Skip to main content
zhaochenyang20
Projects
benchmark_over_sample_2
Workspace
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Zhaochenyang20's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs
4
Name
4 visualized
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
qwen2.5-3b_baseline_05-12-00_0.7
1-4
of 4
Settings
New report
Add panels
actor
9
1-6 of 9
actor/ppo_kl
actor/ppo_kl
10
20
30
40
Step
-2
-1
0
1
2
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
actor/pg_loss
actor/pg_loss
10
20
30
40
Step
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
actor/pg_clipfrac_lower
actor/pg_clipfrac_lower
10
20
30
40
Step
-2
-1
0
1
2
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
actor/pg_clipfrac
actor/pg_clipfrac
10
20
30
40
Step
-2
-1
0
1
2
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
actor/lr
actor/lr
10
20
30
40
Step
0
5e-7
0.000001
0.0000015
0.000002
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
actor/kl_loss
actor/kl_loss
10
20
30
40
Step
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
qwen2.5-3b_baseline_05-14-24_1.0
qwen2.5-3b_baseline_05-13-36_0.9
qwen2.5-3b_baseline_05-12-48_0.8
qwen2.5-3b_baseline_05-12-00_0.7
critic
12
1-6 of 12
global_seqlen
6
1-6 of 6
perf
7
1-6 of 7
prompt_length
4
1-4 of 4
response_length
4
1-4 of 4
Add section