＆quot“ get_labels”但是对于迷你象征？

Question

我目前正在进行几次学习教程的练习，并遇到了一个使用Omniglot数据集的练习。 this 。我希望复制它，但使用Miniimagenet数据集。这个代码块是我一直面临问题的地方。

N_WAY = 5  
N_SHOT = 5  
N_QUERY = 10  
N_EVALUATION_TASKS = 100


test_set.get_labels = lambda: [
    instance[1] for instance in test_set._flat_character_images
]
test_sampler = TaskSampler(
    test_set, n_way=N_WAY, n_shot=N_SHOT, n_query=N_QUERY, n_tasks=N_EVALUATION_TASKS
)

test_loader = DataLoader(
    test_set,
    batch_sampler=test_sampler,
    num_workers=12,
    pin_memory=True,
    collate_fn=test_sampler.episodic_collate_fn,
)

我应该在此修改什么：

test_set.get_labels = lambda: [
        instance[1] for instance in test_set._flat_character_images
    ]

为了允许它运行？

这是错误代码：

    AttributeError                            Traceback (most recent call last)
<ipython-input-7-15761ef243cd> in <module>
      9 ]
     10 test_sampler = TaskSampler(
---> 11     test_set, n_way=N_WAY, n_shot=N_SHOT, n_query=N_QUERY, n_tasks=N_EVALUATION_TASKS
     12 )
     13 

D:\anaconda3\envs\FSL\lib\site-packages\easyfsl\samplers\task_sampler.py in __init__(self, dataset, n_way, n_shot, n_query, n_tasks)
     39 
     40         self.items_per_label = {}
---> 41         for item, label in enumerate(dataset.get_labels()):
     42             if label in self.items_per_label.keys():
     43                 self.items_per_label[label].append(item)

<ipython-input-7-15761ef243cd> in <lambda>()
      6 # The sampler needs a dataset with a "get_labels" method. Check the code if you have any doubt!
      7 test_set.get_labels = lambda: [
----> 8     instance[1] for instance in test_set._flat_images
      9 ]
     10 test_sampler = TaskSampler(

D:\anaconda3\envs\FSL\lib\site-packages\torch\utils\data\dataset.py in __getattr__(self, attribute_name)
     81             return function
     82         else:
---> 83             raise AttributeError
     84 
     85     @classmethod

AttributeError: