Skip to main content
llychinalz
Projects
Flash-DAPO
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Llychinalz's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs
3
Name
3 visualized
flash-w8a8-TIS5-T1-cleanedX
flash-w8a8-TIS5-T1-cleanedX
flash-baseline-T1-cleanedX
flash-baseline-T1-cleanedX
flash-baseline-T1-cleanedX
flash-baseline-T1-cleanedX
1-3
of 3
Add panels
Panel Section
6
Pinned
1-6 of 6
val-core/math_dapo/acc/mean@32
val-core/math_dapo/acc/mean@32
0
50
100
150
200
250
Step
0.1
0.2
0.3
0.4
critic/rewards/mean
critic/rewards/mean
50
100
150
200
250
Step
-0.6
-0.4
-0.2
0
0.2
0.4
actor/pg_clipfrac
actor/pg_clipfrac
50
100
150
200
250
Step
0.002
0.004
0.006
0.008
0.01
train/vllm_kl
train/vllm_kl
50
100
150
200
250
Step
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
actor/entropy_loss
actor/entropy_loss
50
100
150
200
250
Step
0.4
0.6
0.8
1
1.2
perf/throughput
perf/throughput
50
100
150
200
250
Step
200
300
400
actor
6
1-6 of 6
critic
13
1-6 of 13
global_seqlen
6
1-6 of 6
perf
6
1-6 of 6
prompt_length
4
1-4 of 4
response_length
4
1-4 of 4
timing_per_token_ms
3
timing_s
10
1-6 of 10
train
2
training
4
1-4 of 4
val-aux
91
1-6 of 91
val-core
4
1-4 of 4
Add section