请教一下各位大佬

已经安装好CUDA，cuDNN，paddlepaddle_gpu

用这行代码os.environ['CUDA_VISIBLE_DEVICES'] = '0'训练，但是打开任务管理器GPU0使用率是1%，CPU100%

"""
训练目标检测模型

"""
import os
# 选择使用0号卡
os.environ['CUDA_VISIBLE_DEVICES'] = '0'
from paddlex.det import transforms
import paddlex as pdx


# 定义训练和验证时的transforms
train_transforms = transforms.Compose([
    transforms.RandomHorizontalFlip(),
    transforms.Normalize(),
    transforms.ResizeByShort(short_size=600, max_size=1333),
    transforms.Padding(coarsest_stride=32)
])

eval_transforms = transforms.Compose([
    transforms.Normalize(),
    transforms.ResizeByShort(short_size=600, max_size=1333),
    transforms.Padding(coarsest_stride=32),
])

# 定义训练和验证所用的数据集
train_dataset = pdx.datasets.VOCDetection(
    data_dir='sample_det',
    file_list='sample_det/train_list.txt',
    label_list='sample_det/labels.txt',
    transforms=train_transforms,
    shuffle=True)
eval_dataset = pdx.datasets.VOCDetection(
    data_dir='sample_det',
    file_list='sample_det/val_list.txt',
    label_list='sample_det/labels.txt',
    transforms=eval_transforms)

# 初始化模型，并进行训练
# 可使用VisualDL查看训练指标
# VisualDL启动方式: visualdl --logdir output/faster_rcnn_r50_fpn/vdl_log --port 8001
# 浏览器打开 https://0.0.0.0:8001即可
# 其中0.0.0.0为本机访问，如为远程服务, 改成相应机器IP
# num_classes 需要设置为包含背景类的类别数，即: 目标类别数量 + 1
num_classes = len(train_dataset.labels) + 1
model = pdx.det.FasterRCNN(num_classes=num_classes, backbone='ResNet50')
model.train(
    num_epochs=50,
    train_dataset=train_dataset,
    train_batch_size=2,
    eval_dataset=eval_dataset,
    save_interval_epochs=20,
    learning_rate=0.0025,
    warmup_steps=500,
    warmup_start_lr=1.0/1200,
    lr_decay_epochs=[10, 20, 40],
    save_dir='output_det/faster_rcnn_r50_fpn',
    use_vdl=True
    )

咋调用GPU环境？