Skip to main content
andreaskoepf
Projects
openrlhf_train_ppo
Workspace
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Andreaskoepf's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs
2
Name
2 visualized
ppo_0128T17:18
ppo_0128T17:18
ppo_0128T16:29
ppo_0128T16:29
1-2
of 2
Settings
Add panels
train
11
1-6 of 11
train/values
train/values
20
40
60
80
train/global_step
0.2
0.4
0.6
0.8
ppo_0128T17:18
ppo_0128T16:29
train/total_length
train/total_length
20
40
60
80
train/global_step
180
200
220
240
ppo_0128T17:18
ppo_0128T16:29
train/reward
train/reward
20
40
60
80
train/global_step
0.2
0.4
0.6
0.8
ppo_0128T17:18
ppo_0128T16:29
train/return
train/return
20
40
60
80
train/global_step
0.2
0.4
0.6
0.8
ppo_0128T17:18
ppo_0128T16:29
train/response_length
train/response_length
20
40
60
80
train/global_step
60
80
100
120
ppo_0128T17:18
ppo_0128T16:29
train/policy_loss
train/policy_loss
20
40
60
80
train/global_step
-0.2
-0.15
-0.1
-0.05
0
0.05
ppo_0128T17:18
ppo_0128T16:29
System
25
1-6 of 25
Add section