part2 complete
This commit is contained in:
2
hw3/bash/2-2-experiments.sh
Normal file
2
hw3/bash/2-2-experiments.sh
Normal file
@@ -0,0 +1,2 @@
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --exp_name cartpole_rtg_no_baseline
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg -na --use_baseline --exp_name cartpole_na_rtg_baseline
|
||||
8
hw3/bash/2-3-experiments.sh
Normal file
8
hw3/bash/2-3-experiments.sh
Normal file
@@ -0,0 +1,8 @@
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --exp_name cartpole_rtg_no_baseline
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --use_baseline --exp_name cartpole_rtg_baseline
|
||||
# with na
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg -na --use_baseline --exp_name cartpole_na_rtg_baseline
|
||||
# add bgs (default 5) and blr (default 5e-3) to the experiments
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --use_baseline --baseline_gradient_steps 3 --exp_name cartpole_rtg_baseline_bgs3
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --use_baseline --baseline_learning_rate 0.001 --exp_name cartpole_rtg_baseline_blr1e-3
|
||||
python run.py --env_name CartPole-v1 -n 100 -b 5000 -rtg --use_baseline --baseline_gradient_steps 3 --baseline_learning_rate 0.001 --exp_name cartpole_rtg_baseline_bgs3_blr1e-3
|
||||
10
hw3/bash/2-4-experiments.sh
Normal file
10
hw3/bash/2-4-experiments.sh
Normal file
@@ -0,0 +1,10 @@
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -na --use_baseline --exp_name halfcheetah_na_baseline
|
||||
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -na --use_baseline --baseline_gradient_steps 3 --exp_name halfcheetah_na_baseline_bgs3
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -na --use_baseline --baseline_learning_rate 0.001 --exp_name halfcheetah_na_baseline_blr1e-3
|
||||
# with reward to go
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -rtg -na --use_baseline --exp_name halfcheetah_na_rtg_baseline
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -rtg -na --use_baseline --baseline_gradient_steps 3 --exp_name halfcheetah_na_rtg_baseline_bgs3
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -rtg -na --use_baseline --baseline_learning_rate 0.001 --exp_name halfcheetah_na_rtg_baseline_blr1e-3
|
||||
# Berkely parameters
|
||||
python run.py --env_name HalfCheetah-v4 -n 100 -b 5000 -rtg --use_baseline --baseline_gradient_steps 5 --baseline_learning_rate 0.01 --exp_name halfcheetah_na_rtg_baseline_bgs5_blr1e-2
|
||||
Reference in New Issue
Block a user