philippds commited on
Commit
be44b15
·
verified ·
1 Parent(s): 5a18705

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +76 -74
README.md CHANGED
@@ -1,74 +1,76 @@
1
- ---
2
- library_name: hivex
3
- original_train_name: AerialWildfireSuppression_difficulty_1_task_4_run_id_0_train
4
- tags:
5
- - hivex
6
- - hivex-aerial-wildfire-suppression
7
- - reinforcement-learning
8
- - multi-agent-reinforcement-learning
9
- model-index:
10
- - name: hivex-AWS-PPO-baseline-task-4-difficulty-1
11
- results:
12
- - task:
13
- type: sub-task
14
- name: protect_village
15
- task-id: 4
16
- difficulty-id: 1
17
- dataset:
18
- name: hivex-aerial-wildfire-suppression
19
- type: hivex-aerial-wildfire-suppression
20
- metrics:
21
- - type: crash_count
22
- value: 0.6166666761040688 +/- 0.22360679529482064
23
- name: Crash Count
24
- verified: true
25
- - type: extinguishing_trees
26
- value: 0.4333333492279053 +/- 1.9379256515825043
27
- name: Extinguishing Trees
28
- verified: true
29
- - type: extinguishing_trees_reward
30
- value: 2.166666603088379 +/- 9.689627618168343
31
- name: Extinguishing Trees Reward
32
- verified: true
33
- - type: fire_out
34
- value: 0.05833333507180214 +/- 0.11180340220699042
35
- name: Fire Out
36
- verified: true
37
- - type: fire_too_close_to_city
38
- value: 0.425 +/- 0.37257991018529735
39
- name: Fire too Close to City
40
- verified: true
41
- - type: preparing_trees
42
- value: 210.27499961853027 +/- 101.36103492999625
43
- name: Preparing Trees
44
- verified: true
45
- - type: preparing_trees_reward
46
- value: 210.27499961853027 +/- 101.36103492999625
47
- name: Preparing Trees Reward
48
- verified: true
49
- - type: water_drop
50
- value: 3.3749999940395354 +/- 2.19507356485688
51
- name: Water Drop
52
- verified: true
53
- - type: water_pickup
54
- value: 3.2916666686534883 +/- 2.260695820449611
55
- name: Water Pickup
56
- verified: true
57
- - type: cumulative_reward
58
- value: -69.60107120275498 +/- 211.17136032627994
59
- name: Cumulative Reward
60
- verified: true
61
- ---
62
-
63
- This model serves as the baseline for the **Aerial Wildfire Suppression** environment, trained and tested on task <code>4</code> with difficulty <code>1</code> using the Proximal Policy Optimization (PPO) algorithm.<br><br>
64
-
65
- Environment: **Aerial Wildfire Suppression**<br>
66
- Task: <code>4</code><br>
67
- Difficulty: <code>1</code><br>
68
- Algorithm: <code>PPO</code><br>
69
- Episode Length: <code>3000</code><br>
70
- Training <code>max_steps</code>: <code>1800000</code><br>
71
- Testing <code>max_steps</code>: <code>180000</code><br><br>
72
-
73
- Train & Test [Scripts](https://github.com/hivex-research/hivex)<br>
74
- Download the [Environment](https://github.com/hivex-research/hivex-environments)
 
 
 
1
+ ---
2
+ library_name: hivex
3
+ original_train_name: AerialWildfireSuppression_difficulty_1_task_4_run_id_0_train
4
+ tags:
5
+ - hivex
6
+ - hivex-aerial-wildfire-suppression
7
+ - reinforcement-learning
8
+ - multi-agent-reinforcement-learning
9
+ model-index:
10
+ - name: hivex-AWS-PPO-baseline-task-4-difficulty-1
11
+ results:
12
+ - task:
13
+ type: sub-task
14
+ name: protect_village
15
+ task-id: 4
16
+ difficulty-id: 1
17
+ dataset:
18
+ name: hivex-aerial-wildfire-suppression
19
+ type: hivex-aerial-wildfire-suppression
20
+ metrics:
21
+ - type: crash_count
22
+ value: 0.6166666761040688 +/- 0.22360679529482064
23
+ name: Crash Count
24
+ verified: true
25
+ - type: extinguishing_trees
26
+ value: 0.4333333492279053 +/- 1.9379256515825043
27
+ name: Extinguishing Trees
28
+ verified: true
29
+ - type: extinguishing_trees_reward
30
+ value: 2.166666603088379 +/- 9.689627618168343
31
+ name: Extinguishing Trees Reward
32
+ verified: true
33
+ - type: fire_out
34
+ value: 0.05833333507180214 +/- 0.11180340220699042
35
+ name: Fire Out
36
+ verified: true
37
+ - type: fire_too_close_to_city
38
+ value: 0.425 +/- 0.37257991018529735
39
+ name: Fire too Close to City
40
+ verified: true
41
+ - type: preparing_trees
42
+ value: 210.27499961853027 +/- 101.36103492999625
43
+ name: Preparing Trees
44
+ verified: true
45
+ - type: preparing_trees_reward
46
+ value: 210.27499961853027 +/- 101.36103492999625
47
+ name: Preparing Trees Reward
48
+ verified: true
49
+ - type: water_drop
50
+ value: 3.3749999940395354 +/- 2.19507356485688
51
+ name: Water Drop
52
+ verified: true
53
+ - type: water_pickup
54
+ value: 3.2916666686534883 +/- 2.260695820449611
55
+ name: Water Pickup
56
+ verified: true
57
+ - type: cumulative_reward
58
+ value: -69.60107120275498 +/- 211.17136032627994
59
+ name: Cumulative Reward
60
+ verified: true
61
+ ---
62
+
63
+ This model serves as the baseline for the **Aerial Wildfire Suppression** environment, trained and tested on task <code>4</code> with difficulty <code>1</code> using the Proximal Policy Optimization (PPO) algorithm.<br><br>
64
+
65
+ Environment: **Aerial Wildfire Suppression**<br>
66
+ Task: <code>4</code><br>
67
+ Difficulty: <code>1</code><br>
68
+ Algorithm: <code>PPO</code><br>
69
+ Episode Length: <code>3000</code><br>
70
+ Training <code>max_steps</code>: <code>1800000</code><br>
71
+ Testing <code>max_steps</code>: <code>180000</code><br><br>
72
+
73
+ Train & Test [Scripts](https://github.com/hivex-research/hivex)<br>
74
+ Download the [Environment](https://github.com/hivex-research/hivex-environments)
75
+
76
+ [hivex-paper]: https://arxiv.org/abs/2501.04180