hyper_parameters_log.txt

NEW TRAINING
System_info: uname_result(system='Linux', node='vj-TUF-GAMING', release='5.4.0-137-generic', version='#154~18.04.1-Ubuntu SMP Tue Jan 10 16:58:20 UTC 2023', machine='x86_64')
GPU_info: None
BATCH_SIZE: 128
GAMMA: 0.99
EPS_START: 0.9
EPS_END: 0.05
EPS_DECAY: 1000
TAU: 0.005
LR: 0.0001
H: 500
NUM_EPISODES: 20
avg_episode_len: 2.8
avg_episode_reward: -6.8
avg_episode_len_clipped: 2.8
avg_episode_reward_clipped: -6.8
total_training_time: 5.587502717971802
avg_training_time_per_episode: 0.2793751358985901
POLICY_NETWORK: DQN(
  (h_layer1): Linear(in_features=88, out_features=128, bias=True)
  (h_layer2): Linear(in_features=128, out_features=128, bias=True)
  (h_layer3): Linear(in_features=128, out_features=128, bias=True)
  (out_layer): Linear(in_features=128, out_features=6, bias=True)
)
TARGET_NETWORK: DQN(
  (h_layer1): Linear(in_features=88, out_features=128, bias=True)
  (h_layer2): Linear(in_features=128, out_features=128, bias=True)
  (h_layer3): Linear(in_features=128, out_features=128, bias=True)
  (out_layer): Linear(in_features=128, out_features=6, bias=True)
)