phuocphn

'''
As with NLLLoss , the input given is expected to contain log-probabilities… 
The targets are given as probabilities (i.e. without taking the logarithm).
https://discuss.pytorch.org/t/kldivloss-returns-negative-value/62148
https://discuss.pytorch.org/t/kl-divergence-produces-negative-values/16791/16
https://discuss.pytorch.org/t/kullback-leibler-divergence-loss-function-giving-negative-values/763/16

'''

import torch

phuocphn

import torch
import torch.nn as nn
from lsq import Conv2dLSQ, LinearLSQ, ActLSQ

__all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101',
           'resnet152', 'resnext50_32x4d', 'resnext101_32x8d',
           'wide_resnet50_2', 'wide_resnet101_2']

def conv3x3(in_planes, out_planes, stride=1, groups=1, dilation=1):
    """3x3 convolution with padding"""

phuocphn

import torch
import torch.nn as nn
import torch.optim as optim
from torch.autograd import Variable
import torch.nn.functional as F

import matplotlib.pyplot as plt

import numpy as np

phuocphn

import torch
import numpy as np 

torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(0)
torch.manual_seed(0)

phuocphn

import torch
import numpy as np 

torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(0)
torch.manual_seed(0)

phuocphn

import torch
import numpy as np 

torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(0)
torch.manual_seed(0)


nbit = 3

phuocphn

import math
import numpy as np

import torch.nn.functional as F
import torch 
import torch.nn as nn


torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False

phuocphn

import torch as t
import numpy as np

class GradScale(t.nn.Module):
    def forward(self, x, scale):
        y = x
        y_grad = x / scale
        return (y - y_grad).detach() + y_grad

phuocphn

import os
import tempfile
import torch
import torch.distributed as dist
import torch.nn as nn
import torch.optim as optim
import torch.multiprocessing as mp
import numpy as np
import random
from torch.nn.parallel import DistributedDataParallel as DDP

phuocphn

'''
Author: komuw/pycrypto_AES.py

Security issues:
+ https://stackoverflow.com/questions/2641720/for-aes-cbc-encryption-whats-the-importance-of-the-iv
+ https://defuse.ca/cbcmodeiv.htm
+ https://passingcuriosity.com/2009/aes-encryption-in-python-with-m2crypto/
'''

from Crypto.Cipher import AES