Kim Seonghyeon rosinality

no side-effects

rosinality / mathology.html

Created January 23, 2015 07:34

Very simple latex sketchpad

	<!DOCTYPE html>
	<html lang="en">
	<head>
	<meta charset="UTF-8">
	<title>Mathology</title>
	<style>
	body {
	font-family: Arial, Helvetica, sans-serif;
	}

rosinality / adasoft.py

Last active July 21, 2021 03:42

Adaptive Softmax implementation for PyTorch

	import torch
	from torch import nn
	from torch.autograd import Variable

	class AdaptiveSoftmax(nn.Module):
	def __init__(self, input_size, cutoff):
	super().__init__()

	self.input_size = input_size
	self.cutoff = cutoff

rosinality / logtest.py

Last active April 8, 2017 14:22

Simple live log plotting tool for Visdom

rosinality / mhsampler-in-pytorch.ipynb

Created April 15, 2017 02:40

Metropolis-Hastings sampler in PyTorch. Made to explore possibility of bayesian computation in PyTorch.

Sorry, something went wrong. Reload?

Sorry, we cannot display this file.

Sorry, this file is invalid so it cannot be displayed.

rosinality / discriminator_example.py

Created December 9, 2017 15:37

Implementation of Spectral Normalization for PyTorch

	from torch import nn
	from torch.nn import init
	from torch.nn import functional as F

	def init_conv(conv, glu=True):
	init.kaiming_normal(conv.weight)
	if conv.bias is not None:
	conv.bias.data.zero_()

	class ConvBlock(nn.Module):

rosinality / perceptual_loss.py

Created February 7, 2020 13:08

Perceptual loss implementation sample

	import torch
	from torch import nn
	from torchvision.models import vgg16, vgg16_bn, vgg19, vgg19_bn


	class PerceptualLoss(nn.Module):
	def __init__(self, arch, indices, weights, normalize=True, min_max=(-1, 1)):
	super().__init__()

	vgg = (

rosinality / perceptual_loss.py

Created February 7, 2020 13:08

Perceptual loss implementation sample

	import torch
	from torch import nn
	from torchvision.models import vgg16, vgg16_bn, vgg19, vgg19_bn


	class PerceptualLoss(nn.Module):
	def __init__(self, arch, indices, weights, normalize=True, min_max=(-1, 1)):
	super().__init__()

	vgg = (