metadata
library_name: hivex
original_train_name: AerialWildfireSuppression_difficulty_5_task_6_run_id_2_train
tags:
- hivex
- hivex-aerial-wildfire-suppression
- reinforcement-learning
- multi-agent-reinforcement-learning
model-index:
- name: hivex-AWS-PPO-baseline-task-6-difficulty-5
results:
- task:
type: sub-task
name: drop_water
task-id: 6
difficulty-id: 5
dataset:
name: hivex-aerial-wildfire-suppression
type: hivex-aerial-wildfire-suppression
metrics:
- type: crash_count
value: 0.022751575300935654 +/- 0.009602885208751202
name: Crash Count
verified: true
- type: extinguishing_trees
value: 0.2077373639680445 +/- 0.1507774190085061
name: Extinguishing Trees
verified: true
- type: extinguishing_trees_reward
value: 1.0386868014931678 +/- 0.753887068245034
name: Extinguishing Trees Reward
verified: true
- type: preparing_trees
value: 275.6179458618164 +/- 12.486733727275665
name: Preparing Trees
verified: true
- type: preparing_trees_reward
value: 275.6179458618164 +/- 12.486733727275665
name: Preparing Trees Reward
verified: true
- type: water_drop
value: 0.9767484277486801 +/- 0.00989079886411895
name: Water Drop
verified: true
- type: water_pickup
value: 0.0005922459880821407 +/- 0.0015700193469281448
name: Water Pickup
verified: true
- type: cumulative_reward
value: 274.48731155395507 +/- 13.254462171454286
name: Cumulative Reward
verified: true
This model serves as the baseline for the Aerial Wildfire Suppression environment, trained and tested on task 6 with difficulty 5 using the Proximal Policy Optimization (PPO) algorithm.
Environment: Aerial Wildfire Suppression
Task: 6
Difficulty: 5
Algorithm: PPO
Episode Length: 3000
Training max_steps: 1800000
Testing max_steps: 180000
Train & Test Scripts
Download the Environment