Deployed f569f4c with MkDocs version: 1.6.1

configs/ppo/chickendinnerwinner.yaml

@@ -0,0 +1,16 @@
+anneal_lr: true
+clip_coef: 0.2
+clip_vloss: true
+ent_coef: 0.001
+gae_lambda: 0.95
+gamma: 0.99
+learning_rate: 0.0001
+max_grad_norm: 0.5
+norm_adv: true
+num_envs: 32
+num_minibatches: 32
+num_steps: 64
+target_kl: 0.02
+total_timesteps: 12288000
+update_epochs: 4
+vf_coef: 1.0

configs/ppo/debug.yaml

@@ -0,0 +1,16 @@
+learning_rate: 0.0003
+total_timesteps: 409600
+num_envs: 32
+num_steps: 32
+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

configs/ppo/default.yaml

@@ -0,0 +1,19 @@
+# Default PPO Configuration
+# Standard hyperparams from original codebase.
+
+learning_rate: 0.00025
+total_timesteps: 10000000
+num_envs: 100
+num_steps: 128
+anneal_lr: true
+gamma: 0.99
+gae_lambda: 0.95
+num_minibatches: 4
+update_epochs: 4
+norm_adv: true
+clip_coef: 0.1
+clip_vloss: true
+ent_coef: 0.01
+vf_coef: 0.5
+max_grad_norm: 0.5
+target_kl: null

configs/ppo/dev_larger_timesteps_larger_rolloutsteps.yaml

@@ -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

configs/ppo/smoke_test.yaml

@@ -0,0 +1,19 @@
+# Fast PPO Configuration
+# Lower timestep count for quick iterations/testing.
+
+learning_rate: 0.0005
+total_timesteps: 1024
+num_envs: 32
+num_steps: 32
+anneal_lr: true
+gamma: 0.99
+gae_lambda: 0.95
+num_minibatches: 4
+update_epochs: 4
+norm_adv: true
+clip_coef: 0.2
+clip_vloss: true
+ent_coef: 0.01
+vf_coef: 0.5
+max_grad_norm: 0.5
+target_kl: null

configs/ppo/stable.yaml

@@ -0,0 +1,19 @@
+# Stable PPO Configuration
+# Standard hyperparams with lower LR and larger batch.
+
+learning_rate: 0.0001
+total_timesteps: 10000000
+num_envs: 100
+num_steps: 256
+anneal_lr: true
+gamma: 0.99
+gae_lambda: 0.95
+num_minibatches: 8
+update_epochs: 4
+norm_adv: true
+clip_coef: 0.1
+clip_vloss: true
+ent_coef: 0.01
+vf_coef: 0.5
+max_grad_norm: 0.5
+target_kl: null