Fixed holes at 1024

Maybe fixed memory issue -> to confirmed

nodes.py

@@ -291,11 +291,7 @@ class Trellis2SimplifyMesh:
         elif method=="Meshlib":
             mesh.simplify_with_meshlib(target = target_face_num)
         else:
-            raise Exception("Unknown simplification method")
-            
-        mm.soft_empty_cache()
-        torch.cuda.empty_cache()
-        gc.collect()              
+            raise Exception("Unknown simplification method")             
         
         return (mesh,)          
         
@@ -425,8 +421,6 @@ class Trellis2PostProcessMesh:
         mesh.faces = new_faces.to(mesh.device) 
         
         del cumesh
-        mm.soft_empty_cache()
-        torch.cuda.empty_cache()
         gc.collect()          
                 
         return (mesh,)
@@ -443,6 +437,7 @@ class Trellis2UnWrapAndRasterizer:
                 "mesh_cluster_smooth_strength": ("INT",{"default":1}),                
                 "texture_size": ("INT",{"default":1024, "min":512, "max":16384}),
                 "texture_alpha_mode": (["OPAQUE","MASK","BLEND"],{"default":"OPAQUE"}),
+                "double_side_material": ("BOOLEAN",{"default":True}),
             },
         }
 
@@ -452,7 +447,7 @@ class Trellis2UnWrapAndRasterizer:
     CATEGORY = "Trellis2Wrapper"
     OUTPUT_NODE = True
 
-    def process(self, mesh, mesh_cluster_threshold_cone_half_angle_rad, mesh_cluster_refine_iterations, mesh_cluster_global_iterations, mesh_cluster_smooth_strength, texture_size, texture_alpha_mode):
+    def process(self, mesh, mesh_cluster_threshold_cone_half_angle_rad, mesh_cluster_refine_iterations, mesh_cluster_global_iterations, mesh_cluster_smooth_strength, texture_size, texture_alpha_mode, double_side_material):
         aabb = [[-0.5, -0.5, -0.5], [0.5, 0.5, 0.5]]
         
         vertices = mesh.vertices
@@ -587,7 +582,7 @@ class Trellis2UnWrapAndRasterizer:
             metallicFactor=1.0,
             roughnessFactor=1.0,
             alphaMode=alpha_mode,
-            #doubleSided=True if not remesh else False,
+            doubleSided=double_side_material,
         )        
         
         vertices_np = out_vertices.cpu().numpy()
@@ -609,8 +604,6 @@ class Trellis2UnWrapAndRasterizer:
         )   
 
         del cumesh
-        mm.soft_empty_cache()
-        torch.cuda.empty_cache()
         gc.collect()          
                 
         return (textured_mesh,)
@@ -782,7 +775,6 @@ class Trellis2PostProcessAndUnWrapAndRasterizer:
             else:
                 resolution = int(dual_contouring_resolution)
             
-            print(f"Dual Contouring resolution: {resolution}")
             # Perform Dual Contouring remeshing (rebuilds topology)
             cumesh.init(*CuMesh.remeshing.remesh_narrow_band_dc(
                 vertices, faces,
@@ -919,8 +911,6 @@ class Trellis2PostProcessAndUnWrapAndRasterizer:
         )        
         
         del cumesh
-        mm.soft_empty_cache()
-        torch.cuda.empty_cache()
         gc.collect()          
         
         return (textured_mesh,)    
@@ -1033,8 +1023,6 @@ class Trellis2Remesh:
         mesh.faces = new_faces.to(mesh.device) 
         
         del cumesh
-        mm.soft_empty_cache()
-        torch.cuda.empty_cache()
         gc.collect()          
                 
         return (mesh,)        

trellis2/pipelines/trellis2_image_to_3d.py

@@ -83,11 +83,18 @@ class Trellis2ImageTo3DPipeline(Pipeline):
 
         Args:
             path (str): The path to the model. Can be either local path or a Hugging Face repository.
-        """
+        """        
         pipeline = super(Trellis2ImageTo3DPipeline, Trellis2ImageTo3DPipeline).from_pretrained(path)
         new_pipeline = Trellis2ImageTo3DPipeline()
         new_pipeline.__dict__ = pipeline.__dict__
         args = pipeline._pretrained_args
+        
+        # if os.name=='nt':
+            # args['models']['sparse_structure_decoder'] = os.path.join(folder_paths.models_dir,"microsoft","TRELLIS-image-large","ckpts","ss_dec_conv3d_16l8_fp16")
+        # else:
+            # args['models']['sparse_structure_decoder'] = os.path.join("models","microsoft","TRELLIS-image-large","ckpts","ss_dec_conv3d_16l8_fp16")          
+        
+        # print(f"Sparse Structure Decoder: {args['models']['sparse_structure_decoder']}")
 
         new_pipeline.sparse_structure_sampler = getattr(samplers, args['sparse_structure_sampler']['name'])(**args['sparse_structure_sampler']['args'])
         new_pipeline.sparse_structure_sampler_params = args['sparse_structure_sampler']['params']
@@ -550,11 +557,13 @@ class Trellis2ImageTo3DPipeline(Pipeline):
         torch.manual_seed(seed)
         cond_512 = self.get_cond([image], 512)
         cond_1024 = self.get_cond([image], 1024) if pipeline_type != '512' else None
-        ss_res = {'512': 32, '1024': 64, '1024_cascade': 32, '1536_cascade': 32}[pipeline_type]
+        ss_res = {'512': 32, '1024': 32, '1024_cascade': 32, '1536_cascade': 32}[pipeline_type]
+        
         coords = self.sample_sparse_structure(
             cond_512, ss_res,
             num_samples, sparse_structure_sampler_params
-        )
+        )            
+        
         if pipeline_type == '512':
             shape_slat = self.sample_shape_slat(
                 cond_512, self.models['shape_slat_flow_model_512'],