domain/images512x512/demo_imgs"
     example_img_names = os.listdir(image_folder)
     render_model, sample_steps, DiT_model, \
-        vae_triplane, image_encoder, dinov2, dino_img_processor, clip_image_processor, std, mean, ws_avg, Faceverse, device, input_process_model = model_define()
+        vae_triplane, image_encoder, dinov2, dino_img_processor, clip_image_processor, std, mean, ws_avg,  device, input_process_model = model_define()
     controlnet_path = './pretrained_model/control'
     controlnet = ControlNetModel.from_pretrained(
         controlnet_path, torch_dtype=torch.float16