在Atlas 800训练服务器运行模型库YOLOv5模型训练脚本报错

按照gitgee项目的说明，链接：https://gitee.com/mindspore/models/tree/master/official/cv/YOLOv5#%E6%8E%A8%E7%90%86%E8%BF%87%E7%A8%8B

在快速入门部分，使用以下命令：dyh@train-01:~/models-master/official/cv/YOLOv5$ python3 train.py --data_dir =/home/dyh/data/coco --device_target="Ascend"

报错：

Traceback (most recent call last): 

  File "train.py", line 213, in <module> 

    run_train() 

  File "/home/dyh/models-master/official/cv/YOLOv5/model_utils/moxing_adapter.py", line 167, in wrapped_func 

    run_func(*args, **kwargs) 

  File "train.py", line 134, in run_train 

    network = YOLOV5(is_training=True, version=dict_version[config.yolov5_version]) 

  File "/home/dyh/models-master/official/cv/YOLOv5/src/yolo.py", line 356, in __init__ 

    self.mean = ms.Tensor(np.array([0.485 * 255, 0.456 * 255, 0.406 * 255], 

  File "/usr/local/lib/python3.8/dist-packages/mindspore/common/tensor.py", line 2374, in reshape 

    return tensor_operator_registry.get('reshape')()(self, new_shape) 

  File "/usr/local/lib/python3.8/dist-packages/mindspore/ops/primitive.py", line 296, in __call__ 

    return _run_op(self, self.name, args) 

  File "/usr/local/lib/python3.8/dist-packages/mindspore/common/api.py", line 98, in wrapper 

    results = fn(*arg, **kwargs) 

  File "/usr/local/lib/python3.8/dist-packages/mindspore/ops/primitive.py", line 733, in _run_op 

    output = real_run_op(obj, op_name, args) 

RuntimeError: TbeInitialize running failed. 

****************************************************解答*****************************************************

建议用mindyolo这个套件，modelzoo的很多算法都不更新了。