基于PaddlePaddle的散射成像调控深度学习训练工具.zip

# Scatter Generate Tool datasets ----------- # base简介 > 实现基本的数据集基类与数据加载器 ## 1.主要内容 - 实现数据集训练与评估的基类 - 实现测试数据集 - 实现数据集加载器的基类 - 实现训练、评估、测试的数据集加载器 ## 2.主要接口 - TestDataset ```python class TestDataset(Dataset): def __init__(self, image_dir: str): super(TestDataset, self).__init__() ``` - TrainDataLoader ```python class TrainDataLoader(BaseDataLoader): """训练数据加载器 """ def __init__(self, dataset: Dataset, sample_transforms: List[Any] = [], batch_transforms: List[Any] = [], batch_size: int = 1, shuffle: bool = True, drop_last: bool = False, worker_num: int = 1): super(TrainDataLoader, self).__init__( dataset, sample_transforms, batch_transforms, batch_size, shuffle, drop_last, worker_num ) ``` - EvalDataLoader ```python class EvalDataLoader(BaseDataLoader): """验证数据加载器 """ def __init__(self, dataset: Dataset, sample_transforms: List[Any] = [], batch_transforms: List[Any] = [], batch_size: int = 1, worker_num: int = 1): super(EvalDataLoader, self).__init__( dataset, sample_transforms, batch_transforms, batch_size, False, False, worker_num ) ``` - TestDataLoader ```python class TestDataLoader(BaseDataLoader): """测试数据加载器 """ def __init__(self, dataset: Dataset, sample_transforms: List[Any] = [], batch_transforms: List[Any] = [], batch_size: int = 1, worker_num: int = 1): super(TestDataLoader, self).__init__( dataset, sample_transforms, batch_transforms, batch_size, False, False, worker_num ) ``` ----------- # normal简介 > 实现一种通用基础的数据集类 ## 1.主要内容 - 实现Normal训练数据集类 - 实现Normal评估数据集类 ## 2.主要接口 - TrainNormalDataset ```python class TrainNormalDataset(NormalDataset): def __init__(self, data_dir: str, data_txt: str): super(TrainNormalDataset, self).__init__( data_dir, data_txt, True ) ``` - EvalNormalDataset ```python class EvalNormalDataset(NormalDataset): def __init__(self, data_dir: str, data_txt: str): super(TrainNormalDataset, self).__init__( data_dir, data_txt, False ) ```