huawei-noah / Efficient-Computing

Efficient computing methods developed by Huawei Noah's Ark Lab
1.19k stars 208 forks source link

训练报错 #125

Closed lopiyi closed 8 months ago

lopiyi commented 9 months ago

Training start...

 Epoch  iou_loss  dfl_loss  cls_loss

0%| | 0/433 [00:00<?, ?it/s] ERROR in training steps: Default process group has not been initialized, please make sure to call init_process_group. 0%| | 0/433 [00:01<?, ?it/s]
ERROR in training steps. ERROR in training loop or eval/save model. Traceback (most recent call last): File "E:\Pycharm\Gold-YOLO\tools\train.py", line 129, in main(args) File "E:\Pycharm\Gold-YOLO\tools\train.py", line 119, in main trainer.train() File "E:\Pycharm\Gold-YOLO\yolov6\core\engine.py", line 109, in train self.train_in_loop(self.epoch) File "E:\Pycharm\Gold-YOLO\yolov6\core\engine.py", line 127, in train_in_loop self.print_details() File "E:\Pycharm\Gold-YOLO\yolov6\core\engine.py", line 339, in print_details self.mean_loss = (self.mean_loss * self.step + self.loss_items) / (self.step + 1) TypeError: unsupported operand type(s) for +: 'Tensor' and 'NoneType'

以上是完整报错内容,大致定位是ERROR in training steps: Default process group has not been initialized, please make sure to call init_process_group.环境是win11,不知如何修改,感谢作者指点!!

lose4578 commented 8 months ago

You can refer this issue #67