tdpeuter/2026SEL3-project-Brittle_Star_Controller_Modularity
tdpeuter/2026SEL3-project-Brittle_Star_Controller_Modularity
Archived
1
Fork 0
Code Issues Releases Activity

fix: custom reward function dependent on env reward + extensions

Browse source
This commit is contained in:
Robin Meersman 2026-04-17 14:23:57 +02:00
parent 3efee5d746
commit ad17401c4c
9 changed files with 71 additions and 50 deletions
Download patch file Download diff file Expand all files Collapse all files

2
configs/ppo/debug.yaml
View file

@ -1,5 +1,5 @@
learning_rate: 0.0003
total_timesteps: 409600
total_timesteps: 409600
num_envs: 32
num_steps: 32
anneal_lr: true

2
configs/ppo/fast.yaml
View file

@ -3,7 +3,7 @@
learning_rate: 0.0005
total_timesteps: 500000
num_envs: 8
num_envs: 32
num_steps: 128
anneal_lr: true
gamma: 0.99

16
configs/ppo/fast2.yaml Normal file
View file

@ -0,0 +1,16 @@
learning_rate: 0.0003
total_timesteps: 1228800
num_envs: 32
num_steps: 64
anneal_lr: true
gamma: 0.99
gae_lambda: 0.95
num_minibatches: 32
update_epochs: 4
norm_adv: true
clip_coef: 0.2
clip_vloss: true
ent_coef: 0.005
vf_coef: 1.0
max_grad_norm: 0.5
target_kl: null