huantwang / supersonic Goto Github PK

> conda activate ss
>  cd /home/sys/SUPERSONIC

>  python SuperSonic/policy_search/supersonic_main.py --env BanditCSREnv-v0  --datapath "tasks/CSR/DATA" --mode policy --total_steps 10 2>/dev/null

> python SuperSonic/policy_search/supersonic_main.py --env BanditCSREnv-v0  --datapath "tasks/CSR/DATA" --mode config --iterations 10 --task CSR 2>/dev/null

> python SuperSonic/policy_search/supersonic_main.py --env BanditCSREnv-v0  --datapath "tasks/CSR/DATA" --mode deploy --training_iterations 50 --task CSR 2>/dev/null
...
== Status ==
Memory usage on this node: 4.0/62.0 GiB
Using FIFO scheduling algorithm.
Resources requested: 0/50 CPUs, 0/1 GPUs, 0.0/36.13 GiB heap, 0.0/12.45 GiB objects
Result logdir: /home/sys/SUPERSONIC/SuperSonic/logs/model_save/SAC
Number of trials: 1 (1 TERMINATED)
+------------------------+------------+-------+--------+------------------+------+----------+
| Trial name             | status     | loc   |   iter |   total time (s) |   ts |   reward |
|------------------------+------------+-------+--------+------------------+------+----------|
| SAC_csr_rl_b5add_00000 | TERMINATED |       |     50 |          61.4566 | 9650 |        0 |
+------------------------+------------+-------+--------+------------------+------+----------+
-------------------------------

optimization finished, ready to use our script to evaluate performance
-------------------------------

start executing, usually takes 3 mins

> python tasks/CSR/run.py

Traceback (most recent call last):
  File "tasks/CSR/run.py", line 1, in <module>
    from AE.utils.Calculate import CalculateCSRDemo
ModuleNotFoundError: No module named 'AE'