Question by
UjfalusiAbel · Jan 16 at 01:55 PM ·
artificial intelligenceagent
ML-agents config.yaml file seems incorrect when I want to train 2 brains
Hello!
I wanted to train 2 types of agents simultaneously in a competitive setting, but I can't start training on 2 behaviours. My .yaml should be correct, I even copied some examples from to official repository to see if I misconfigured something but still says it is incorrect. I need some help in it please, there is the config. Thanks in advance!
default_settings: null
behaviors:
Builder:
trainer_type: poca
hyperparameters:
batch_size: 2048
buffer_size: 20480
learning_rate: 0.0003
beta: 0.005
epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: constant
network_settings:
normalize: false
hidden_units: 512
num_layers: 2
vis_encode_type: simple
reward_signals:
extrinsic:
gamma: 0.99
strength: 1.0
keep_checkpoints: 5
max_steps: 30000000
time_horizon: 1000
summary_freq: 10000
self_play:
save_steps: 50000
team_change: 200000
swap_steps: 1000
window: 10
play_against_latest_model_ratio: 0.5
initial_elo: 1200.0
Destroyer:
trainer_type: poca
hyperparameters:
batch_size: 2048
buffer_size: 20480
learning_rate: 0.0003
beta: 0.005
epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: constant
network_settings:
normalize: false
hidden_units: 512
num_layers: 2
vis_encode_type: simple
reward_signals:
extrinsic:
gamma: 0.99
strength: 1.0
keep_checkpoints: 5
max_steps: 30000000
time_horizon: 1000
summary_freq: 10000
self_play:
save_steps: 50000
team_change: 200000
swap_steps: 4000
window: 10
play_against_latest_model_ratio: 0.5
initial_elo: 1200.0
env_settings:
env_path: null
env_args: null
base_port: 5005
num_envs: 1
seed: -1
engine_settings:
width: 84
height: 84
quality_level: 5
time_scale: 20
target_frame_rate: -1
capture_frame_rate: 60
no_graphics: false
environment_parameters: null
checkpoint_settings:
run_id: test
initialize_from: null
load_model: false
resume: false
force: false
train_model: false
inference: false
results_dir: results
torch_settings:
device: null
debug: false
Comment
Your answer
Follow this Question
Related Questions
using Navmesh or ML Agent ?! 0 Answers
ml-agents - playing the game with pretrained model 0 Answers
NavMech rotation problem 1 Answer
Navmesh Agent Multiple surfaces types 0 Answers
How to fix the rotation bug? 1 Answer