hyperparameters:
batch_size: 2048
buffer_size: 20480
learning_rate: 0.0001
learning_rate: 0.0002
beta: 0.005
epsilon: 0.2
lambd: 0.95
behaviors:
WalkerStatic:
trainer_type: ppo_transfer
num_epoch: 3
learning_rate_schedule: linear
encoder_layers: 2
policy_layers: 2
value_layers: 2
forward_layers: 2
inverse_layers: 1
feature_size: 64
reuse_encoder: true
in_epoch_alter: true
use_op_buffer: true
network_settings:
normalize: true
hidden_units: 512
num_layers: 3
vis_encode_type: simple
reward_signals:
extrinsic:
gamma: 0.995
strength: 1.0
keep_checkpoints: 5
max_steps: 20000000
time_horizon: 1000
summary_freq: 30000
threaded: true