Skip to main content
aicrowd
Projects
flatland-paper
Reports
APEX Mixed IL and RL
Log in
Sign up
Share
Comment
Star
APEX Mixed IL and RL
We train APEX with DQfD sytle Imitation loss for Imitation Learning Samples. The ratio of training IL and RL is kept constant at 25%
AIcrowd
Created on August 8
|
Last edited on August 8
Comment
Section 1
Add markdown, images, and
LaTeXLaTeX
L
a
T
e
X
custom_metrics/episode_score_normalized_mean
custom_metrics/episode_score_normalized_mean
Select runs that logged custom_metrics/episode_score_normalized_mean
to visualize data in this line chart.
evaluation/custom_metrics/episode_score_normalized_max
evaluation/custom_metrics/episode_score_normalized_max
Select runs that logged evaluation/custom_metrics/episode_score_normalized_max
to visualize data in this line chart.
MARWIL
0
Run set
3
Add a comment