Skip to main content
patjj
Projects
msmarco_search
Workspace
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Pj20's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs
9
Name
5 visualized
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
1-9
of 9
Settings
Add panels
val
1
actor
6
1-6 of 6
actor/ppo_kl
actor/ppo_kl
100
200
300
400
500
Step
0
0.01
0.02
0.03
0.04
0.05
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
actor/pg_loss
actor/pg_loss
100
200
300
400
500
Step
-0.6
-0.4
-0.2
0
0.2
0.4
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
actor/pg_clipfrac
actor/pg_clipfrac
100
200
300
400
500
Step
0
0.005
0.01
0.015
0.02
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
actor/lr
actor/lr
100
200
300
400
500
Step
0
5e-7
0.000001
0.0000015
0.000002
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
actor/grad_norm
actor/grad_norm
100
200
300
400
500
Step
20
40
60
80
100
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
actor/entropy_loss
actor/entropy_loss
100
200
300
400
500
Step
0.1
0.2
0.3
0.4
0.5
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
msmarco_search_3b_health_sparse
critic
23
1-6 of 23
global_seqlen
6
1-6 of 6
mfu
2
prompt_length
4
1-4 of 4
response_length
4
1-4 of 4
reward
5
1-5 of 5
Add section