环境
- Ubuntu: 16.04
- CUDA: 10.0
- CUDNN: 7.6.5
执行情况
$ python3 train.py
{'train': {'data_loader': {'batch_size': 1, 'resize_shape': [600, 600], 'train_and_val': 0.8, 'train_img': './data/sample/train_data/train_images', 'train_txt': './data/sample/train_data/train.txt'}, 'save_model': {'h5_savePath': './saved_models/model_epoch{epoch}.h5', 'save_freq': 5000, 'pb_save_path': './saved_models/train'}, 'base_model': {'input_shape': [600, 600, 3], 'pretrain_model': 'imagenet', 'trainable': True, 'model_name': 'None', 'num_classes': 2}, 'log_dir': './logs', 'epochs': 50, 'learning_rate': 0.001}, 'test': {'pb_load_path': './saved_models/train', 'test_img': '/aisdata/dataset/table_images/