├── .github ├── CODEOWNERS └── ISSUE_TEMPLATE │ ├── bug_report.md │ ├── config.yml │ ├── propose_addition.md │ └── propose_change.md ├── .gitignore ├── CHANGELOG.md ├── CODE_OF_CONDUCT.md ├── CONTRIBUTING.md ├── LICENSE ├── README.md ├── examples └── template.json ├── fragments ├── environment │ ├── README.md │ └── environment-fragment.md ├── runtime │ ├── README.md │ └── runtime-fragment.md ├── training │ ├── README.md │ └── training-fragment.md └── usage │ ├── README.md │ └── usage-fragment.md └── model-metadata ├── README.md └── model-metadata.md /.github/CODEOWNERS: -------------------------------------------------------------------------------- 1 | # Default owners 2 | * @duckontheweb 3 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/bug_report.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Bug report 3 | about: Report a bug/inconsistency in the spec 4 | title: "[TITLE]" 5 | labels: bug 6 | assignees: duckontheweb 7 | 8 | --- 9 | 10 | ### Problem Description 11 | 12 | *Please provide a clear and concise description of the problem. If possible, link to specific locations within the 13 | spec for reference.* 14 | 15 | ### Proposed Solution 16 | 17 | *How do you proposed we solve this problem within the spec?* 18 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/config.yml: -------------------------------------------------------------------------------- 1 | blank_issues_enabled: true 2 | contact_links: 3 | - name: Radiant MLHub Slack Organization 4 | url: https://mlhubearth.slack.com/app_redirect?channel=geo-ml-model-catalog 5 | about: Slack channel for discussion of this spec -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/propose_addition.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Propose Addition 3 | about: Propose an addition to the spec 4 | title: "[TITLE]" 5 | labels: addition 6 | assignees: duckontheweb 7 | 8 | --- 9 | 10 | ### Problem Description 11 | 12 | *Please provide a brief description of the problem or gap that this proposal addresses. Include concrete examples 13 | whenever possible.* 14 | 15 | ### Proposal 16 | 17 | *Please provide a description of the additional field(s) that you are proposing within the spec. You may include 18 | detailed descriptions of the field names, types, etc. if you wish, but more general proposals are also welcome.* -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/propose_change.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Propose Change 3 | about: Propose a change to an existing part of the spec 4 | title: "[TITLE]" 5 | labels: change 6 | assignees: duckontheweb 7 | 8 | --- 9 | 10 | ### Problem Description 11 | 12 | *Please provide a brief description of the problem or gap that this proposal addresses. Include concrete examples 13 | and/or links to existing portions of the spec whenever possible.* 14 | 15 | ### Proposal 16 | 17 | *Please provide a description of the changes that you are proposing within the spec. You may include 18 | detailed descriptions of changes to field names, types, etc. if you wish, but more general proposals are also welcome.* 19 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | .idea 2 | 3 | # Created by .ignore support plugin (hsz.mobi) 4 | ### VisualStudio template 5 | ## Ignore Visual Studio temporary files, build results, and 6 | ## files generated by popular Visual Studio add-ons. 7 | ## 8 | ## Get latest from https://github.com/github/gitignore/blob/master/VisualStudio.gitignore 9 | 10 | # User-specific files 11 | *.rsuser 12 | *.suo 13 | *.user 14 | *.userosscache 15 | *.sln.docstates 16 | 17 | # User-specific files (MonoDevelop/Xamarin Studio) 18 | *.userprefs 19 | 20 | # Mono auto generated files 21 | mono_crash.* 22 | 23 | # Build results 24 | [Dd]ebug/ 25 | [Dd]ebugPublic/ 26 | [Rr]elease/ 27 | [Rr]eleases/ 28 | x64/ 29 | x86/ 30 | [Ww][Ii][Nn]32/ 31 | [Aa][Rr][Mm]/ 32 | [Aa][Rr][Mm]64/ 33 | bld/ 34 | [Bb]in/ 35 | [Oo]bj/ 36 | [Ll]og/ 37 | [Ll]ogs/ 38 | 39 | # Visual Studio 2015/2017 cache/options directory 40 | .vs/ 41 | # Uncomment if you have tasks that create the project's static files in wwwroot 42 | #wwwroot/ 43 | 44 | # Visual Studio 2017 auto generated files 45 | Generated\ Files/ 46 | 47 | # MSTest test Results 48 | [Tt]est[Rr]esult*/ 49 | [Bb]uild[Ll]og.* 50 | 51 | # NUnit 52 | *.VisualState.xml 53 | TestResult.xml 54 | nunit-*.xml 55 | 56 | # Build Results of an ATL Project 57 | [Dd]ebugPS/ 58 | [Rr]eleasePS/ 59 | dlldata.c 60 | 61 | # Benchmark Results 62 | BenchmarkDotNet.Artifacts/ 63 | 64 | # .NET Core 65 | project.lock.json 66 | project.fragment.lock.json 67 | artifacts/ 68 | 69 | # ASP.NET Scaffolding 70 | ScaffoldingReadMe.txt 71 | 72 | # StyleCop 73 | StyleCopReport.xml 74 | 75 | # Files built by Visual Studio 76 | *_i.c 77 | *_p.c 78 | *_h.h 79 | *.ilk 80 | *.meta 81 | *.obj 82 | *.iobj 83 | *.pch 84 | *.pdb 85 | *.ipdb 86 | *.pgc 87 | *.pgd 88 | *.rsp 89 | *.sbr 90 | *.tlb 91 | *.tli 92 | *.tlh 93 | *.tmp 94 | *.tmp_proj 95 | *_wpftmp.csproj 96 | *.log 97 | *.vspscc 98 | *.vssscc 99 | .builds 100 | *.pidb 101 | *.svclog 102 | *.scc 103 | 104 | # Chutzpah Test files 105 | _Chutzpah* 106 | 107 | # Visual C++ cache files 108 | ipch/ 109 | *.aps 110 | *.ncb 111 | *.opendb 112 | *.opensdf 113 | *.sdf 114 | *.cachefile 115 | *.VC.db 116 | *.VC.VC.opendb 117 | 118 | # Visual Studio profiler 119 | *.psess 120 | *.vsp 121 | *.vspx 122 | *.sap 123 | 124 | # Visual Studio Trace Files 125 | *.e2e 126 | 127 | # TFS 2012 Local Workspace 128 | $tf/ 129 | 130 | # Guidance Automation Toolkit 131 | *.gpState 132 | 133 | # ReSharper is a .NET coding add-in 134 | _ReSharper*/ 135 | *.[Rr]e[Ss]harper 136 | *.DotSettings.user 137 | 138 | # TeamCity is a build add-in 139 | _TeamCity* 140 | 141 | # DotCover is a Code Coverage Tool 142 | *.dotCover 143 | 144 | # AxoCover is a Code Coverage Tool 145 | .axoCover/* 146 | !.axoCover/settings.json 147 | 148 | # Coverlet is a free, cross platform Code Coverage Tool 149 | coverage*.json 150 | coverage*.xml 151 | coverage*.info 152 | 153 | # Visual Studio code coverage results 154 | *.coverage 155 | *.coveragexml 156 | 157 | # NCrunch 158 | _NCrunch_* 159 | .*crunch*.local.xml 160 | nCrunchTemp_* 161 | 162 | # MightyMoose 163 | *.mm.* 164 | AutoTest.Net/ 165 | 166 | # Web workbench (sass) 167 | .sass-cache/ 168 | 169 | # Installshield output folder 170 | [Ee]xpress/ 171 | 172 | # DocProject is a documentation generator add-in 173 | DocProject/buildhelp/ 174 | DocProject/Help/*.HxT 175 | DocProject/Help/*.HxC 176 | DocProject/Help/*.hhc 177 | DocProject/Help/*.hhk 178 | DocProject/Help/*.hhp 179 | DocProject/Help/Html2 180 | DocProject/Help/html 181 | 182 | # Click-Once directory 183 | publish/ 184 | 185 | # Publish Web Output 186 | *.[Pp]ublish.xml 187 | *.azurePubxml 188 | # Note: Comment the next line if you want to checkin your web deploy settings, 189 | # but database connection strings (with potential passwords) will be unencrypted 190 | *.pubxml 191 | *.publishproj 192 | 193 | # Microsoft Azure Web App publish settings. Comment the next line if you want to 194 | # checkin your Azure Web App publish settings, but sensitive information contained 195 | # in these scripts will be unencrypted 196 | PublishScripts/ 197 | 198 | # NuGet Packages 199 | *.nupkg 200 | # NuGet Symbol Packages 201 | *.snupkg 202 | # The packages folder can be ignored because of Package Restore 203 | **/[Pp]ackages/* 204 | # except build/, which is used as an MSBuild target. 205 | !**/[Pp]ackages/build/ 206 | # Uncomment if necessary however generally it will be regenerated when needed 207 | #!**/[Pp]ackages/repositories.config 208 | # NuGet v3's project.json files produces more ignorable files 209 | *.nuget.props 210 | *.nuget.targets 211 | 212 | # Microsoft Azure Build Output 213 | csx/ 214 | *.build.csdef 215 | 216 | # Microsoft Azure Emulator 217 | ecf/ 218 | rcf/ 219 | 220 | # Windows Store app package directories and files 221 | AppPackages/ 222 | BundleArtifacts/ 223 | Package.StoreAssociation.xml 224 | _pkginfo.txt 225 | *.appx 226 | *.appxbundle 227 | *.appxupload 228 | 229 | # Visual Studio cache files 230 | # files ending in .cache can be ignored 231 | *.[Cc]ache 232 | # but keep track of directories ending in .cache 233 | !?*.[Cc]ache/ 234 | 235 | # Others 236 | ClientBin/ 237 | ~$* 238 | *~ 239 | *.dbmdl 240 | *.dbproj.schemaview 241 | *.jfm 242 | *.pfx 243 | *.publishsettings 244 | orleans.codegen.cs 245 | 246 | # Including strong name files can present a security risk 247 | # (https://github.com/github/gitignore/pull/2483#issue-259490424) 248 | #*.snk 249 | 250 | # Since there are multiple workflows, uncomment next line to ignore bower_components 251 | # (https://github.com/github/gitignore/pull/1529#issuecomment-104372622) 252 | #bower_components/ 253 | 254 | # RIA/Silverlight projects 255 | Generated_Code/ 256 | 257 | # Backup & report files from converting an old project file 258 | # to a newer Visual Studio version. Backup files are not needed, 259 | # because we have git ;-) 260 | _UpgradeReport_Files/ 261 | Backup*/ 262 | UpgradeLog*.XML 263 | UpgradeLog*.htm 264 | ServiceFabricBackup/ 265 | *.rptproj.bak 266 | 267 | # SQL Server files 268 | *.mdf 269 | *.ldf 270 | *.ndf 271 | 272 | # Business Intelligence projects 273 | *.rdl.data 274 | *.bim.layout 275 | *.bim_*.settings 276 | *.rptproj.rsuser 277 | *- [Bb]ackup.rdl 278 | *- [Bb]ackup ([0-9]).rdl 279 | *- [Bb]ackup ([0-9][0-9]).rdl 280 | 281 | # Microsoft Fakes 282 | FakesAssemblies/ 283 | 284 | # GhostDoc plugin setting file 285 | *.GhostDoc.xml 286 | 287 | # Node.js Tools for Visual Studio 288 | .ntvs_analysis.dat 289 | node_modules/ 290 | 291 | # Visual Studio 6 build log 292 | *.plg 293 | 294 | # Visual Studio 6 workspace options file 295 | *.opt 296 | 297 | # Visual Studio 6 auto-generated workspace file (contains which files were open etc.) 298 | *.vbw 299 | 300 | # Visual Studio LightSwitch build output 301 | **/*.HTMLClient/GeneratedArtifacts 302 | **/*.DesktopClient/GeneratedArtifacts 303 | **/*.DesktopClient/ModelManifest.xml 304 | **/*.Server/GeneratedArtifacts 305 | **/*.Server/ModelManifest.xml 306 | _Pvt_Extensions 307 | 308 | # Paket dependency manager 309 | .paket/paket.exe 310 | paket-files/ 311 | 312 | # FAKE - F# Make 313 | .fake/ 314 | 315 | # CodeRush personal settings 316 | .cr/personal 317 | 318 | # Python Tools for Visual Studio (PTVS) 319 | __pycache__/ 320 | *.pyc 321 | 322 | # Cake - Uncomment if you are using it 323 | # tools/** 324 | # !tools/packages.config 325 | 326 | # Tabs Studio 327 | *.tss 328 | 329 | # Telerik's JustMock configuration file 330 | *.jmconfig 331 | 332 | # BizTalk build output 333 | *.btp.cs 334 | *.btm.cs 335 | *.odx.cs 336 | *.xsd.cs 337 | 338 | # OpenCover UI analysis results 339 | OpenCover/ 340 | 341 | # Azure Stream Analytics local run output 342 | ASALocalRun/ 343 | 344 | # MSBuild Binary and Structured Log 345 | *.binlog 346 | 347 | # NVidia Nsight GPU debugger configuration file 348 | *.nvuser 349 | 350 | # MFractors (Xamarin productivity tool) working folder 351 | .mfractor/ 352 | 353 | # Local History for Visual Studio 354 | .localhistory/ 355 | 356 | # BeatPulse healthcheck temp database 357 | healthchecksdb 358 | 359 | # Backup folder for Package Reference Convert tool in Visual Studio 2017 360 | MigrationBackup/ 361 | 362 | # Ionide (cross platform F# VS Code tools) working folder 363 | .ionide/ 364 | 365 | # Fody - auto-generated XML schema 366 | FodyWeavers.xsd 367 | 368 | ### macOS template 369 | # General 370 | .DS_Store 371 | .AppleDouble 372 | .LSOverride 373 | 374 | # Icon must end with two \r 375 | Icon 376 | 377 | # Thumbnails 378 | ._* 379 | 380 | # Files that might appear in the root of a volume 381 | .DocumentRevisions-V100 382 | .fseventsd 383 | .Spotlight-V100 384 | .TemporaryItems 385 | .Trashes 386 | .VolumeIcon.icns 387 | .com.apple.timemachine.donotpresent 388 | 389 | # Directories potentially created on remote AFP share 390 | .AppleDB 391 | .AppleDesktop 392 | Network Trash Folder 393 | Temporary Items 394 | .apdisk 395 | 396 | ### Python template 397 | # Byte-compiled / optimized / DLL files 398 | __pycache__/ 399 | *.py[cod] 400 | *$py.class 401 | 402 | # C extensions 403 | *.so 404 | 405 | # Distribution / packaging 406 | .Python 407 | build/ 408 | develop-eggs/ 409 | dist/ 410 | downloads/ 411 | eggs/ 412 | .eggs/ 413 | lib/ 414 | lib64/ 415 | parts/ 416 | sdist/ 417 | var/ 418 | wheels/ 419 | share/python-wheels/ 420 | *.egg-info/ 421 | .installed.cfg 422 | *.egg 423 | MANIFEST 424 | 425 | # PyInstaller 426 | # Usually these files are written by a python script from a template 427 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 428 | *.manifest 429 | *.spec 430 | 431 | # Installer logs 432 | pip-log.txt 433 | pip-delete-this-directory.txt 434 | 435 | # Unit test / coverage reports 436 | htmlcov/ 437 | .tox/ 438 | .nox/ 439 | .coverage 440 | .coverage.* 441 | .cache 442 | nosetests.xml 443 | coverage.xml 444 | *.cover 445 | *.py,cover 446 | .hypothesis/ 447 | .pytest_cache/ 448 | cover/ 449 | 450 | # Translations 451 | *.mo 452 | *.pot 453 | 454 | # Django stuff: 455 | *.log 456 | local_settings.py 457 | db.sqlite3 458 | db.sqlite3-journal 459 | 460 | # Flask stuff: 461 | instance/ 462 | .webassets-cache 463 | 464 | # Scrapy stuff: 465 | .scrapy 466 | 467 | # Sphinx documentation 468 | docs/_build/ 469 | 470 | # PyBuilder 471 | .pybuilder/ 472 | target/ 473 | 474 | # Jupyter Notebook 475 | .ipynb_checkpoints 476 | 477 | # IPython 478 | profile_default/ 479 | ipython_config.py 480 | 481 | # pyenv 482 | # For a library or package, you might want to ignore these files since the code is 483 | # intended to run in multiple environments; otherwise, check them in: 484 | # .python-version 485 | 486 | # pipenv 487 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 488 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 489 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 490 | # install all needed dependencies. 491 | #Pipfile.lock 492 | 493 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 494 | __pypackages__/ 495 | 496 | # Celery stuff 497 | celerybeat-schedule 498 | celerybeat.pid 499 | 500 | # SageMath parsed files 501 | *.sage.py 502 | 503 | # Environments 504 | .env 505 | .venv 506 | env/ 507 | venv/ 508 | ENV/ 509 | env.bak/ 510 | venv.bak/ 511 | 512 | # Spyder project settings 513 | .spyderproject 514 | .spyproject 515 | 516 | # Rope project settings 517 | .ropeproject 518 | 519 | # mkdocs documentation 520 | /site 521 | 522 | # mypy 523 | .mypy_cache/ 524 | .dmypy.json 525 | dmypy.json 526 | 527 | # Pyre type checker 528 | .pyre/ 529 | 530 | # pytype static type analyzer 531 | .pytype/ 532 | 533 | # Cython debug symbols 534 | cython_debug/ 535 | 536 | ### Windows template 537 | # Windows thumbnail cache files 538 | Thumbs.db 539 | Thumbs.db:encryptable 540 | ehthumbs.db 541 | ehthumbs_vista.db 542 | 543 | # Dump file 544 | *.stackdump 545 | 546 | # Folder config file 547 | [Dd]esktop.ini 548 | 549 | # Recycle Bin used on file shares 550 | $RECYCLE.BIN/ 551 | 552 | # Windows Installer files 553 | *.cab 554 | *.msi 555 | *.msix 556 | *.msm 557 | *.msp 558 | 559 | # Windows shortcuts 560 | *.lnk 561 | 562 | ### Linux template 563 | *~ 564 | 565 | # temporary files which can be created if a process still has a handle open of a deleted file 566 | .fuse_hidden* 567 | 568 | # KDE directory preferences 569 | .directory 570 | 571 | # Linux trash folder which might appear on any partition or disk 572 | .Trash-* 573 | 574 | # .nfs files are created when an open file is removed but is still being accessed 575 | .nfs* 576 | 577 | -------------------------------------------------------------------------------- /CHANGELOG.md: -------------------------------------------------------------------------------- 1 | # Changelog 2 | 3 | All notable changes to this project will be documented in this file. 4 | 5 | The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/) 6 | and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.html). 7 | 8 | ## [Unreleased] 9 | 10 | ### Added 11 | 12 | ### Changed 13 | 14 | ### Removed 15 | 16 | ### Fixed 17 | 18 | ## [v0.1.0] - 2021-05-19 19 | 20 | Initial release. 21 | 22 | 23 | [Unreleased]: 24 | [v0.1.0]: -------------------------------------------------------------------------------- /CODE_OF_CONDUCT.md: -------------------------------------------------------------------------------- 1 | 2 | # Contributor Covenant Code of Conduct 3 | 4 | ## Our Pledge 5 | 6 | In the interest of fostering an open and welcoming environment, we as 7 | contributors and maintainers pledge to make participation in our project and 8 | our community a harassment-free experience for everyone, regardless of age, body 9 | size, disability, ethnicity, sex characteristics, gender identity and expression, 10 | level of experience, education, socio-economic status, nationality, personal 11 | appearance, race, religion, or sexual identity and orientation. 12 | 13 | ## Our Standards 14 | 15 | Examples of behavior that contributes to creating a positive environment 16 | include: 17 | 18 | * Using welcoming and inclusive language 19 | * Being respectful of differing viewpoints and experiences 20 | * Gracefully accepting constructive criticism 21 | * Focusing on what is best for the community 22 | * Showing empathy towards other community members 23 | 24 | Examples of unacceptable behavior by participants include: 25 | 26 | * The use of sexualized language or imagery and unwelcome sexual attention or 27 | advances 28 | * Trolling, insulting/derogatory comments, and personal or political attacks 29 | * Public or private harassment 30 | * Publishing others' private information, such as a physical or electronic 31 | address, without explicit permission 32 | * Other conduct which could reasonably be considered inappropriate in a 33 | professional setting 34 | 35 | ## Our Responsibilities 36 | 37 | Project maintainers are responsible for clarifying the standards of acceptable 38 | behavior and are expected to take appropriate and fair corrective action in 39 | response to any instances of unacceptable behavior. 40 | 41 | Project maintainers have the right and responsibility to remove, edit, or 42 | reject comments, commits, code, wiki edits, issues, and other contributions 43 | that are not aligned to this Code of Conduct, or to ban temporarily or 44 | permanently any contributor for other behaviors that they deem inappropriate, 45 | threatening, offensive, or harmful. 46 | 47 | ## Scope 48 | 49 | This Code of Conduct applies within all project spaces, and it also applies when 50 | an individual is representing the project or its community in public spaces. 51 | Examples of representing a project or community include using an official 52 | project e-mail address, posting via an official social media account, or acting 53 | as an appointed representative at an online or offline event. Representation of 54 | a project may be further defined and clarified by project maintainers. 55 | 56 | ## Enforcement 57 | 58 | Instances of abusive, harassing, or otherwise unacceptable behavior may be 59 | reported by contacting the project team at [support@radiant.earth](mailto:support@radiant.earth). All 60 | complaints will be reviewed and investigated and will result in a response that 61 | is deemed necessary and appropriate to the circumstances. The project team is 62 | obligated to maintain confidentiality with regard to the reporter of an incident. 63 | Further details of specific enforcement policies may be posted separately. 64 | 65 | Project maintainers who do not follow or enforce the Code of Conduct in good 66 | faith may face temporary or permanent repercussions as determined by other 67 | members of the project's leadership. 68 | 69 | ## Attribution 70 | 71 | This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 1.4, 72 | available at https://www.contributor-covenant.org/version/1/4/code-of-conduct.html 73 | 74 | [homepage]: https://www.contributor-covenant.org 75 | 76 | For answers to common questions about this code of conduct, see 77 | https://www.contributor-covenant.org/faq 78 | -------------------------------------------------------------------------------- /CONTRIBUTING.md: -------------------------------------------------------------------------------- 1 | # Contributing 2 | 3 | Please review our [Code of Conduct](./CODE_OF_CONDUCT.md) before contributing to the spec. 4 | 5 | ## Spec Development 6 | 7 | The primary methods for contributing to the spec itself are through Issues and/or Pull Requests in this repo, and 8 | everyone is encouraged to contribute through either of these methods. Please be sure to check existing Issues and Pull 9 | Requests to see if someone has already addressed your question or problem. 10 | 11 | ## Design and Direction 12 | 13 | In addition to direct contributions to the spec itself, we welcome input and feedback on the high-level design from 14 | users and community members. The best way to contribute to this discussion is on the 15 | [`#geo-ml-model-catalog`](https://mlhubearth.slack.com/app_redirect?channel=geo-ml-model-catalog) Slack channel. 16 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright [yyyy] [name of copyright owner] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Geospatial ML Model Catalog \(GMLMC\) Specification 2 | 3 | | :exclamation: DEPRECATED | 4 | |:----------------------------| 5 | | Work on a geospatial model cataloging spec has been moved to the [STAC ML Model Extension](https://github.com/stac-extensions/ml-model). This repository has been archived and will no longer be actively maintained. | 6 | 7 | ## Purpose 8 | 9 | The number and variety of machine learning (ML) models that utilize geospatial data (e.g. satellite imagery, airborne 10 | observations, and physical model estimates) is growing rapidly. Many of these ML models have the potential to be 11 | deployed to derive useful insights for many global problems, or become benchmark baselines that empower development of 12 | more accurate and complex models. For this to happen, these models should be made more discoverable and usable by ML 13 | practitioners and data scientists. The GMLMC specification aims to address this goal through a common metadata 14 | definition for ML models that operate on geospatial data. 15 | 16 | At a high level, this specification should provide sufficient information to enable search and discovery of geospatial 17 | ML models and answer the following questions: 18 | 19 | * Is this model applicable to my domain (e.g. land cover, agricultural monitoring, etc.)? 20 | * What kind of input data are required to use the model? 21 | * Is this model applicable to the geographic region I am interested in? 22 | * How well does this model perform under the kinds of conditions under which I will be using it? 23 | 24 | ## Getting Started 25 | 26 | The best place to start is with the [Model Metadata](./model-metadata) spec. This describes the top-level metadata 27 | document for a geospatial ML model. The Model Metadata spec describes the fields that may be present in the top-level 28 | metadata document with links more detailed descriptions and definitions of more complex fields. 29 | 30 | ## Relation to STAC 31 | 32 | The [SpatioTemporal Asset Catalog (STAC) spec](https://github.com/radiantearth/stac-spec) is a mature and well-defined 33 | standard for cataloging geospatial assets. STAC has mature specifications for describing remotely sensed data, as well 34 | as labeled data related to machine learning applications. The 35 | [Electro-Optical](https://github.com/stac-extensions/eo), 36 | [Projection](https://github.com/stac-extensions/projection), 37 | [View](https://github.com/stac-extensions/view), and 38 | [SAR](https://github.com/stac-extensions/sar) extensions, among others, provide a thorough description of remotely 39 | sensed data from a variety of sources. Combined with the [Label](https://github.com/stac-extensions/label) extension, 40 | these provide a thorough description of the data required for training an ML model using remotely sensed imagery. 41 | 42 | The Geospatial ML Model Catalog (GMLMC) specification takes advantage of the STAC specification to describe training data 43 | associated with a model, and we hope that discussions relating to the GMLMC spec will help guide development of those 44 | standards as well. 45 | 46 | ## Examples 47 | 48 | Example are currently a work in progress and we hope to have some soon. There is a [template 49 | file](./examples/template.json) that demonstrates the overall structure and types in the JSON document. 50 | 51 | ## Stability & Versioning 52 | 53 | This specification is a work in progress and should be considered unstable until further notice. The specification 54 | is not currently versioned, but we plan to begin versioning once we have a basic working draft. 55 | 56 | ## Contributing 57 | 58 | See the [Contributing guidelines](./CONTRIBUTING.md) for more information on how you can contribute to the spec. 59 | -------------------------------------------------------------------------------- /examples/template.json: -------------------------------------------------------------------------------- 1 | { 2 | "version": "0.1.0", 3 | "model_id": "template", 4 | "model_type": { 5 | "learning_approach": "Supervised", 6 | "prediction_type": "Classification", 7 | "architecture": "RCNN", 8 | "description": "Description here" 9 | }, 10 | "license": "MIT", 11 | "contacts": [ 12 | { 13 | "name": "Some Person", 14 | "organization": "My Organization", 15 | "email": "some.person@myorganization.com" 16 | } 17 | ], 18 | "citation": { 19 | "doi": "10.1000/xyz123", 20 | "citation": "RECOMMENDED CITATION HERE", 21 | "publications": [ 22 | { 23 | "doi": "10.1111/xyz123", 24 | "citation": "PUBLICATION CITATION HERE" 25 | } 26 | ] 27 | }, 28 | "training": { 29 | "created_at": "2020-04-12T23:20:50.52Z", 30 | "environment": { 31 | "processor": { 32 | "number_of_cores": 8, 33 | "processor_type": "GPU" 34 | }, 35 | "operating_system": "Windows", 36 | "programming_language": "Python", 37 | "dependencies": "https://www.github.com/myorg/myrepo/blob/main/somefolder/requirements.txt" 38 | }, 39 | "data": [ 40 | "https://api.radiant.earth/mlhub/v1/collections/ref_african_crops_kenya_01_source" 41 | ] 42 | }, 43 | "inputs": [ 44 | { 45 | "name": "input_var_1", 46 | "type": "float32", 47 | "shape": [50, 256, 256], 48 | "description": "Description of what the inputs should look like" 49 | } 50 | ], 51 | "outputs": [ 52 | { 53 | "name": "output_var_2", 54 | "type": "float32", 55 | "shape": [50, 1], 56 | "description": "Description of what the outputs of the model are" 57 | } 58 | ], 59 | "runtimes": [ 60 | { 61 | "type": "docker", 62 | "properties": { 63 | "format": "Docker", 64 | "link": "myregistryhost:5000/organization/model_container:1.0.0", 65 | "code_examples": [ 66 | "https://www.github.com/myorganization/greatexamples" 67 | ], 68 | "host_requirements": "???" 69 | } 70 | } 71 | ], 72 | "usage_recommendations": { 73 | "recommendations": [ 74 | { 75 | "spatial": [ 76 | [34.18, 0.47 ,34.37, 0.71] 77 | ], 78 | "temporal": [ 79 | ["2018-04-12T00:00:00Z", "2020-03-12T00:00:00Z"] 80 | ], 81 | "description": "Description of usage recomendation" 82 | } 83 | ] 84 | } 85 | } -------------------------------------------------------------------------------- /fragments/environment/README.md: -------------------------------------------------------------------------------- 1 | # Environment Fragment 2 | 3 | This fragment provides a thorough description of a compute environment used for training or running an ML model. The 4 | aim it to capture enough detail to enable new users to replicate the environment for training or inferencing. This 5 | object is used in both the `runtime` and `training` fields of the top-level [Model 6 | Metadata](../../model-metadata/model-metadata.md) document. 7 | 8 | ## Directory Contents 9 | 10 | * [`environment-fragment.md`](./environment-fragment.md) - The detailed Environment Fragment Specification. 11 | -------------------------------------------------------------------------------- /fragments/environment/environment-fragment.md: -------------------------------------------------------------------------------- 1 | # Environment Fragment Specification 2 | 3 | This document describes the structure and content of an Environment object. 4 | 5 | ## Environment Fields 6 | | Field Name | Type | Description | 7 | |-------------------------|--------------------|-----------------------------------------------------------------------| 8 | | `processor` | [Processor] | A description of the processor used in the environment. | 9 | | `operating_system` | string | The name of the environment's operating system. For training environments this indicates the operating system used to train the model. For runtime environments this indicates an operating system on which the model may be run. | 10 | | `programming_language` | string | A string identifying the programming language used in the environment. This should be a lower-cased string with spaces replaced by underscores (`_`). For instance, JavaScript would be `"javascript"` and Objective-C would be `"objective-c"`. | 11 | | `dependencies` | string | A link to a file defining the dependencies for creating the environment. This file must be interpretable using common tools in the given `programming_language`. | 12 | 13 | ### Processor 14 | 15 | Describes the processor used in the environment. 16 | 17 | | Field Name | Type | Description | 18 | |-------------------|-----------|------------------------------------------------------------------------| 19 | | `number_of_cores` | number | The number of cores required by the environment. | 20 | | `processor_type` | string | The type of processor used. Must be either `"CPU"` or `"GPU"` | 21 | 22 | [Processor]: #processor -------------------------------------------------------------------------------- /fragments/runtime/README.md: -------------------------------------------------------------------------------- 1 | # Model Runtime Fragment 2 | 3 | This fragment defines representations of a model as a file or container that can be used to generate inferences on new 4 | data. There are 2 types of model runtimes supported by the spec: 5 | 6 | * Containerized: An image that can be used to run the model 7 | * Serialized: A model saved in a well-known file format 8 | 9 | ## Containerized Model Runtimes 10 | 11 | Containerized model runtimes take the form of images that can be used to run a model using a container platform. 12 | [Docker] is currently the only supported containerized runtime. 13 | 14 | ## Serialized Model Runtimes 15 | 16 | Serialized model runtimes represent a file (or group of files) that serialize the model in such a way that it can be 17 | the model can be reproduced in an ML framework and used to generate inferences. [ONNX] is currently the only supported 18 | containerized runtime. 19 | 20 | ## Directory Contents 21 | 22 | * [`runtime-fragment.md`](./runtime-fragment.md) - The detailed Model Runtime Fragment specification. 23 | 24 | [Docker]: https://www.docker.com/ 25 | [ONNX]: https://onnx.ai/index.html -------------------------------------------------------------------------------- /fragments/runtime/runtime-fragment.md: -------------------------------------------------------------------------------- 1 | # Model Runtime Fragment Specification 2 | 3 | A model runtime provides a thorough description of the tools and frameworks needed to generate inferences using the 4 | model. This could be a file (or collection of files) representing the serialized model, or a containerized version of 5 | the model that can be run on a well-known platform. 6 | 7 | In order to make client-side parsing easier, all Model Runtime objects contain the same 2 top-level fields (see [Model 8 | Runtime Fields]). The structure of the `properties` field will depend on the specific type of runtime being described. 9 | 10 | 11 | ## Model Runtime Fields 12 | 13 | | Field Name | Type | Description | 14 | |---------------------|---------------------|-----------------------------------------------------------------------| 15 | | `type` | string | The type of model runtime. Either `"docker"` or `"onnx"`. 16 | | `properties` | [ONNX Fields]\|[Docker Fields] | Describes the properties of the runtime environment (depends on the `"type"`). | 17 | 18 | ## Serialized Models 19 | 20 | The spec currently only supports models serialized in the [ONNX] format. 21 | 22 | ### ONNX Fields 23 | 24 | | Field Name | Type | Description | 25 | |------------------|----------------|-----------------------------------------------------------------------| 26 | | `format` | string | Always `"ONNX"` for the ONNX serialization format. | 27 | | `link` | string | URL to the ONNX file. | 28 | | `environment` | [Environment] | Description of the environment required to run the model. | 29 | | `code_examples` | \[string\] | A list of URLs to code examples demonstrating model usage. | 30 | 31 | ## Containerized Models 32 | 33 | The spec currently supports model containerized using [Docker]. 34 | 35 | ### Docker Fields 36 | 37 | | Field Name | Type | Description | 38 | |-----------------------|-----------------------|-----------------------------------------------------------------------| 39 | | `format` | string | Always `"Docker"` for the Docker container format. | 40 | | `link` | string | Full identifier for the model image. See the [Docker Tag Reference] for details. | 41 | | `code_examples` | \[string\] | A list of URLs to code examples demonstrating model usage. | 42 | | `host_requirements` | [Host Requirements] | Description of the host requirements for running the container. | 43 | 44 | ### Host Requirements 45 | 46 | > **TODO:** Flesh out these fields (memory, CPU, networking, etc.). Maybe look to AWS and Azure requirements for Docker services. 47 | 48 | [ONNX]: https://onnx.ai/index.html 49 | [Docker]: https://www.docker.com/ 50 | [Docker Tag Reference]: https://docs.docker.com/engine/reference/commandline/tag/#extended-description 51 | [Environment]: ../environment/environment-fragment.md 52 | [Model Runtime Fields]: #model-runtime-fields 53 | [Host Requirements]: #host-requirements 54 | [ONNX Fields]: #onnx-fields 55 | [Docker Fields]: #docker-fields 56 | -------------------------------------------------------------------------------- /fragments/training/README.md: -------------------------------------------------------------------------------- 1 | # Model Training Fragment 2 | 3 | The Model Training fragment captures essential information about how the model was trained, including the 4 | compute environment, training datasets, and the date on which the model was trained. The goal of this section 5 | is to capture enough information that a new developer could reasonably reproduce the trained model. This also 6 | contributes to the ability to reproduce published model results. 7 | 8 | ## Directory Contents 9 | 10 | * [`training-fragment.md`](./training-fragment.md) - The detailed Model Training Fragment Specification. -------------------------------------------------------------------------------- /fragments/training/training-fragment.md: -------------------------------------------------------------------------------- 1 | # Model Training Fragment 2 | 3 | This document describes the structure and content of a Model Training object. 4 | 5 | | Field Name | Type | Description | 6 | |---------------------|---------------------|-----------------------------------------------------------------------| 7 | | `created_at` | string | **Required.** Approximate date and time at which the model was trained. It is formatted according to [RFC 3339, section 5.6]. | 8 | | `environment` | [Environment] | **Required.** Description of the environment used to train the model. | 9 | | `data` | [string] | **Required.** A list of URLs, each of which points to a [STAC Collection] representing input data used to train the model. | 10 | 11 | #### created_at 12 | 13 | This field is intended to capture the approximate date and time at which the model was trained in order to give 14 | other developers a sense of the currency of the model. The model publisher is free to choose the exact event that 15 | this timestamp represents (e.g. start of model training, completion of model training, etc.). Publishers may also 16 | choose to use `00` for all time values if they wish to only represent a date (e.g. 17 | `2020-02-01T00:00:00Z`). This format keeps the spec in line the datetime format requirements from 18 | the STAC spec. 19 | 20 | #### data 21 | 22 | A list of URLs, each of which points to a [STAC Collection] representing input data used to train 23 | the model. The link may reference a hosted static Collection or a Collection that is part of a 24 | [STAC API] implementation. If the Collection is part of a STAC API implementation, then the 25 | `/collections/{collection_id}` endpoint should be used. Collections must implement the [ML AOI 26 | Extension], which describes how individual Items are split into `train`, `test`, and `validation` 27 | sets. As per the ML AOI Extension, Items in the Collection will then link to Items describing the 28 | input data itself, as well as labels (for supervised learning models). 29 | 30 | ##### Input Data 31 | 32 | Items from the Collection linked to in the `data` field will in turn reference Items in an input data 33 | Collection. This Collection should implement any [STAC Extensions] required to thoroughly describe 34 | the source data. For instance, a collection of optical satellite imagery should probably implement 35 | the [Projection], [View Geometry], and [Electro-Optical] extensions. 36 | 37 | ##### Label Data 38 | 39 | For supervised learning models, Items from the Collection linked to in the `data` field will also 40 | reference Items in a label Collection. Items in this label Collection must implement the [Label 41 | Extension] and should also implement any other [STAC Extensions] that the model publisher 42 | believes are necessary to thoroughly describe the dataset. 43 | 44 | The STAC [Label Extension] is intended to support the use of labeled AOIs with machine learning models and supports 45 | the use of both raster labels and vector labels. The extension supports labels for the following machine learning model 46 | types: 47 | 48 | * Classification 49 | * Regression 50 | * Object detection 51 | * Segmentation 52 | 53 | Model publishers are encouraged to review the [Label Extension] in detail to better understand its capabilities and how 54 | to use it to properly catalog training data. *The extension is still in the "Proposal" stage and is undergoing active 55 | development.* Model publishers are encouraged to contribute to the development of that specification to ensure that it 56 | meets the needs of the community. See the "Get Involved" section of the [STAC Spec site] for details on how to join the 57 | discussion. 58 | 59 | 60 | [RFC 3339, section 5.6]: https://tools.ietf.org/html/rfc3339#section-5.6 61 | [STAC Collection]: https://github.com/radiantearth/stac-spec/tree/master/collection-spec 62 | [Environment]: ../environment/environment-fragment.md 63 | [STAC Extensions]: https://stac-extensions.github.io/ 64 | [Projection]: https://github.com/stac-extensions/projection 65 | [View Geometry]: https://github.com/stac-extensions/view 66 | [Electro-Optical]: https://github.com/stac-extensions/eo 67 | [Label Extension]: https://github.com/stac-extensions/label 68 | [Input Datasets]: #input-datasets 69 | [STAC Spec site]: https://stacspec.org/ 70 | [ML AOI Extension]: https://github.com/stac-extensions/ml-aoi 71 | [STAC API]: https://github.com/radiantearth/stac-api-spec 72 | -------------------------------------------------------------------------------- /fragments/usage/README.md: -------------------------------------------------------------------------------- 1 | # Usage Recommendations Fragment 2 | 3 | This fragment describes conditions under which the model publishers believe the model will perform 4 | as expected. 5 | 6 | While the [Model Training] section gives users a clear picture of the training conditions, the model 7 | publisher may also have evidence or intuition that the model will also perform well when used in a 8 | different context. For instance, a model might have been trained on data from a particular 9 | geographic region and time period, but based on some combination of intuition, previous experience, 10 | and informal experimentation, the model publisher may be confident that it will perform similarly in 11 | a new geographic region. This section is meant to capture that information in a systematic way, 12 | while also providing publishers the opportunity to describe their recommendations in a more 13 | free-form style. 14 | 15 | ## Directory Contents 16 | 17 | * [`usage-fragment.md`](./usage-fragment.md) - The detailed Usage Recommendations Fragment 18 | specification. 19 | 20 | [Model Training]: ../training -------------------------------------------------------------------------------- /fragments/usage/usage-fragment.md: -------------------------------------------------------------------------------- 1 | # Usage Recommendations Specification 2 | 3 | The Usage Recommendations section defines conditions under which the model publishers believe the 4 | model will perform as expected. It also provides a mechanism for publishers to explicitly call out 5 | conditions under which they believe the model is likely to not performa as expected. It is important 6 | to note that these are *recommendations* and that they **in no way provide any guarantee of model 7 | performance under the given conditions**. 8 | 9 | If no Usage Recommendation section is provided, then users should assume that the model will only 10 | perform as expected under conditions that match those described in the [Model Training] section. 11 | This also applies to any fields that are not present in the Usage Recommendations section. For 12 | instance, if no `temporal` field is provided, users should assume the model will only perform as 13 | expected for data in the time range given in the [Model Training] section. 14 | 15 | 16 | ## Usage Recommendations Fields 17 | 18 | The `recommendations` and `cautions` fields in this section are arrays of objects that describe 19 | conditions under which the model either should perform as expected or should *not* behave as 20 | expected, respectively. 21 | 22 | If the Usage Recommendations section is present, then at least one of these fields is required and 23 | must be a non-empty array. 24 | 25 | | Field Name | Type | Description | 26 | | ----------------- | --------------------- | ---------------------------------------------------------------------------------------------- | 27 | | `recommendations` | \[[Usage Conditions]] | An array of objects describing conditions under which the model *should* perform as expected. | 28 | | `cautions` | \[[Usage Conditions]] | An array of objects describing conditions under which the model *may not* perform as expected. | 29 | 30 | ## Usage Conditions 31 | 32 | This object is used to described conditions under which model usage is either recommended (if found 33 | in the `recommendations` list) or discouraged (if found in the `cautions` list). 34 | 35 | | Field Name | Type | Description | 36 | | ----------- | ------------------------ | ------------------------------------------------------------------------------------------ | 37 | | spatial | [Spatial Extent Object] | The spatial extents to which the recommendation or caution applies. | 38 | | temporal | [Temporal Extent Object] | The temporal extents to which the recommendation or caution applies. | 39 | | description | string | A human-readable description of conditions to which the recommendation or caution applies. | 40 | 41 | ### Spatial Extent Object 42 | 43 | | Element | Type | Description | 44 | | ------- | ------------ | -------------------------------------------------------------------------------------------- | 45 | | bbox | \[\[number]] | **REQUIRED.** Potential *spatial extents* within which the model should perform as expected. | 46 | 47 | The `bbox` element is an *array* whose elements are themselves arrays. Each inner array is a bounding 48 | box that describes a spatial extent within which the model should perform as expected. 49 | 50 | The format of the inner bounding box arrays follows the specification outlined in the **bbox** 51 | section of the [STAC Spatial Extent Object] documentation. However, the `bbox` element defined in 52 | this spec differs from the `bbox` element in the STAC Spatial Extent Object in that there is no 53 | first element that always represents the overall spatial extent. Such an overall extent might give 54 | users the incorrect impression that the model should perform as expected anywhere within that 55 | spatial extent, when in fact there may be only specific areas in which this is true. 56 | 57 | ### Temporal Extent Object 58 | 59 | | Element | Type | Description | 60 | | -------- | ------------------ | --------------------------------------------------------------------------------------------- | 61 | | interval | \[\[string\|null]] | **REQUIRED.** Potential *temporal extents* within which the model should perform as expected. | 62 | 63 | The `interval` element is an *array* whose elements are themselves arrays. Each inner array is an 64 | interval array containing 2 elements, each of which may be either `null` or a timestamp. Timestamps 65 | must follow the specifications outlined in the [STAC Temporal Extent Object] documentation. 66 | 67 | The format of the inner interval arrays follows the specification outlined in the [STAC Temporal 68 | Extent Object] documentation. However, the `interval` element defined in this spec differs from 69 | the `interval` element in the STAC Temporal Extent Object in that there is no first element that 70 | always represents the overall temporal extent. Such an overall extent might give users the 71 | incorrect impression that the model should perform as expected anywhere within that temporal range, 72 | when in fact there may be only specific ranges in which this is true. 73 | 74 | 75 | [Model Training]: ../training 76 | [Spatial Extent Object]: #spatial-extent-object 77 | [Temporal Extent Object]: #temporal-extent-object 78 | [Usage Conditions]: #usage-conditions 79 | [STAC Spatial Extent Object]: https://github.com/radiantearth/stac-spec/blob/v1.0.0-rc.4/collection-spec/collection-spec.md#spatial-extent-object 80 | [STAC Temporal Extent Object]: https://github.com/radiantearth/stac-spec/blob/v1.0.0-rc.4/collection-spec/collection-spec.md#temporal-extent-object 81 | -------------------------------------------------------------------------------- /model-metadata/README.md: -------------------------------------------------------------------------------- 1 | # Model Metadata 2 | 3 | The Model Metadata document is the top-level structure containing all metadata pertaining to a geospatial ML model. 4 | Consumers of the metadata document will start here and navigate the structure of this object to extract the information 5 | they need regarding the model. 6 | 7 | The attributes in the top-level Model Metadata document may be simple types (e.g. a string `model_id`) or more complex 8 | structures (e.g. a `training_data` section). In most cases, these attributes are documented directly in this section. 9 | However, in the case of more complicated sections (like `training_data`), the definitions have been moved to their own 10 | directory in the [`fragments`](../fragments) directory. In either case, you can start with the 11 | [`model-metadata.md`](./model-metadata.md) specification and follow links to the appropriate definitions. 12 | 13 | ## Directory Contents 14 | 15 | * [`model-metadata.md`](./model-metadata.md) - The detailed Model Metadata specification, including attribute 16 | definitions. -------------------------------------------------------------------------------- /model-metadata/model-metadata.md: -------------------------------------------------------------------------------- 1 | # Model Metadata Specification 2 | 3 | This document describes the structure and content of a top-level Model Metadata object. 4 | 5 | ## Model Metadata Fields 6 | 7 | | Field Name | Type | Description | 8 | | ----------------------- | ----------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- | 9 | | `version` | string | The GMLMC spec version that this metadata document implements. | 10 | | `model_id` | string | **REQUIRED.** A unique string identifier for the model. This ID must be unique across the provider. | 11 | | `model_type` | [Model Type] | **REQUIRED.** Describes the learning approach, type of prediction, and model architecture. | 12 | | `license` | string | **REQUIRED.** The model's license(s). Either a SPDX [License identifier], `various` if multiple licenses apply, or `proprietary` for all other cases. | 13 | | `contacts` | \[[Contacts]\] | **REQUIRED.** List of names and contact information for question and issues related to the model. | 14 | | `citation` | [Citation] | Citation information related to the model. | 15 | | `training` | [Model Training] | **REQUIRED.** A description of the data and environment used to train the model. | 16 | | `inputs` | \[[Model Input]\] | **REQUIRED.** A list of [Model Input] objects describing the names and types of model inputs. | 17 | | `outputs` | \[[Model Output]\] | **REQUIRED.** A list of [Model Output] objects describing the names and types of model outputs. | 18 | | `runtimes` | \[[Runtime]\] | A list of [Runtime] objects describing serialized or containerized versions of the model that can be used to generate inferences. | 19 | | `usage_recommendations` | [Usage Recommendations] | A description of the recommended conditions under which the model can be used. | 20 | 21 | ### Model Type 22 | 23 | This object defines the type of model based on the learning approach and type of prediction. It also allows the 24 | publisher to provide a free-text description of the model type to cover any nuances that are not 25 | adequately described by the other fields. 26 | 27 | | Field Name | Type | Description | 28 | | ------------------- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | 29 | | `learning_approach` | string | **REQUIRED.** The learning approach used to train the model. It is STRONGLY RECOMMENDED that you use one of the values in the [Learning Approach] section below, but other values are allowed. | 30 | | `prediction_type` | string | **REQUIRED.** The type of prediction that the model makes. It is STRONGLY RECOMMENDED that you use one of the values in the [Prediction Type] section below, but other values are allowed. | 31 | | `architecture` | string | **REQUIRED.** Identifies the architecture employed by the model (e.g. RCNN, U-Net, etc.). This may be any string identifier, but publishers are encouraged to use well-known identifiers whenever possible. | 32 | | `description` | string | A free text description of the model type. | 33 | 34 | #### Learning Approach 35 | 36 | Describes the learning approach used to train the model. It is STRONGLY RECOMMENDED that you use one of the 37 | following values, but other values are allowed. 38 | 39 | * `"Supervised"` 40 | * `"Unsupervised"` 41 | * `"Semi-Supervised"` 42 | * `"Reinforcement Learning"` 43 | 44 | #### Prediction Type 45 | Describes the type of predictions made by the model. It is STRONGLY RECOMMENDED that you use one of the 46 | following values, but other values are allowed. Note that not all Prediction Type values are valid 47 | for a given [Learning Approach]. 48 | 49 | * `"Object Detection"` 50 | * `"Classification"` 51 | * `"Segmentation"` 52 | * `"Regression"` 53 | 54 | ### Contacts 55 | 56 | This object describes contact information for individuals associated with the model. These contacts 57 | may be different than the model authors listed in [Citation] section. 58 | 59 | | Field Name | Type | Description | 60 | | -------------- | ------ | ----------------------------------------------------------------- | 61 | | `name` | string | **REQUIRED.** The full name of the contact. | 62 | | `organization` | string | The name of the organization to which this contact is affiliated. | 63 | | `email` | string | **REQUIRED.** The email for this contact. | 64 | 65 | ### Citation 66 | 67 | This object indicates from which publication the model originates and how the model itself should be cited or 68 | referenced. The object follows the [STAC Scientific Citation Extension], with 2 notable changes: ( 69 | * There is no `sci:` prefix on the field names 70 | * If `doi` is present, then `citation` is also required 71 | 72 | All field descriptions are adapted from the corresponding [STAC Scientific Citation Extension] 73 | definitions. *As per the Scientific Citation Extension spec, at least one field is required.* 74 | 75 | | Field Name | Type | Description | 76 | | -------------- | ------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | 77 | | `doi` | string | The DOI of the data, e.g. `10.1000/xyz123`. This MUST NOT be a DOIs link. | 78 | | `citation` | string | The recommended human-readable reference (citation) to be used by publications citing the model. No specific citation style is suggested, but the citation should contain all information required to find the publication distinctively. **Required if `doi` is present.** | 79 | | `publications` | \[[Publication Object]\] | List of relevant publications referencing and describing the data. | 80 | 81 | #### Publication Object 82 | 83 | | Field Name | Type | Description | 84 | | ---------- | ------ | ---------------------------------------------------------------------------- | 85 | | `doi` | string | See the description of `doi` field in the [Citation] description above. | 86 | | `citation` | string | See the description of `citation` field in the [Citation] description above. | 87 | 88 | ### Model Input 89 | 90 | This object describes a single model input parameter. 91 | 92 | | Field Name | Type | Description | 93 | | ------------- | ------ | ---------------------------------------------------------------------------------------------- | 94 | | `name` | string | **REQUIRED.** The name of the parameter. | 95 | | `type` | string | **REQUIRED.** The data type of the parameter (e.g. `float32`). | 96 | | `shape` | [int] | **REQUIRED.** The shape of the parameter as an array of integers. | 97 | | `description` | string | A human-readable description of the parameter that indicates what type of content is required. | 98 | 99 | ### Model Output 100 | 101 | This object describes a single model output. 102 | 103 | | Field Name | Type | Description | 104 | | ------------- | ------ | ---------------------------------------------------------------------------------------------- | 105 | | `name` | string | **REQUIRED.** The name of the parameter. | 106 | | `type` | string | **REQUIRED.** The data type of the parameter (e.g. `float32`). | 107 | | `shape` | [int] | **REQUIRED.** The shape of the parameter as an array of integers. | 108 | | `description` | string | A human-readable description of the parameter that indicates what type of content it contains. | 109 | 110 | [License identifier]: https://spdx.org/licenses/ 111 | [STAC Scientific Citation Extension]: https://github.com/radiantearth/stac-spec/tree/v1.0.0-rc.1/extensions/scientific 112 | [Runtime]: ../fragments/runtime 113 | [Model Training]: ../fragments/training 114 | [Learning Approach]: #learning-approach 115 | [Prediction Type]: #prediction-type 116 | [Model Type]: #model-type 117 | [Contacts]: #contacts 118 | [Citation]: #citation 119 | [Model Input]: #model-input 120 | [Model Output]: #model-output 121 | [Usage Recommendations]: ../fragments/usage 122 | [Publication Object]: #publication-object 123 | --------------------------------------------------------------------------------