floodsung

import numpy as np
import gym
from gym.spaces import Discrete,Box

# -------------------------------------------
# Policies
# -------------------------------------------

class DeterministicDiscreteActionLinearPolicy(object):

dtinth

require 'midilib'

seq = MIDI::Sequence.new()
File.open(ARGV[0], 'rb') { |file| seq.read(file) }

events = [ ]
id = 0

seq.tracks.each do |track|

sorenbouma

import gym
import numpy as np 
import matplotlib.pyplot as plt
env = gym.make('CartPole-v0')
env.render(close=True)
#vector of means(mu) and standard dev(sigma) for each paramater
mu=np.random.uniform(size=state.shape)
sigma=np.random.uniform(low=0.001,size=state.shape)

zsal

#Most code from John Schulman's MLSS talk on Deep Reinforcement Learning
#http://rl-gym-doc.s3-website-us-west-2.amazonaws.com/mlss/lab1.html#szitalorincz06

import numpy as np
import gym
from gym.spaces import Discrete, Box

# ================================================================
# Policies
# ================================================================

fnurl

import os
import sys
import yaml
import json

# base url to use
base_url = "http://localhost:1313"

# The attribute mapping for docsearch.
#

kashif

Cross Entropy Method

How do we solve for the policy optimization problem which is to maximize the total reward given some parametrized policy?

Discounted future reward

To begin with, for an episode the total reward is the sum of all the rewards. If our environment is stochastic, we can never be sure if we will get the same rewards the next time we perform the same actions. Thus the more we go into the future the more the total future reward may diverge. So for that reason it is common to use the discounted future reward where the parameter discount is called the discount factor and is between 0 and 1.

A good strategy for an agent would be to always choose an action that maximizes the (discounted) future reward. In other words we want to maximize the expected reward per episode.

squiter

#!/bin/bash

set -eou pipefail

version=8.6.4.1
patchinfo=299124-linux-x86_64-threaded
dir=ActiveTcl$version.$patchinfo
package=$dir.tar.gz
url=http://downloads.activestate.com/ActiveTcl/releases/$version/$package

eph2795

import gym
from tqdm import tqdm_notebook
import numpy as np
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings('ignore')

def get_random_policy():
    return np.random.choice(n_actions, tuple(bins))

brookisme

#
# add to bashrc
#

# files
alias sampletree='mkdir -p sample/{train,test,valid}'
lsn(){ matchdir=`pwd`/$2; find $matchdir -type f | grep -v sample | shuf -n $1 | awk -F`pwd` '{print "."$NF}' ; }
# shuffle mv/cp
cpn(){ matchdir=`pwd`/$2; find $matchdir -type f | grep -v sample | shuf -n $1 | awk -F`pwd` '{print "."$NF" sample"$NF}' | xargs -t -n2 cp ; }
mvn(){ matchdir=`pwd`/$2; todir=`pwd`/$3; find $matchdir -type f | grep -v sample | shuf -n $1 | awk -F`pwd` -v todir="$todir" '{print $0" "todir}' | xargs -t -n2 mv ; }

pat-coady