poca-SoccerTwosMemory / configuration.yaml
brettgoehre's picture
22th Push
fd2023e verified
raw
history blame
2.67 kB
default_settings: null
behaviors:
SoccerTwos:
trainer_type: poca
hyperparameters:
batch_size: 60000
buffer_size: 180000
learning_rate: 0.0003
beta: 0.005
epsilon: 0.3
lambd: 0.9
num_epoch: 3
learning_rate_schedule: constant
beta_schedule: linear
epsilon_schedule: constant
checkpoint_interval: 500000
network_settings:
normalize: true
hidden_units: 512
num_layers: 2
vis_encode_type: simple
memory:
sequence_length: 16
memory_size: 128
goal_conditioning_type: hyper
deterministic: false
reward_signals:
extrinsic:
gamma: 0.999
strength: 1.0
network_settings:
normalize: false
hidden_units: 128
num_layers: 2
vis_encode_type: simple
memory: null
goal_conditioning_type: hyper
deterministic: false
curiosity:
gamma: 0.7
strength: 1.0e-06
network_settings:
normalize: false
hidden_units: 512
num_layers: 2
vis_encode_type: simple
memory: null
goal_conditioning_type: hyper
deterministic: false
learning_rate: 0.0003
encoding_size: 512
rnd:
gamma: 0.7
strength: 1.0e-06
network_settings:
normalize: false
hidden_units: 128
num_layers: 2
vis_encode_type: simple
memory: null
goal_conditioning_type: hyper
deterministic: false
learning_rate: 0.0001
encoding_size: null
init_path: null
keep_checkpoints: 1
even_checkpoints: false
max_steps: 10000000000
time_horizon: 1000
summary_freq: 10000
threaded: false
self_play:
save_steps: 500000
team_change: 10000
swap_steps: 500
window: 30
play_against_latest_model_ratio: 0.9
initial_elo: 1.0
behavioral_cloning: null
env_settings:
env_path: ./SoccerTwos/SoccerTwos.app
env_args: null
base_port: 5005
num_envs: 1
num_areas: 1
timeout_wait: 60
seed: -1
max_lifetime_restarts: 10
restarts_rate_limit_n: 1
restarts_rate_limit_period_s: 60
engine_settings:
width: 84
height: 84
quality_level: 5
time_scale: 20
target_frame_rate: -1
capture_frame_rate: 60
no_graphics: true
no_graphics_monitor: false
environment_parameters: null
checkpoint_settings:
run_id: sept-run-01
initialize_from: null
load_model: false
resume: true
force: false
train_model: false
inference: false
results_dir: results
torch_settings:
device: null
debug: false