ValueError: Sample larger than population or is negative

[sreecharan31]

~\anaconda3\envs\mask2face\lib\random.py in sample(self, population, k)
361 n = len(population)
362 if not 0 <= k <= n:
--> 363 raise ValueError("Sample larger than population or is negative")
364 result = [None] * k
365 setsize = 21 # size of a small set minus size of an empty list

ValueError: Sample larger than population or is negative

[Anuraag287]

Even am encountering the same issue. Did u find any solution?

[sreecharan31]

yeah, it's working with celebA dataset.

[Anuraag287]

ohh, thank you!! and did u train it? I am having trouble training the model due to lack of appropriate computational resources.

[Broyojo]

this is because the config has 15000 train images by default, but the default dataset includes fewer images, I believe there are only 13233 images, so just change the train images number in the config to 13233 or fewer.

[Daryl149]

@Broyojo's solution solved it. Had to reload the config and works like a charm.