Releases: Clay-foundation/model
Releases · Clay-foundation/model
v1.0
Release v1.0 (2024/06/06)
💫 Highlights
- 🎉 Release of Clay Foundation Model v1.0 🎉
🚀 Model
- Datamodule & Model for Clay v1 by @srmsoumya in #253
🗃️ Data Pipeline
- Moved data pipeline code into stacchip a dedicated library for the data wrangling
📖 Documentation
- Clay over aoi by @yellowcap in #116
- Convert run-over-aoi ipynb to md by @yellowcap in #122
- Static Documentation for v0.1 by @brunosan in #118
- Document data sampling strategy by @yellowcap in #124
- Updating location to download imagery script in docs by @MaxLenormand in #198
- Change "Radiant Earth Foundation" to "Radiant Earth" in docs by @kbgg in #214
- Update instructions to re-lock conda-lock.yml file by @weiji14 in #225
- chore(docs): correct typos and improve stylistic consistency by @kelseyjosund in #232
- Partial inputs - Pakistan flood tutorial by @lillythomas in #154
- Obtain patch level metadata (e.g. geospatial bounds and cloud cover), save and demo DEP use case (sim search) by @lillythomas in #172
- Obtain patch level cloud coverage percentages by @lillythomas in #184
- Tutorial on burn scar analysis using embeddings from partial inputs by @yellowcap in #149
- Doc updates by @brunosan in #230
- Update specification.md by @brunosan in #250
- Add notebook showing how to run v1 by @yellowcap in #254
- Documented v1 sampling strategy by @yellowcap in #249
- Updates the mamba install instructions link. by @dbonomo in #252
- Remove clay-v0-tutorials & add visualize embeddings for v1 by @srmsoumya in #256
- Minor spelling fix by @tylere in #260
- Shorten comment line length by @yellowcap in #261
- Refactor docs by moving v0 docs into separate section by @yellowcap in #262
- Docs v1 continued by @yellowcap in #263
- Documented metadata file for normalization and wavelenghts by @yellowcap in #266
- [small change] add source.coop link by @brunosan in #137
- Segmentation on Clay by @srmsoumya in #257
🧰 Maintenance
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #206
- Bump conda-lock from 2.5.1 to 2.5.6 by @weiji14 in #183
- Remove scripts folder by @yellowcap in #265
- Unified v1 and v0 environments by @yellowcap in #264
- Add osx-arm64 platform to conda-lock.yml file and GitHub Actions CI by @weiji14 in #164
- Move pipeline scripts into dedicated folder by @yellowcap in #163
- Memory buildup clean by @brunosan in #158
- Add osx platform to conda lock file and CI by @chuckwondo in #162
- Add scikit-learn by @weiji14 in #159
- Remove float16 dype in favor of float32 by @yellowcap in #143
- Remove stale bands argument from clay model classes by @yellowcap in #144
🛠️ v0.2 updates
- Add normalization parameters to documentation by @yellowcap in #152
- Worldcover embeddings conus by @yellowcap in #153
- Create data/minicubes directory if it doesn't already exist by @weiji14 in #160
- Clay pipeline v04 by @yellowcap in #173
- Shuffle patches only during training by @yellowcap in #135
- Add option to output raw patch embeddings by @yellowcap in #133
🧑🤝🧑 New Contributors
- @chuckwondo made their first contribution in #162
- @MaxLenormand made their first contribution in #198
- @kbgg made their first contribution in #214
- @kelseyjosund made their first contribution in #232
- @dbonomo made their first contribution in #252
- @tylere made their first contribution in #260
- @alkalait, @raulramosudea, @rramosp and @jmarintur contributed to https://github.com/Clay-foundation/earth-text
- @MaceGrim, @4242psherman4242 contributed on embeddings QA in the background.
Full Changelog: v0.0.1...v1.0
v0.0.1
Release v0.0.1 (2024/01/12)
💫 Highlights
- 🎉 First release of Clay Foundation Model 🎉
- Implement MAE with support for position, time, latlon & channel embeddings (#47)
🚀 Model
- Generate embeddings from CLAYModule trained with latlon/time encodings (#96)
- Add multigpu support & UI to test embeddings (#109)
- Refactor model for multi-device usage and easier disabling of masking (#95)
- Rename embeddings file to include MGRS code and store GeoTIFF source_url (#86)
- Callback function to log Masked Autoencoder reconstructions to WandB (#88)
- Adapt model to load 512x512 images from s3 bucket (#85)
- Save embeddings with spatiotemporal metadata to GeoParquet (#73)
- Generate embeddings via prediction loop (#56)
- Initial Vision Transformer architecture with MAE decoder (#37)
🗃️ Data Pipeline
- Adapted sampling strategy (#81)
- Allow ClayDataModule to load GeoTIFF files directly from s3 (#92)
- Let ClayDataModule return same spatiotemporal fields as GeoTIFFDataModule (#91)
- Improve date handling for data pipeline (#76)
- Let LightningDataModule return spatiotemporal metadata (#66)
- check for no data on a tile level in sentinel 1 vv and vh, sentinel 2 and DEM (#60)
- Batch setup (#54)
- LightningDataModule to load GeoTIFF files (#52)
- Ready for batch (#44)
- Tiler module (#41)
- Landcover based sampling strategy (#29)
- Datacube (#27)
📖 Documentation
- Document how the benchmark dataset labels were prepared (#100)
- Document how to finetune pretrained model on downstream task (#99)
- Document how to generate vector embeddings (#98)
- Document how to run the datacube pipeline with a batch job (#97)
- Initialize Jupyter Book documentation (#89)
- Setting the model license to OpenRail-M (#63)
- Create CODE_OF_CONDUCT.md (#53)
🧰 Maintenance
- Bump pytorch from 2.0.0 to 2.1.0, CUDA from 11.8 to 12.0 (#51)
- Add pre-commit hooks with ruff formatter/linter rules (#26)
- Setup GitHub Actions Continuous Integration tests (#25)
- Setup LightningCLI trainer script (#24)
- Initial conda environment and binder links (#15)
🧑🤝🧑 Contributors
Full Changelog: v0.0.0...v0.0.1