Skip to main content
costa-huang
Projects
cleanRL
Reports
Cleanba impala threads ASAP max GPU
Log in
Sign up
Share
Comment
Star
Share
Comment
Star
Cleanba impala threads ASAP max GPU
Costa
Created on April 10
|
Last edited on April 16
Comment
charts/SPS_update
charts/SPS_update
10M
20M
30M
40M
50M
global_step
10000
20000
30000
charts/learning_rate
charts/learning_rate
Select runs that logged charts/learning_rate
to visualize data in this line chart.
losses/policy_loss
losses/policy_loss
10M
20M
30M
40M
50M
global_step
-400
-200
0
200
400
charts/avg_episodic_return
charts/avg_episodic_return
Select runs that logged charts/avg_episodic_return
to visualize data in this line chart.
charts/avg_episodic_return
charts/avg_episodic_return
5
10
15
20
25
Time (minutes)
0
100
200
300
400
system/gpu.0.gpu
system/gpu.0.gpu
5
10
15
20
25
Time (minutes)
10
20
30
40
50
60
70
actor threads (A100)
1
cleanba impala 8 A100 baseline (a0_l1_d4)
1
actor threads asap (A100) prime candidate
1
a0_l1+2+3_t4
1
a0_l1+2+3_t8 w/ no .item() logging
1
a0_l1+2+3_t8_n60
1
a0+1_l2+3+4_t4
1
a0+1+2_l3+4+5_t4
1
a0_l1+2+3_t4 w/o actual env step
1
cleanba impala a0+1_l2+3+4_t4 w/ thread_affinity
1
cleanba PPO baseline
1
TPU
1
a0_l1+2+3_t4 (TPU equivalent)
1
a0_l0_t4 (TPU equivalent)
1
a0_l0_t4
1
a0_l0_t4 old actor threads
1
a0_l0_t4_n30
1
a0_l0_t4_n30_nmb1
1
a0_l0_t4_n120_nmb2
1
a0_l0_t4_n30_nmb2
1
a0_l0_t4_n16_nmb2
1
Add a comment