training_history.json

[
  {
    "episode": 1,
    "epsilon": 0.9995,
    "avg_reward_200": -1.0089,
    "avg_score_200": 0.3427,
    "ema_easy": 0.5,
    "ema_medium": 0.4843,
    "ema_hard": 0.5
  },
  {
    "episode": 100,
    "epsilon": 0.9461,
    "avg_reward_200": -0.2571,
    "avg_score_200": 0.4869,
    "ema_easy": 0.7062,
    "ema_medium": 0.3607,
    "ema_hard": 0.5425
  },
  {
    "episode": 200,
    "epsilon": 0.8922,
    "avg_reward_200": -0.2741,
    "avg_score_200": 0.4821,
    "ema_easy": 0.699,
    "ema_medium": 0.3753,
    "ema_hard": 0.5207
  },
  {
    "episode": 300,
    "epsilon": 0.8382,
    "avg_reward_200": -0.188,
    "avg_score_200": 0.4979,
    "ema_easy": 0.6441,
    "ema_medium": 0.4126,
    "ema_hard": 0.5528
  },
  {
    "episode": 400,
    "epsilon": 0.7843,
    "avg_reward_200": -0.1041,
    "avg_score_200": 0.526,
    "ema_easy": 0.7491,
    "ema_medium": 0.4085,
    "ema_hard": 0.6285
  },
  {
    "episode": 500,
    "epsilon": 0.7304,
    "avg_reward_200": -0.0853,
    "avg_score_200": 0.538,
    "ema_easy": 0.7122,
    "ema_medium": 0.4281,
    "ema_hard": 0.5412
  },
  {
    "episode": 600,
    "epsilon": 0.6765,
    "avg_reward_200": -0.0095,
    "avg_score_200": 0.5571,
    "ema_easy": 0.7304,
    "ema_medium": 0.4528,
    "ema_hard": 0.6262
  },
  {
    "episode": 700,
    "epsilon": 0.6226,
    "avg_reward_200": -0.0167,
    "avg_score_200": 0.5557,
    "ema_easy": 0.7331,
    "ema_medium": 0.4076,
    "ema_hard": 0.6067
  },
  {
    "episode": 800,
    "epsilon": 0.5686,
    "avg_reward_200": -0.0749,
    "avg_score_200": 0.5352,
    "ema_easy": 0.7717,
    "ema_medium": 0.4499,
    "ema_hard": 0.6342
  },
  {
    "episode": 900,
    "epsilon": 0.5147,
    "avg_reward_200": -0.063,
    "avg_score_200": 0.5235,
    "ema_easy": 0.7954,
    "ema_medium": 0.4238,
    "ema_hard": 0.5801
  },
  {
    "episode": 1000,
    "epsilon": 0.4608,
    "avg_reward_200": 0.0078,
    "avg_score_200": 0.5359,
    "ema_easy": 0.7646,
    "ema_medium": 0.4622,
    "ema_hard": 0.7117
  },
  {
    "episode": 1100,
    "epsilon": 0.4069,
    "avg_reward_200": 0.0611,
    "avg_score_200": 0.5559,
    "ema_easy": 0.8171,
    "ema_medium": 0.4475,
    "ema_hard": 0.6804
  },
  {
    "episode": 1200,
    "epsilon": 0.353,
    "avg_reward_200": 0.1037,
    "avg_score_200": 0.5546,
    "ema_easy": 0.8186,
    "ema_medium": 0.4633,
    "ema_hard": 0.7064
  },
  {
    "episode": 1300,
    "epsilon": 0.2991,
    "avg_reward_200": 0.1535,
    "avg_score_200": 0.5693,
    "ema_easy": 0.8344,
    "ema_medium": 0.5137,
    "ema_hard": 0.7561
  },
  {
    "episode": 1400,
    "epsilon": 0.2451,
    "avg_reward_200": 0.208,
    "avg_score_200": 0.5948,
    "ema_easy": 0.7968,
    "ema_medium": 0.5195,
    "ema_hard": 0.7195
  },
  {
    "episode": 1500,
    "epsilon": 0.1912,
    "avg_reward_200": 0.1866,
    "avg_score_200": 0.573,
    "ema_easy": 0.8134,
    "ema_medium": 0.4705,
    "ema_hard": 0.7068
  },
  {
    "episode": 1600,
    "epsilon": 0.1373,
    "avg_reward_200": 0.1404,
    "avg_score_200": 0.5424,
    "ema_easy": 0.8137,
    "ema_medium": 0.4946,
    "ema_hard": 0.6969
  },
  {
    "episode": 1700,
    "epsilon": 0.0834,
    "avg_reward_200": 0.2049,
    "avg_score_200": 0.5579,
    "ema_easy": 0.8332,
    "ema_medium": 0.5128,
    "ema_hard": 0.7636
  },
  {
    "episode": 1800,
    "epsilon": 0.03,
    "avg_reward_200": 0.2726,
    "avg_score_200": 0.5916,
    "ema_easy": 0.8391,
    "ema_medium": 0.5755,
    "ema_hard": 0.7705
  }
]