Mr4k

Mr4k

Mr4k

loss_grad = grad(loss)
loss_grad(circles) # this gives us the gradient with respect to the circles

Mr4k

def loss(circles):
  return ((render(circles) - target) ** 2).sum()

Mr4k

use std::env;

const EMPTY_NODE_ID: usize = 0;

// an intermediate representation for symbol data during the parse step
#[derive(Clone, Debug, PartialEq)]
struct ParseSymbol {
    string: String,
    debug_program_position: usize,
}

Mr4k

import numpy as np
from bintrees import FastAVLTree as AVLTree
import random
import time

def weightedRandom(weights):
	"""
	Draw from a general discrete distribution.
	:param weights: A dictionary of weights that must sum to one.
	:return: A random sample from it the distribution defined by the weights.

Mr4k

import random

class QLearner:
	def __init__(self, num_actions):
	  #play around with these parameters for different results
		self.learning_rate = 0.03
		self.discount_factor = 0.65
		self.num_actions = num_actions
		self.last_state = None
		self.last_action = None