Skip to main content
sorry
Projects
trlx-references
Reports
fix-kl-controller v. main
Log in
Sign up
Share
Comment
Star
Share
Comment
Star
fix-kl-controller v. main
Sorry
Created on March 9
|
Last edited on March 9
Comment
ilql_randomwalks/GPT2Config/1gpu
metrics/optimality@beta=100
metrics/optimality@beta=100
0
50
100
150
Step
0
0.2
0.4
0.6
0.8
metrics/lengths@beta=1
metrics/lengths@beta=1
0
50
100
150
Step
50
60
70
80
90
100
metrics/optimality@beta=0
metrics/optimality@beta=0
0
50
100
150
Step
0
0.1
0.2
0.3
0.4
metrics/lengths@beta=0
metrics/lengths@beta=0
0
50
100
150
Step
50
60
70
80
90
100
metrics/lengths@beta=100
metrics/lengths@beta=100
0
50
100
150
Step
20
40
60
80
100
metrics/optimality@beta=1
metrics/optimality@beta=1
0
50
100
150
Step
0
0.1
0.2
0.3
0.4
Run set
1
Run set
1
ppo_randomwalks/randomwalks/1gpu
Run set
1
Run set
1
Add a comment