Repository navigation
Expand file tree
/
Copy pathopenenv.yaml
More file actions
74 lines (64 loc) · 1.74 KB
/
Copy pathopenenv.yaml
File metadata and controls
74 lines (64 loc) · 1.74 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
name: WildfireContainment-v0
version: 0.1.0
description: >
A dynamic wildfire suppression environment for agentic RL.
An AI agent controls 1 Air Tanker and 2 Ground Crews on a 20x20 grid
to contain fire spread, protect structures, and manage limited resources
under wind and stochastic spot-fire conditions.
author: "Aryan"
license: "BSD-3-Clause"
env_type: "GridEnvironment"
endpoints:
health: /health
reset: /reset
step: /step
state: /state
schema: /schema
action_space:
type: "object"
description: "Actions for 3 units: [Tanker, Crew1, Crew2]"
observation_space:
type: "object"
fields:
fuel_grid: "List[List[float]] 20x20"
fire_grid: "List[List[float]] 20x20"
moisture_grid: "List[List[float]] 20x20"
structure_grid: "List[List[int]] 20x20"
units: "List[UnitStatus]"
wind_dir: "int 0-7"
wind_speed: "float"
tasks:
- id: easy
description: "Low fire spread (ignition 0.05), 5 structures, slow wind."
steps: 3
ideal_action: "contain"
- id: medium
description: "Moderate fire (ignition 0.12), 10 structures, variable wind."
steps: 3
ideal_action: "suppress"
- id: hard
description: "Intense spread (ignition 0.20), frequent spot fires, 20 structures."
steps: 3
ideal_action: "protect"
- id: extreme
description: "Extreme spread (ignition 0.25), rapid spot fires, 25 structures."
steps: 3
ideal_action: "protect"
scoring:
reward_range: [0.01, 0.99]
success_threshold: 0.5
score_formula: mean(step_rewards)
constraints:
max_runtime_seconds: 1200
max_memory_gb: 8
max_vcpu: 2
entrypoint: "server/app.py"
port: 7860
docker:
image: "python:3.11-slim"
inference:
script: "inference.py"
env_vars:
- API_BASE_URL
- MODEL_NAME
- HF_TOKEN