Commit
·
083cb0a
verified
·
0
Parent(s):
1.0.0b1@gradio
Browse files- .gitattributes +35 -0
- .gitignore +240 -0
- README.md +75 -0
- app.py +441 -0
- requirements.txt +1 -0
.gitattributes
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
20 |
+
*.pickle filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.pkl filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
26 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
28 |
+
*.tar filter=lfs diff=lfs merge=lfs -text
|
29 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
30 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
31 |
+
*.wasm filter=lfs diff=lfs merge=lfs -text
|
32 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
33 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
+
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
.gitignore
ADDED
@@ -0,0 +1,240 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Byte-compiled / optimized / DLL files
|
2 |
+
__pycache__/
|
3 |
+
*.py[cod]
|
4 |
+
*$py.class
|
5 |
+
|
6 |
+
# C extensions
|
7 |
+
*.so
|
8 |
+
|
9 |
+
# Distribution / packaging
|
10 |
+
.Python
|
11 |
+
build/
|
12 |
+
develop-eggs/
|
13 |
+
dist/
|
14 |
+
downloads/
|
15 |
+
eggs/
|
16 |
+
.eggs/
|
17 |
+
lib/
|
18 |
+
lib64/
|
19 |
+
parts/
|
20 |
+
sdist/
|
21 |
+
var/
|
22 |
+
wheels/
|
23 |
+
share/python-wheels/
|
24 |
+
*.egg-info/
|
25 |
+
.installed.cfg
|
26 |
+
*.egg
|
27 |
+
MANIFEST
|
28 |
+
|
29 |
+
# PyInstaller
|
30 |
+
# Usually these files are written by a python script from a template
|
31 |
+
# before PyInstaller builds the exe, so as to inject date/other infos into it.
|
32 |
+
*.manifest
|
33 |
+
*.spec
|
34 |
+
|
35 |
+
# Installer logs
|
36 |
+
pip-log.txt
|
37 |
+
pip-delete-this-directory.txt
|
38 |
+
|
39 |
+
# Unit test / coverage reports
|
40 |
+
htmlcov/
|
41 |
+
.tox/
|
42 |
+
.nox/
|
43 |
+
.coverage
|
44 |
+
.coverage.*
|
45 |
+
.cache
|
46 |
+
nosetests.xml
|
47 |
+
coverage.xml
|
48 |
+
*.cover
|
49 |
+
*.py,cover
|
50 |
+
.hypothesis/
|
51 |
+
.pytest_cache/
|
52 |
+
cover/
|
53 |
+
|
54 |
+
# Translations
|
55 |
+
*.mo
|
56 |
+
*.pot
|
57 |
+
|
58 |
+
# Django stuff:
|
59 |
+
*.log
|
60 |
+
local_settings.py
|
61 |
+
db.sqlite3
|
62 |
+
db.sqlite3-journal
|
63 |
+
|
64 |
+
# Flask stuff:
|
65 |
+
instance/
|
66 |
+
.webassets-cache
|
67 |
+
|
68 |
+
# Scrapy stuff:
|
69 |
+
.scrapy
|
70 |
+
|
71 |
+
# Sphinx documentation
|
72 |
+
docs/_build/
|
73 |
+
|
74 |
+
# PyBuilder
|
75 |
+
.pybuilder/
|
76 |
+
target/
|
77 |
+
|
78 |
+
# Jupyter Notebook
|
79 |
+
.ipynb_checkpoints
|
80 |
+
|
81 |
+
# IPython
|
82 |
+
profile_default/
|
83 |
+
ipython_config.py
|
84 |
+
|
85 |
+
# pyenv
|
86 |
+
# For a library or package, you might want to ignore these files since the code is
|
87 |
+
# intended to run in multiple environments; otherwise, check them in:
|
88 |
+
# .python-version
|
89 |
+
|
90 |
+
# pipenv
|
91 |
+
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
|
92 |
+
# However, in case of collaboration, if having platform-specific dependencies or dependencies
|
93 |
+
# having no cross-platform support, pipenv may install dependencies that don't work, or not
|
94 |
+
# install all needed dependencies.
|
95 |
+
#Pipfile.lock
|
96 |
+
|
97 |
+
# poetry
|
98 |
+
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
|
99 |
+
# This is especially recommended for binary packages to ensure reproducibility, and is more
|
100 |
+
# commonly ignored for libraries.
|
101 |
+
# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
|
102 |
+
#poetry.lock
|
103 |
+
|
104 |
+
# pdm
|
105 |
+
# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
|
106 |
+
#pdm.lock
|
107 |
+
# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
|
108 |
+
# in version control.
|
109 |
+
# https://pdm.fming.dev/#use-with-ide
|
110 |
+
.pdm.toml
|
111 |
+
|
112 |
+
# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
|
113 |
+
__pypackages__/
|
114 |
+
|
115 |
+
# Celery stuff
|
116 |
+
celerybeat-schedule
|
117 |
+
celerybeat.pid
|
118 |
+
|
119 |
+
# SageMath parsed files
|
120 |
+
*.sage.py
|
121 |
+
|
122 |
+
# Environments
|
123 |
+
.env
|
124 |
+
.venv
|
125 |
+
env/
|
126 |
+
venv/
|
127 |
+
ENV/
|
128 |
+
env.bak/
|
129 |
+
venv.bak/
|
130 |
+
|
131 |
+
# Spyder project settings
|
132 |
+
.spyderproject
|
133 |
+
.spyproject
|
134 |
+
|
135 |
+
# Rope project settings
|
136 |
+
.ropeproject
|
137 |
+
|
138 |
+
# mkdocs documentation
|
139 |
+
/site
|
140 |
+
|
141 |
+
# mypy
|
142 |
+
.mypy_cache/
|
143 |
+
.dmypy.json
|
144 |
+
dmypy.json
|
145 |
+
|
146 |
+
# Pyre type checker
|
147 |
+
.pyre/
|
148 |
+
|
149 |
+
# pytype static type analyzer
|
150 |
+
.pytype/
|
151 |
+
|
152 |
+
# Cython debug symbols
|
153 |
+
cython_debug/
|
154 |
+
|
155 |
+
# PyCharm
|
156 |
+
# Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio, WebStorm and Rider
|
157 |
+
# Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839
|
158 |
+
|
159 |
+
# User-specific stuff
|
160 |
+
.idea/**/workspace.xml
|
161 |
+
.idea/**/tasks.xml
|
162 |
+
.idea/**/usage.statistics.xml
|
163 |
+
.idea/**/dictionaries
|
164 |
+
.idea/**/shelf
|
165 |
+
|
166 |
+
# AWS User-specific
|
167 |
+
.idea/**/aws.xml
|
168 |
+
|
169 |
+
# Generated files
|
170 |
+
.idea/**/contentModel.xml
|
171 |
+
|
172 |
+
# Sensitive or high-churn files
|
173 |
+
.idea/**/dataSources/
|
174 |
+
.idea/**/dataSources.ids
|
175 |
+
.idea/**/dataSources.local.xml
|
176 |
+
.idea/**/sqlDataSources.xml
|
177 |
+
.idea/**/dynamic.xml
|
178 |
+
.idea/**/uiDesigner.xml
|
179 |
+
.idea/**/dbnavigator.xml
|
180 |
+
|
181 |
+
# Gradle
|
182 |
+
.idea/**/gradle.xml
|
183 |
+
.idea/**/libraries
|
184 |
+
|
185 |
+
# Gradle and Maven with auto-import
|
186 |
+
# When using Gradle or Maven with auto-import, you should exclude module files,
|
187 |
+
# since they will be recreated, and may cause churn. Uncomment if using
|
188 |
+
# auto-import.
|
189 |
+
# .idea/artifacts
|
190 |
+
# .idea/compiler.xml
|
191 |
+
# .idea/jarRepositories.xml
|
192 |
+
# .idea/modules.xml
|
193 |
+
# .idea/*.iml
|
194 |
+
# .idea/modules
|
195 |
+
# *.iml
|
196 |
+
# *.ipr
|
197 |
+
|
198 |
+
# CMake
|
199 |
+
cmake-build-*/
|
200 |
+
|
201 |
+
# Mongo Explorer plugin
|
202 |
+
.idea/**/mongoSettings.xml
|
203 |
+
|
204 |
+
# File-based project format
|
205 |
+
*.iws
|
206 |
+
|
207 |
+
# IntelliJ
|
208 |
+
out/
|
209 |
+
|
210 |
+
# mpeltonen/sbt-idea plugin
|
211 |
+
.idea_modules/
|
212 |
+
|
213 |
+
# JIRA plugin
|
214 |
+
atlassian-ide-plugin.xml
|
215 |
+
|
216 |
+
# Cursive Clojure plugin
|
217 |
+
.idea/replstate.xml
|
218 |
+
|
219 |
+
# SonarLint plugin
|
220 |
+
.idea/sonarlint/
|
221 |
+
|
222 |
+
# Crashlytics plugin (for Android Studio and IntelliJ)
|
223 |
+
com_crashlytics_export_strings.xml
|
224 |
+
crashlytics.properties
|
225 |
+
crashlytics-build.properties
|
226 |
+
fabric.properties
|
227 |
+
|
228 |
+
# Editor-based Rest Client
|
229 |
+
.idea/httpRequests
|
230 |
+
|
231 |
+
# Android studio 3.1+ serialized cache file
|
232 |
+
.idea/caches/build_file_checksums.ser
|
233 |
+
|
234 |
+
.idea/
|
235 |
+
|
236 |
+
# Experimental Files
|
237 |
+
runs/
|
238 |
+
|
239 |
+
# MacOS Files
|
240 |
+
.DS_Store
|
README.md
ADDED
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
title: WildTorch
|
3 |
+
emoji: 🔥
|
4 |
+
colorFrom: red
|
5 |
+
colorTo: yellow
|
6 |
+
sdk: gradio
|
7 |
+
sdk_version: 4.26.0
|
8 |
+
app_file: app.py
|
9 |
+
pinned: false
|
10 |
+
license: mit
|
11 |
+
---
|
12 |
+
# WildTorch
|
13 |
+
|
14 |
+
[](https://github.com/pypa/hatch)
|
15 |
+
[](https://wildtorch.readthedocs.io/)
|
16 |
+
[](https://doi.org/10.5281/zenodo.10968647)
|
17 |
+
|
18 |
+
WildTorch: Leveraging GPU Acceleration for High-Fidelity, Stochastic Wildfire Simulations with PyTorch
|
19 |
+
|
20 |
+
GitHub: [https://github.com/xiazeyu/WildTorch](https://github.com/xiazeyu/WildTorch)
|
21 |
+
|
22 |
+
### Installation
|
23 |
+
|
24 |
+
Install with minimal dependencies:
|
25 |
+
|
26 |
+
```shell
|
27 |
+
pip install wildtorch
|
28 |
+
```
|
29 |
+
|
30 |
+
Install with full dependencies (includes visualization and logging):
|
31 |
+
|
32 |
+
```shell
|
33 |
+
pip install 'wildtorch[full]'
|
34 |
+
```
|
35 |
+
|
36 |
+
### Quick Start
|
37 |
+
|
38 |
+
```shell
|
39 |
+
pip install 'wildtorch[full]'
|
40 |
+
```
|
41 |
+
|
42 |
+
```python
|
43 |
+
import wildtorch as wt
|
44 |
+
|
45 |
+
wildfire_map = wt.dataset.generate_empty_dataset()
|
46 |
+
|
47 |
+
simulator = wt.WildTorchSimulator(
|
48 |
+
wildfire_map=wildfire_map,
|
49 |
+
simulator_constants=wt.SimulatorConstants(p_continue_burn=0.7),
|
50 |
+
initial_ignition=wt.utils.create_ignition(shape=wildfire_map[0].shape),
|
51 |
+
)
|
52 |
+
|
53 |
+
logger = wt.logger.Logger()
|
54 |
+
|
55 |
+
for i in range(200):
|
56 |
+
simulator.step()
|
57 |
+
logger.log_stats(
|
58 |
+
step=i,
|
59 |
+
num_cells_on_fire=wt.metrics.cell_on_fire(simulator.fire_state).item(),
|
60 |
+
num_cells_burned_out=wt.metrics.cell_burned_out(simulator.fire_state).item(),
|
61 |
+
)
|
62 |
+
logger.snapshot_simulation(simulator)
|
63 |
+
|
64 |
+
logger.save_logs()
|
65 |
+
logger.save_snapshots()
|
66 |
+
|
67 |
+
```
|
68 |
+
|
69 |
+
### Demo
|
70 |
+
|
71 |
+
See Our Live Demo at [Hugging Face Space](https://xiazeyu-wildtorch.hf.space/).
|
72 |
+
|
73 |
+
### API Documents
|
74 |
+
|
75 |
+
See at Our [Read the Docs](https://wildtorch.readthedocs.io/).
|
app.py
ADDED
@@ -0,0 +1,441 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import cast
|
2 |
+
import uuid
|
3 |
+
import os
|
4 |
+
|
5 |
+
import gradio as gr
|
6 |
+
import numpy as np
|
7 |
+
import pandas as pd
|
8 |
+
import torch
|
9 |
+
|
10 |
+
import wildtorch as wt
|
11 |
+
|
12 |
+
USE_OFFLINE_DATA = False
|
13 |
+
ENABLE_DOWNLOAD_SNAPSHOTS = False
|
14 |
+
|
15 |
+
if USE_OFFLINE_DATA:
|
16 |
+
wildfire_sim_maps = torch.load('wildfire_sim_maps.pt')
|
17 |
+
else:
|
18 |
+
wildfire_sim_maps = wt.dataset.load_wildfire_sim_maps()
|
19 |
+
# torch.save(wildfire_sim_maps, 'wildfire_sim_maps.pt')
|
20 |
+
|
21 |
+
DEFAULT_SHAPE = (512, 512)
|
22 |
+
DEFAULT_STATE = {
|
23 |
+
'ds': {
|
24 |
+
'name': None,
|
25 |
+
'shape': None,
|
26 |
+
'data': None,
|
27 |
+
},
|
28 |
+
'constants': {
|
29 |
+
'p_h': 0.58,
|
30 |
+
'c_1': 0.045,
|
31 |
+
'c_2': 0.131,
|
32 |
+
'a': 0.078,
|
33 |
+
'theta_w': 0,
|
34 |
+
'v': 10,
|
35 |
+
'p_firebreak': 0.9,
|
36 |
+
'p_continue_burn': 0.6,
|
37 |
+
'device': torch.device('cpu'),
|
38 |
+
'dtype': torch.float32,
|
39 |
+
},
|
40 |
+
'ignition': None,
|
41 |
+
'out_video_path': None,
|
42 |
+
'snapshots_path': None,
|
43 |
+
'checkpoint': None,
|
44 |
+
'logger': None,
|
45 |
+
}
|
46 |
+
|
47 |
+
with (gr.Blocks() as demo):
|
48 |
+
def remove_state_files(in_state):
|
49 |
+
if in_state['out_video_path'] is not None:
|
50 |
+
os.remove(in_state['out_video_path'])
|
51 |
+
if in_state['snapshots_path'] is not None:
|
52 |
+
os.remove(in_state['snapshots_path'])
|
53 |
+
|
54 |
+
|
55 |
+
state_var = gr.State(DEFAULT_STATE, delete_callback=remove_state_files)
|
56 |
+
with gr.Tabs(selected='tab_1') as tabs:
|
57 |
+
with gr.Tab("1. Datasets", interactive=True, id='tab_1') as tab_1:
|
58 |
+
sel_dataset = gr.Dropdown(cast(list, wildfire_sim_maps['name']) + ['empty'], label='Dataset')
|
59 |
+
with gr.Row() as shape_row:
|
60 |
+
sel_shape_h = gr.Number(label="Map Height", visible=False)
|
61 |
+
sel_shape_w = gr.Number(label="Map Width", visible=False)
|
62 |
+
|
63 |
+
with gr.Row() as preview_row:
|
64 |
+
canopy_img = gr.Image(label="canopy")
|
65 |
+
density_img = gr.Image(label="density")
|
66 |
+
slope_img = gr.Image(label="slope")
|
67 |
+
|
68 |
+
tab_1_confirm_btn = gr.Button("Confirm", interactive=True)
|
69 |
+
|
70 |
+
|
71 |
+
@tab_1_confirm_btn.click(inputs=[state_var], outputs=[state_var, tabs])
|
72 |
+
def jump_to_tab_2(in_state):
|
73 |
+
return in_state, gr.Tabs(selected='tab_2')
|
74 |
+
|
75 |
+
with gr.Tab("2. Simulation Constants and Initial Ignition", interactive=False, id='tab_2') as tab_2:
|
76 |
+
with gr.Row():
|
77 |
+
sel_p_h = gr.Slider(label="p_h",
|
78 |
+
info="The probability that a burnable cell adjacent to a burning cell will "
|
79 |
+
"catch fire at the next time step under normal conditions",
|
80 |
+
value=DEFAULT_STATE['constants']['p_h'], minimum=0, maximum=1, step=0.01,
|
81 |
+
interactive=True)
|
82 |
+
sel_p_continue_burn = gr.Slider(label="p_continue_burn",
|
83 |
+
info="The probability that a burning cell will continue to burn "
|
84 |
+
"at the next time step",
|
85 |
+
value=DEFAULT_STATE['constants']['p_continue_burn'], minimum=0,
|
86 |
+
maximum=1,
|
87 |
+
step=0.01,
|
88 |
+
interactive=True)
|
89 |
+
with gr.Row():
|
90 |
+
sel_a = gr.Slider(label="a",
|
91 |
+
info="The coefficient of ground elevation",
|
92 |
+
value=DEFAULT_STATE['constants']['a'], minimum=0, maximum=1, step=0.001,
|
93 |
+
interactive=True)
|
94 |
+
sel_p_firebreak = gr.Slider(label="p_firebreak",
|
95 |
+
info="The probability that a burnable cell will not catch fire even "
|
96 |
+
"if it is adjacent to a burning cell",
|
97 |
+
value=DEFAULT_STATE['constants']['p_firebreak'], minimum=0, maximum=1,
|
98 |
+
step=0.01, interactive=True)
|
99 |
+
with gr.Row():
|
100 |
+
sel_c_1 = gr.Slider(label="c_1",
|
101 |
+
info="The coefficient of wind velocity",
|
102 |
+
value=DEFAULT_STATE['constants']['c_1'], minimum=0, maximum=1, step=0.001,
|
103 |
+
interactive=True)
|
104 |
+
sel_c_2 = gr.Slider(label="c_2",
|
105 |
+
info="The coefficient of wind direction",
|
106 |
+
value=DEFAULT_STATE['constants']['c_2'], minimum=0, maximum=1, step=0.001,
|
107 |
+
interactive=True)
|
108 |
+
with gr.Row():
|
109 |
+
sel_theta_w = gr.Slider(label="theta_w",
|
110 |
+
info="The direction of the wind in degrees, measured clockwise from north",
|
111 |
+
value=DEFAULT_STATE['constants']['theta_w'], minimum=0, maximum=360, step=1,
|
112 |
+
interactive=True)
|
113 |
+
sel_v = gr.Slider(label="v",
|
114 |
+
info="The wind velocity, unit in m/s",
|
115 |
+
value=DEFAULT_STATE['constants']['v'], minimum=0, maximum=60, step=1,
|
116 |
+
interactive=True)
|
117 |
+
with gr.Row():
|
118 |
+
sel_device = gr.Dropdown(label="device", choices=['cpu', 'cuda', 'mps'],
|
119 |
+
info="The device to use",
|
120 |
+
value='cpu', allow_custom_value=True, interactive=True)
|
121 |
+
sel_dtype = gr.Dropdown(label="data type", choices=['float16', 'float32', 'float64'],
|
122 |
+
info="The data type to use",
|
123 |
+
value='float32', interactive=True)
|
124 |
+
|
125 |
+
|
126 |
+
@gr.on(triggers=[sel_p_h.input, sel_c_1.input, sel_c_2.input, sel_a.input,
|
127 |
+
sel_theta_w.input, sel_v.input, sel_p_firebreak.input,
|
128 |
+
sel_p_continue_burn.input, sel_device.input, sel_dtype.input],
|
129 |
+
inputs=[state_var, sel_p_h, sel_c_1, sel_c_2, sel_a, sel_theta_w, sel_v,
|
130 |
+
sel_p_firebreak, sel_p_continue_burn, sel_device, sel_dtype],
|
131 |
+
outputs=[state_var])
|
132 |
+
def update_constants_state(in_state, in_p_h, in_c_1, in_c_2, in_a, in_theta_w, in_v, in_p_firebreak,
|
133 |
+
in_p_continue_burn,
|
134 |
+
in_device, in_dtype):
|
135 |
+
in_state['constants']['p_h'] = in_p_h
|
136 |
+
in_state['constants']['c_1'] = in_c_1
|
137 |
+
in_state['constants']['c_2'] = in_c_2
|
138 |
+
in_state['constants']['a'] = in_a
|
139 |
+
in_state['constants']['theta_w'] = in_theta_w
|
140 |
+
in_state['constants']['v'] = in_v
|
141 |
+
in_state['constants']['p_firebreak'] = in_p_firebreak
|
142 |
+
in_state['constants']['p_continue_burn'] = in_p_continue_burn
|
143 |
+
in_state['constants']['device'] = torch.device(in_device)
|
144 |
+
in_state['constants']['dtype'] = {
|
145 |
+
'float16': torch.float16,
|
146 |
+
'float32': torch.float32,
|
147 |
+
'float64': torch.float64,
|
148 |
+
}[in_dtype]
|
149 |
+
return in_state
|
150 |
+
|
151 |
+
|
152 |
+
sel_ignition_mode = gr.Dropdown(label="Initial Ignition", choices=['random', 'center', 'custom'],
|
153 |
+
interactive=True)
|
154 |
+
|
155 |
+
gr.Markdown(
|
156 |
+
'to use custom ignition, please use the crop to fix the size, and then draw on the image. Please '
|
157 |
+
'click on the green button once done. Drawing on the black will be good choices.')
|
158 |
+
with gr.Row():
|
159 |
+
custom_ignition_paint = gr.Paint(label="custom ignition", image_mode='L', interactive=True,
|
160 |
+
brush=gr.Brush(default_size=3, color_mode='fixed'))
|
161 |
+
ignition_img_over_map = gr.Image(label="ignition over map")
|
162 |
+
|
163 |
+
|
164 |
+
@gr.on(triggers=[sel_shape_h.input, sel_shape_w.input],
|
165 |
+
inputs=[state_var, sel_shape_h, sel_shape_w],
|
166 |
+
outputs=[state_var, canopy_img, density_img, slope_img, tab_2, custom_ignition_paint])
|
167 |
+
def update_preview_row(in_state, in_h, in_w):
|
168 |
+
shape = in_h, in_w
|
169 |
+
data = wt.dataset.generate_empty_dataset(shape)
|
170 |
+
in_state['ds']['shape'] = shape
|
171 |
+
in_state['ds']['data'] = data
|
172 |
+
in_state['ignition'] = None
|
173 |
+
return in_state, gr.Image(wt.utils.colorize_array(np.array(data[0]))), gr.Image(
|
174 |
+
wt.utils.colorize_array(np.array(data[1]))), gr.Image(
|
175 |
+
wt.utils.colorize_array(np.array(data[2]))), gr.Tab(interactive=True), gr.ImageEditor(
|
176 |
+
crop_size=(shape[1], shape[0]))
|
177 |
+
|
178 |
+
|
179 |
+
@sel_dataset.change(inputs=[state_var, sel_dataset],
|
180 |
+
outputs=[state_var, sel_shape_h, sel_shape_w, tab_2, custom_ignition_paint,
|
181 |
+
canopy_img, density_img, slope_img])
|
182 |
+
def update_shape_row(in_state, in_dataset):
|
183 |
+
if in_dataset == 'empty':
|
184 |
+
shape = DEFAULT_SHAPE
|
185 |
+
data = wt.dataset.generate_empty_dataset(shape)
|
186 |
+
editable = True
|
187 |
+
else:
|
188 |
+
idx_dict = {item['name']: index for index, item in enumerate(wildfire_sim_maps)}
|
189 |
+
shape = tuple(cast(torch.Tensor, wildfire_sim_maps[idx_dict[in_dataset]]['shape']).tolist())
|
190 |
+
data = wt.dataset.transform_wildfire_sim_map(wildfire_sim_maps[idx_dict[in_dataset]])
|
191 |
+
editable = False
|
192 |
+
in_state['ds']['name'] = in_dataset
|
193 |
+
in_state['ds']['shape'] = shape
|
194 |
+
in_state['ds']['data'] = data
|
195 |
+
return in_state, gr.Number(value=shape[0], interactive=editable, visible=True), gr.Number(
|
196 |
+
value=shape[1],
|
197 |
+
interactive=editable,
|
198 |
+
visible=True), gr.Tab(interactive=True), gr.ImageEditor(interactive=True,
|
199 |
+
crop_size=(shape[1], shape[0])), gr.Image(
|
200 |
+
wt.utils.colorize_array(np.array(data[0]))), gr.Image(
|
201 |
+
wt.utils.colorize_array(np.array(data[1]))), gr.Image(
|
202 |
+
wt.utils.colorize_array(np.array(data[2])))
|
203 |
+
|
204 |
+
|
205 |
+
tab_2_confirm_btn = gr.Button("Confirm", interactive=False)
|
206 |
+
|
207 |
+
|
208 |
+
@sel_ignition_mode.input(
|
209 |
+
inputs=[state_var, sel_ignition_mode, custom_ignition_paint],
|
210 |
+
outputs=[state_var, ignition_img_over_map, tab_2_confirm_btn])
|
211 |
+
def update_ignition_img(in_state, in_mode, in_custom):
|
212 |
+
ignition = torch.zeros(in_state['ds']['shape'], dtype=torch.bool)
|
213 |
+
|
214 |
+
if in_mode == 'random':
|
215 |
+
ignition = wt.utils.create_ignition(shape=in_state['ds']['shape'], mode='random-single')
|
216 |
+
elif in_mode == 'center':
|
217 |
+
ignition = wt.utils.create_ignition(shape=in_state['ds']['shape'], mode='center')
|
218 |
+
elif in_mode == 'custom':
|
219 |
+
if in_custom['composite'] is not None:
|
220 |
+
ignition_ndarray = in_custom['composite'] != 0
|
221 |
+
ignition = torch.tensor(ignition_ndarray)
|
222 |
+
else:
|
223 |
+
return in_state, gr.Image(
|
224 |
+
wt.utils.colorize_array(wt.utils.compose_vis_wildfire_map(in_state['ds']['data']),
|
225 |
+
cmap='grey')), gr.Button(interactive=False)
|
226 |
+
|
227 |
+
in_state['ignition'] = ignition
|
228 |
+
ignition_ndarray = wt.utils.to_ndarray(ignition)
|
229 |
+
|
230 |
+
ignition__over_map = wt.utils.overlay_arrays(
|
231 |
+
wt.utils.colorize_array(ignition_ndarray),
|
232 |
+
wt.utils.colorize_array(wt.utils.compose_vis_wildfire_map(in_state['ds']['data']),
|
233 |
+
cmap='grey'),
|
234 |
+
0.5
|
235 |
+
)
|
236 |
+
|
237 |
+
return in_state, gr.Image((ignition__over_map * 255).astype(np.uint8)), gr.Button(interactive=True)
|
238 |
+
|
239 |
+
|
240 |
+
@custom_ignition_paint.change(
|
241 |
+
inputs=[state_var, custom_ignition_paint],
|
242 |
+
outputs=[state_var, sel_ignition_mode, ignition_img_over_map, tab_2_confirm_btn])
|
243 |
+
def update_ignition_img_over_map(in_state, in_custom):
|
244 |
+
if in_custom['composite'] is not None:
|
245 |
+
ignition_ndarray = in_custom['composite'] != 0
|
246 |
+
ignition = torch.tensor(ignition_ndarray)
|
247 |
+
else:
|
248 |
+
return in_state, gr.Dropdown(), gr.Image(), gr.Button()
|
249 |
+
in_state['ignition'] = ignition
|
250 |
+
|
251 |
+
ignition__over_map = wt.utils.overlay_arrays(
|
252 |
+
wt.utils.colorize_array(ignition_ndarray),
|
253 |
+
wt.utils.colorize_array(wt.utils.compose_vis_wildfire_map(in_state['ds']['data']),
|
254 |
+
cmap='grey'),
|
255 |
+
0.5
|
256 |
+
)
|
257 |
+
|
258 |
+
return in_state, gr.Dropdown(value='custom'), gr.Image(
|
259 |
+
(ignition__over_map * 255).astype(np.uint8)), gr.Button(interactive=True)
|
260 |
+
|
261 |
+
with gr.Tab("3. Simulation Control", interactive=False, id='tab_3') as tab_3:
|
262 |
+
@tab_2_confirm_btn.click(inputs=[state_var], outputs=[state_var, tabs, tab_3])
|
263 |
+
def update_tab_34_components(in_state):
|
264 |
+
return in_state, gr.Tabs(selected='tab_3'), gr.Tab(interactive=True)
|
265 |
+
|
266 |
+
|
267 |
+
with gr.Row():
|
268 |
+
with gr.Column():
|
269 |
+
gr.Markdown("## Memory Control")
|
270 |
+
checkpoint_cb = gr.Checkbox(label="Checkpoint -> Memory", value=False, interactive=True)
|
271 |
+
run_from_cp_cb = gr.Checkbox(label="Begin from Memory", value=False, interactive=True)
|
272 |
+
reset_btn = gr.Button("Reset Memory", interactive=True)
|
273 |
+
with gr.Column():
|
274 |
+
gr.Markdown("## Misc Control")
|
275 |
+
sel_steps = gr.Number(label="Number of Steps", value=200, minimum=1, step=1, interactive=True)
|
276 |
+
auto_run_cb = gr.Checkbox(label="Auto Run", value=False, interactive=True)
|
277 |
+
auto_reseed_cb = gr.Checkbox(label="Auto Regenerate Seed when open Tab", value=False,
|
278 |
+
interactive=True)
|
279 |
+
track_p_burn_cb = gr.Checkbox(label="Track p(burn), slow", value=False, interactive=True)
|
280 |
+
with gr.Column():
|
281 |
+
gr.Markdown("## Random Seed Control")
|
282 |
+
sel_seed = gr.Number(label="Random Seed", value=torch.Generator().seed(), minimum=0, step=1,
|
283 |
+
interactive=True)
|
284 |
+
random_seed_btn = gr.Button("Randomize Seed", interactive=True)
|
285 |
+
|
286 |
+
|
287 |
+
@random_seed_btn.click(inputs=[state_var], outputs=[state_var, sel_seed])
|
288 |
+
def randomize_seed(in_state):
|
289 |
+
return in_state, torch.Generator().seed()
|
290 |
+
|
291 |
+
with gr.Row():
|
292 |
+
run_btn = gr.Button("Run Simulation", interactive=True)
|
293 |
+
download_snap_btn = gr.DownloadButton(label="Download Snapshots", interactive=False, visible=False)
|
294 |
+
|
295 |
+
progress_bar = gr.Progress(track_tqdm=True)
|
296 |
+
|
297 |
+
with gr.Row():
|
298 |
+
output_video = gr.Video(label="Simulation Video", interactive=False, autoplay=True)
|
299 |
+
|
300 |
+
stats_plot = gr.LinePlot(title="Simulation Stats", interactive=True, height=600,
|
301 |
+
width=600, )
|
302 |
+
|
303 |
+
with gr.Tab("4. Advanced Simulation", interactive=False, id='tab_4') as tab_4:
|
304 |
+
|
305 |
+
sel_tab4_step = gr.Slider(label='Step', minimum=0, step=1, value=0, interactive=True)
|
306 |
+
with gr.Row():
|
307 |
+
cof_tb = gr.Textbox(label='cell_on_fire', interactive=False)
|
308 |
+
cbo_tb = gr.Textbox(label='cell_burned_out', interactive=False)
|
309 |
+
with gr.Row():
|
310 |
+
fire_state_img = gr.Image(label="Fire State", interactive=False)
|
311 |
+
p_burn_plot = gr.Image(label="p(burn)", interactive=False) # gr.Plot is bad at presenting
|
312 |
+
stats_df = gr.DataFrame()
|
313 |
+
|
314 |
+
|
315 |
+
@sel_tab4_step.input(inputs=[state_var, sel_tab4_step],
|
316 |
+
outputs=[state_var, fire_state_img, p_burn_plot, cof_tb, cbo_tb])
|
317 |
+
def update_tab4_step(in_state, in_user_step):
|
318 |
+
|
319 |
+
o_fsi, o_pbp = gr.Image(), gr.Image()
|
320 |
+
o_cof_tb, o_cbo_tb = gr.Textbox(), gr.Textbox()
|
321 |
+
|
322 |
+
if in_state['logger'] is not None:
|
323 |
+
snapshot = in_state['logger'].snapshots[in_user_step]
|
324 |
+
log = in_state['logger'].logs[in_user_step]
|
325 |
+
o_fsi = gr.Image(
|
326 |
+
value=wt.utils.colorize_array(wt.utils.compose_vis_fire_state(snapshot['fire_state'])))
|
327 |
+
if len(in_state['logger'].p_burns) > 0:
|
328 |
+
p_burn_arr = in_state['logger'].p_burns[in_user_step].cpu().numpy()
|
329 |
+
o_pbp = gr.Image(
|
330 |
+
value=wt.utils.colorize_array(p_burn_arr))
|
331 |
+
o_cof_tb = gr.Textbox(value=str(log['num_cells_on_fire']))
|
332 |
+
o_cbo_tb = gr.Textbox(value=str(log['num_cells_burned_out']))
|
333 |
+
|
334 |
+
return in_state, o_fsi, o_pbp, o_cof_tb, o_cbo_tb
|
335 |
+
|
336 |
+
|
337 |
+
@tab_3.select(
|
338 |
+
inputs=[state_var, auto_run_cb, sel_steps, sel_seed, auto_reseed_cb, checkpoint_cb, run_from_cp_cb,
|
339 |
+
track_p_burn_cb],
|
340 |
+
outputs=[state_var, output_video, tab_4, stats_plot, download_snap_btn, sel_seed, sel_tab4_step,
|
341 |
+
stats_df])
|
342 |
+
def auto_run_simulation(in_state, in_auto_run, in_steps, in_seed, in_auto_reseed, in_checkpoint_cb,
|
343 |
+
in_run_from_cp_cb, in_track_p_burn_cb):
|
344 |
+
o_s = in_state
|
345 |
+
o_v = gr.Video()
|
346 |
+
o_t = gr.Tab()
|
347 |
+
o_lp = gr.LinePlot()
|
348 |
+
o_dsb = gr.DownloadButton()
|
349 |
+
o_ts = gr.Slider()
|
350 |
+
o_sdf = pd.DataFrame()
|
351 |
+
if in_auto_reseed:
|
352 |
+
in_seed = torch.Generator().seed()
|
353 |
+
if in_auto_run:
|
354 |
+
o_s, o_v, o_t, o_lp, o_dsb, o_ts, o_sdf = run_simulation(in_state,
|
355 |
+
in_steps,
|
356 |
+
in_seed,
|
357 |
+
in_checkpoint_cb,
|
358 |
+
in_run_from_cp_cb,
|
359 |
+
in_track_p_burn_cb)
|
360 |
+
return o_s, o_v, o_t, o_lp, o_dsb, in_seed, o_ts, o_sdf
|
361 |
+
|
362 |
+
|
363 |
+
@reset_btn.click(inputs=[state_var],
|
364 |
+
outputs=[state_var])
|
365 |
+
def reset_simulation(in_state):
|
366 |
+
if in_state['checkpoint'] is not None:
|
367 |
+
in_state['checkpoint'] = None
|
368 |
+
gr.Info('Checkpoint Cleared.')
|
369 |
+
return in_state
|
370 |
+
|
371 |
+
|
372 |
+
@run_btn.click(inputs=[state_var, sel_steps, sel_seed, checkpoint_cb, run_from_cp_cb, track_p_burn_cb],
|
373 |
+
outputs=[state_var, output_video, tab_4, stats_plot, download_snap_btn, sel_tab4_step,
|
374 |
+
stats_df])
|
375 |
+
def run_simulation(in_state, in_steps, in_seed, in_checkpoint_cb, in_run_from_cp_cb, in_track_p_burn_cb,
|
376 |
+
in_progress=gr.Progress(track_tqdm=True)):
|
377 |
+
if in_state['out_video_path'] is None:
|
378 |
+
in_state['out_video_path'] = f'runs/{str(uuid.uuid4())}.mp4'
|
379 |
+
simulator = wt.WildTorchSimulator(
|
380 |
+
wildfire_map=in_state['ds']['data'],
|
381 |
+
simulator_constants=wt.SimulatorConstants(
|
382 |
+
p_h=in_state['constants']['p_h'],
|
383 |
+
c_1=in_state['constants']['c_1'],
|
384 |
+
c_2=in_state['constants']['c_2'],
|
385 |
+
a=in_state['constants']['a'],
|
386 |
+
theta_w=in_state['constants']['theta_w'],
|
387 |
+
v=in_state['constants']['v'],
|
388 |
+
p_firebreak=in_state['constants']['p_firebreak'],
|
389 |
+
p_continue_burn=in_state['constants']['p_continue_burn'],
|
390 |
+
device=in_state['constants']['device'],
|
391 |
+
dtype=in_state['constants']['dtype'],
|
392 |
+
),
|
393 |
+
maximum_step=in_steps,
|
394 |
+
initial_ignition=in_state['ignition'],
|
395 |
+
seed=in_seed,
|
396 |
+
)
|
397 |
+
|
398 |
+
if in_state['checkpoint'] is not None and in_run_from_cp_cb:
|
399 |
+
simulator.load_checkpoint(in_state['checkpoint'], restore_seed=False)
|
400 |
+
|
401 |
+
logger = wt.logger.Logger(disable_writing=True, verbose=False)
|
402 |
+
|
403 |
+
for i in in_progress.tqdm(range(in_steps)):
|
404 |
+
simulator.step()
|
405 |
+
logger.snapshot_simulation(simulator)
|
406 |
+
logger.log_stats(
|
407 |
+
step=i,
|
408 |
+
num_cells_on_fire=wt.metrics.cell_on_fire(simulator.fire_state).item(),
|
409 |
+
num_cells_burned_out=wt.metrics.cell_burned_out(simulator.fire_state).item(),
|
410 |
+
)
|
411 |
+
if in_track_p_burn_cb:
|
412 |
+
logger.log_p_burn(simulator)
|
413 |
+
|
414 |
+
gr.Info('Simulation Completed. Generating Video ...')
|
415 |
+
|
416 |
+
in_state['logger'] = logger
|
417 |
+
|
418 |
+
if in_checkpoint_cb:
|
419 |
+
in_state['checkpoint'] = simulator.checkpoint
|
420 |
+
|
421 |
+
if ENABLE_DOWNLOAD_SNAPSHOTS:
|
422 |
+
logger.snapshots_filepath = in_state['snapshots_path'] = f'runs/{str(uuid.uuid4())}.pt'
|
423 |
+
logger.save_snapshots()
|
424 |
+
can_download_snapshots = True
|
425 |
+
else:
|
426 |
+
can_download_snapshots = False
|
427 |
+
|
428 |
+
wt.utils.animate_snapshots(logger.snapshots, simulator.wildfire_map,
|
429 |
+
out_filename=in_state['out_video_path'])
|
430 |
+
|
431 |
+
m_stats_df = pd.DataFrame(logger.logs)
|
432 |
+
m_stats_df = m_stats_df.melt(id_vars=["step"], var_name="key", value_name="value")
|
433 |
+
|
434 |
+
o_stats_df = pd.DataFrame(logger.logs)
|
435 |
+
return in_state, gr.Video(value=in_state['out_video_path']), gr.Tab(interactive=True), gr.LinePlot(
|
436 |
+
m_stats_df, x='step', y='value', color="key", color_legend_position="bottom",
|
437 |
+
tooltip=["step", "key", "value"], container=False, ), gr.DownloadButton(
|
438 |
+
value=in_state['snapshots_path'], interactive=can_download_snapshots,
|
439 |
+
visible=can_download_snapshots), gr.Slider(maximum=in_steps - 1), o_stats_df
|
440 |
+
|
441 |
+
demo.queue().launch()
|
requirements.txt
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
wildtorch[full]
|