stage2 训练改batch

brep2sdf/train.py

@@ -443,37 +443,63 @@ class Trainer:
 
                 # 累积损失
                 losses.append(loss)
-                if epoch % 1 == 0:
+                if epoch % 100 == 0:
                     loss_detailss.append(loss_details)
 
-            
+                if len(losses) % 30 == 0:
-
+                    # 多个损失平均后反向传播
-            # 多个损失平均后反向传播
+                    loss_tensor = torch.stack(losses)
-            loss_tensor = torch.stack(losses)
+                    mean_loss = (loss_tensor * weights).sum()
-            mean_loss = (loss_tensor * weights).sum()
+                    mean_loss.backward()
-            mean_loss.backward()
 
-            # 更新参数
+                    # 更新参数
-            self.scheduler.optimizer.step()
+                    self.scheduler.optimizer.step()
-            self.scheduler.step(mean_loss, epoch)
+                    self.scheduler.step(mean_loss, epoch)
 
-            # 清空梯度
+                    # 清空梯度
-            self.scheduler.optimizer.zero_grad()
+                    self.scheduler.optimizer.zero_grad()
+                    torch.cuda.empty_cache()
+                
+                    # 如果你想查看详细的损失信息，可以在这里添加日志记录
+                    if epoch % 100 == 0:
+                        logger.info(f'Train [Stage 2] Epoch: {epoch:4d}\t'
+                                    f'Loss: {loss:.6f}')
+                        loss_details_tensor = torch.stack(loss_detailss)  # shape: [num_patches, 5]
+
+                        # 对每个子项取加权平均（如果需要 weights）
+                        weighted_avg = (loss_details_tensor * weights.view(-1, 1)).sum(dim=0) / weights.sum()
+
+                        subloss_names = ["manifold", "normals", "eikonal", "offsurface", "psdf"]
+                        logger.info("  ".join([f"{name}: {weighted_avg[i].item():.6f}" for i, name in enumerate(subloss_names)]))
+                    losses = []
+                    loss_detailss = []
+
+            if len(losses) > 0:
+                # 多个损失平均后反向传播
+                loss_tensor = torch.stack(losses)
+                mean_loss = (loss_tensor * weights).sum()
+                mean_loss.backward()
+
+                # 更新参数
+                self.scheduler.optimizer.step()
+                self.scheduler.step(mean_loss, epoch)
+
+                # 清空梯度
+                self.scheduler.optimizer.zero_grad()
+                torch.cuda.empty_cache()
+            
+                # 如果你想查看详细的损失信息，可以在这里添加日志记录
+                if epoch % 100 == 0:
+                    logger.info(f'Train [Stage 2] Epoch: {epoch:4d}\t'
+                                f'Loss: {loss:.6f}')
+                    loss_details_tensor = torch.stack(loss_detailss)  # shape: [num_patches, 5]
 
-            # 清理缓存
+                    # 对每个子项取加权平均（如果需要 weights）
-            torch.cuda.empty_cache()
+                    weighted_avg = (loss_details_tensor * weights.view(-1, 1)).sum(dim=0) / weights.sum()
-        
-            # 如果你想查看详细的损失信息，可以在这里添加日志记录
-            if epoch % 1 == 0:
-                logger.info(f'Train [Stage 2] Epoch: {epoch:4d}\t'
-                            f'Loss: {loss:.6f}')
-                loss_details_tensor = torch.stack(loss_detailss)  # shape: [num_patches, 5]
 
-                # 对每个子项取加权平均（如果需要 weights）
+                    subloss_names = ["manifold", "normals", "eikonal", "offsurface", "psdf"]
-                weighted_avg = (loss_details_tensor * weights.view(-1, 1)).sum(dim=0) / weights.sum()
+                    logger.info("  ".join([f"{name}: {weighted_avg[i].item():.6f}" for i, name in enumerate(subloss_names)]))
 
-                subloss_names = ["manifold", "normals", "eikonal", "offsurface", "psdf"]
-                logger.info("  ".join([f"{name}: {weighted_avg[i].item():.6f}" for i, name in enumerate(subloss_names)]))
 
         avg_loss = sum(losses) / len(losses)
         logger.info(f"Total Loss: {total_loss:.6f} | Avg Loss: {avg_loss:.6f}")