使用paddlex_restful中的api接口训练发生报错如下：

This log file path is /home/zksc/paddlex_workspace/projects/P0076/T0287/err.log
注意：标志为WARNING/INFO类的仅为警告或提示类信息，非错误信息
2022-11-28 10:39:25,269-WARNING: type object 'QuantizationTransformPass' has no attribute '_supported_quantizable_op_type'
2022-11-28 10:39:25,269-WARNING: If you want to use training-aware and post-training quantization, please use Paddle >= 1.8.4 or develop version
Warning: import ppdet from source directory without installing, run 'python setup.py install' to install ppdet firstly
Process Process-3:
Traceback (most recent call last):
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex_restful/restful/project/operate.py", line 93, in _call_paddlex_train
train(task_path, dataset_path, params['train'])
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex_restful/restful/project/train/detection.py", line 224, in train
model = detector(num_classes=num_classes, backbone=params.backbone)
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex/cv/models/detector.py", line 972, in __init__
backbone = self._get_backbone('DarkNet', norm_type=norm_type)
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex/cv/models/detector.py", line 101, in _get_backbone
backbone = getattr(ppdet.modeling, backbone_name)(**params)
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex/ppdet/modeling/backbones/darknet.py", line 275, in __init__
self.conv0 = ConvBNLayer(
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddlex-2.1.0-py3.9.egg/paddlex/ppdet/modeling/backbones/darknet.py", line 58, in __init__
self.conv = nn.Conv2D(
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/nn/layer/conv.py", line 644, in __init__
super(Conv2D, self).__init__(
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/nn/layer/conv.py", line 133, in __init__
self.weight = self.create_parameter(
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/fluid/dygraph/layers.py", line 423, in create_parameter
return self._helper.create_parameter(temp_attr, shape, dtype, is_bias,
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/fluid/layer_helper_base.py", line 376, in create_parameter
return self.main_program.global_block().create_parameter(
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/fluid/framework.py", line 3572, in create_parameter
initializer(param, self)
File "/home/zksc/anaconda3/envs/paddleX/lib/python3.9/site-packages/paddle/fluid/initializer.py", line 365, in __call__
out_var = _C_ops.gaussian_random(
OSError: (External) CUDA error(3), initialization error.
[Hint: 'cudaErrorInitializationError'. The API call failed because the CUDA driver and runtime could not be initialized. ] (at /paddle/paddle/phi/backends/gpu/cuda/cuda_info.cc:243)
[operator < gaussian_random > error]

尝试在paddlex_restful下的command启动文件中添加mp.set_start_method("spwan")，但是没有效果

求助这个问题应该怎么解决？