Skip to main content

AirSim-SAC-steps

Learning to perform a custom OpenAI Gym task of landing a drone autonomously in Microsoft AirSim (goal score): AirSim-v0 (1000). Soft Actor-Critic (SAC) is trained on the cost function for this task. SAC achieves the goal scores for the tasks, verified by looking at the "mean episode reward (RL)" chart. The learned SAC models are used to generate expert demonstrations of the task. GAIL is then trained on these demos to learn to imitate the tasks.
Created on October 3|Last edited on November 4

Section 1

Add markdown, images, and LaTeX\LaTeX





Parameter importance with respect to
mean episode reward (RL)

Config parameter
Importance
Correlation
Loading...
meta
["--env","AirSim-v0","--algo","sac","--exp-id","2","-rl","-eval","-best","-tb","-wandb","--env-kwargs","name:'Drone3'","rew_complexity:'complex'"]
["--env","AirSim-v0","--algo","sac","--exp-id","1","-rl","-eval","-best","-tb","-wandb","--env-kwargs","name:'Drone3'"]
12h 57m 17s
19h 8m 1s
config
env_kwargs
prabhasa
-
complex
-
2
1
-
false
sac
-
summary
46637
68881
1680
2460
1601552857
1601175100
1.21022
1.43717
6720
9840
11
7
2.8200e-8
1.4400e-8
1225.2
972.3
58.03
41.68
05001k1.5k2k2.5kStep020040060080010001200
Run set
2