metadata
library_name: hivex
original_train_name: WindFarmControl_pattern_2_task_1_run_id_0_train
tags:
- hivex
- hivex-wind-farm-control
- reinforcement-learning
- multi-agent-reinforcement-learning
model-index:
- name: hivex-WFC-PPO-baseline-task-1-pattern-2
results:
- task:
type: sub-task
name: avoid_damage
task-id: 1
pattern-id: 2
dataset:
name: hivex-wind-farm-control
type: hivex-wind-farm-control
metrics:
- type: cumulative_reward
value: 4817.209194335937 +/- 40.181277476160616
name: Cumulative Reward
verified: true
- type: avoid_damage_reward
value: 4817.198291015625 +/- 42.24024614907985
name: Avoid Damage Reward
verified: true
- type: individual_performance
value: 0.0 +/- 0.0
name: Individual Performance
verified: true
This model serves as the baseline for the Wind Farm Control environment, trained and tested on task 1 with pattern 2 using the Proximal Policy Optimization (PPO) algorithm.
Environment: Wind Farm Control
Task: 1
Pattern: 2
Algorithm: PPO
Episode Length: 5000
Training max_steps: 8000000
Testing max_steps: 8000000
Train & Test Scripts
Download the Environment