justheuristic justheuristic

Consider the following hypothetical scenario:

1. You are a small hypothetical group of individuals with low influence (e.g. PhD students)

1.1 your goal is to prevent the creation of misaligned superhuman AGI, in a scenario where it would otherwise be created;
1.2 your group has some technical expertize, some compute, low resources, no political power;
1.3 the group cares about avoiding irreversible harm to other humans (e.g. causing death of anyone outside of the group itself would be strongly prohibited);

	"""
	This script creates an HTML file that displays success probabilities for Changeling the Dreaming 20th Anniversary Edition.
	See lines 27-34 for parameters that you might want to change. Note that large num_attempts will slow down the script.
	You can run the code in a Jupyter Notebook or similar to display the roll table interactively in your notebook.
	CtD20 differs from prior editions: willpower success can be cancelled by a "1" and speciality grants two successes on 10.
	If you want to change how this works (e.g. VtM 3e willpower can't be cancelled), find and modify def roll_dice(...) below.

	License: do whatever you want with this file, see Unlicense for details https://en.wikipedia.org/wiki/Unlicense
	"""
	import random

	This is free and unencumbered software released into the public domain.

	Anyone is free to copy, modify, publish, use, compile, sell, or
	distribute this software, either in source code form or as a compiled
	binary, for any purpose, commercial or non-commercial, and by any
	means.

	In jurisdictions that recognize copyright laws, the author or authors
	of this software dedicate any and all copyright interest in the
	software to the public domain. We make this dedication for the benefit

	import time
	from math import isfinite
	import pandas as pd
	import selenium.webdriver

	data = pd.read_csv('avtomaticheskaia-obrabotka-tekstov-2020-autumn - avtomaticheskaia-obrabotka-tekstov-2020-autumn.csv')
	# ^-- file format: https://docs.google.com/spreadsheets/d/1GcdT7eID_wNMjjxV0v6T0kOBRHn6Bg79qO80vFY_zgg/edit#gid=781547499

	assignment_title = 'Week9 - Word alignment'

	# authors: jheuristic, qwicen
	import torch
	import torch.nn as nn
	import threading

	device = 'cuda' if torch.cuda.is_available() else 'cpu'

	class TrainableSparseMatrix(nn.Module):
	def __init__(self, indices, values, **kwargs):
	"""

	import torch, torch.nn as nn, torch.nn.functional as F
	import numpy as np

	class OneCycleSchedule:
	""" A simplified torch lr schedule that updates learning rate before every opt.step """

	def __init__(self, optimizer, **kwargs):
	"""
	:type optimizer: torch.optim.Optimizer
	:param kwargs: see self.update_learning_rate