import numpy as np
import random
import torch
[docs]def seed(s=0):
"""
Seeds every possible random number generator
with a specific seed to guarantee the same
results with the same seed (hopefully). Currently
supported RNGs are::
torch.manual_seed
torch.cuda.manual_seed
np.random.seed
random.seed
If you use RNGs outside of this list in your project,
then there's no guarantee of reproducibility.
Args:
s (int, optional): Seed to use. Defaults to 0.
"""
torch.backends.cudnn.benchmark = True
torch.manual_seed(s)
torch.cuda.manual_seed(s)
np.random.seed(s)
random.seed(s)