Sofian Mejjoute Ryu1845

NICONICO/ニコニコ Community OSINT

Author: EF1500

Introduction

Niconico is like Japanese BiliBili, it is a video hosting platform where users can host videos, create communities and other various activities. Depending on the age of the profile/community, you can discover various information that is crucial to any investigation. However, it may be difficult to pick out information if you are new to OSINT or do not speak the language. This writeup aims to make it easier to find information about Niconico communities/users.

Information you can uncover from Niconico may include:

Email Addresses
Skype Usernames
Friend Usernames/Communities

_{masterpiece, best quality, 1girl, green hair, sweater, looking at viewer, upper body, beanie, outdoors, watercolor, night, turtleneck}

Waifu Diffusion 1.4 Anime Release Notes

Download Pages

HuggingFace Diffusers: https://huggingface.co/hakurei/waifu-diffusion
Pickled Models: https://huggingface.co/hakurei/waifu-diffusion-v1-4

Waifu Diffusion 1.3 Release Notes

HuggingFace Page for model download: https://huggingface.co/hakurei/waifu-diffusion-v1-3

Waifu Diffusion 1.4 Overview

_{An image generated at resolution 512x512 then upscaled to 1024x1024 with Waifu Diffusion 1.3 Epoch 7.}

Goals

Improving image generation at different aspect ratios using conditional masking during training. This will allow for the entire image to be seen during training instead of center cropped images, which will allow for better results when generating full body images, portraits, and improving the composition.
Expanded the input context from 77 tokens to 231 tokens or perhaps to an unlimited amount of tokens. Out of 77 tokens for input, only 75 are useable. This does not give nearly enough room for complex prompting that requires a lot of detail.

	"""
	Simplified Implementation of the Linear Recurrent Unit
	------------------------------------------------------
	We present here a simplified JAX implementation of the Linear Recurrent Unit (LRU).
	The state of the LRU is driven by the input $(u_k)_{k=1}^L$ of sequence length $L$
	according to the following formula (and efficiently parallelized using an associative scan):
	$x_{k} = \Lambda x_{k-1} +\exp(\gamma^{\log})\odot (B u_{k})$,
	and the output is computed at each timestamp $k$ as follows: $y_k = C x_k + D u_k$.
	In our code, $B,C$ follow Glorot initialization, with $B$ scaled additionally by a factor 2
	to account for halving the state variance by taking the real part of the output projection.

	#!/usr/bin/env python

	import math
	import matplotlib.pyplot as plt
	import torch
	import torch.nn as nn

	from sklearn.datasets import make_moons
	from torch import Tensor
	from tqdm import tqdm

	#!/usr/bin/env python3

	"""A simple JAX process launcher for multiple devices on a single host.

	You must import jax_local_cluster somewhere inside the script you are launching.
	"""

	import argparse
	from functools import partial
	import os

	"""
	clipbooru.py

	Go through a list of (md5).(ext), downloads them from danbooru, and compute CLIP
	embeddings. Images are downloaded in parallel while inference is being done, so
	this should run as fast as your hardware/network can handle.
	"""

	# requirements.txt
	"""

	#!/usr/bin/env bash

	# Small shell script to more easily automatically download and transcribe live stream VODs.
	# This uses YT-DLP, ffmpeg and the CPP version of Whisper: https://github.com/ggerganov/whisper.cpp
	# Use `./transcribe-vod help` to print help info.

	# MIT License

	# Copyright (c) 2022 Daniils Petrovs

	#!/bin/python

	####
	# Designed for use with https://github.com/AUTOMATIC1111/stable-diffusion-webui
	# Run this file from the root of the repo
	#
	# Usage in prompt: Put embedding name in prompt (using moxxi.pt and borderlands.pt)
	# e.g. moxxi on a beach, in the style of borderlands sharp, clear lines, detailed
	# or
	# e.g. <moxxi> on a beach, in the style of <borderlands> sharp, clear lines, detailed