Sohojoe

here is the content

Sohojoe

{
    "data": [
        [
            "addString",
            "When you wake, you taste blood and you smell decay. Cold envelops your body. You hear water trickling softly – a steady "
        ],
        [   "addString",
            "drip",
            "withFXAnimation",
            "FXAnimDrip"

Sohojoe

Crowded Fiction
http://crowdedfiction.com

Vidya Gamer
http://vidyagamer.com

Joe Booth
http://joebooth.com

Sohojoe

{
  "Title": "A Princess of Mars",
  "Author": "Edgar Rice Burroughs",
  "Content": [
    {
      "ContentModelType": 3,
      "Text": "Table of Contents"
    },
    {
      "ContentModelType": 8,

Sohojoe

#Duel DDQN V0.2

Code moved to: - https://github.com/analog-rl/Duel_DDQN

OpenAI Gym: Algorthem page - https://gym.openai.com/algorithms/alg_zy3YHp0RTVOq6VXpocB20g

python duel.py CartPole-v0 --gamma 0.995 --no_display --batch_size 100 --gym_record th

Sohojoe

Sohojoe

#Duel DDQN V0.3 Duel DDQN - https://github.com/analog-rl/Duel_DDQN use this link:-

branch - https://github.com/analog-rl/Duel_DDQN/tree/b938bc593d4a206a1eb1af5773e9b6f205719454

python duel.py MountainCar-v0 --no_display --gym_record mcar-th --episodes 10000

Sohojoe

osh/kerlym#4 ...

I tried DQN & DDQN - ran both for 55k Episodes and both fail to converge

1. DQN - kerlym -e Pong-v0 -n simple_cnn -t 1 -P -f 200 -u 250 -o 0.5 -D 0 -a dqn -S 1 -i https://gym.openai.com/evaluations/eval_4M7KAHHnSFKmw4cuhkTBkg

2. DDQN -

Sohojoe

.idea

Sohojoe

// App.Config
// Created by Sohojoe on 2017/01/28