preparing release v0.10

GammaLearn v10.0 introduces new domain adaptation techniques with DeepJDot and DeepCoral. The data loading of the vision datasets has been refactored to make it more general. The containerization is improved with an environment layer using mamba. The documentation has been improved.

Changelog since release v0.9 of gammalearn:
* build docker dev from base mamba image (@vuillaume)
* remove leftover docs/CLI directory causing bug (@vuillaume)
* Cities (@michael.dellaiera)
* Improve documentation and adds a CLI submenu (@vuillaume)
* add unit test for gl_dl1_to_dl2 (@vuillaume)
* Fix DL1 to DL2 entry point (@vuillaume)
* removing line converting parameters from float64 to float32 (@vuillaume)
* Build doc in glearn env, fixing doc build (@vuillaume)
* fix issue with latest sphinx (@vuillaume)
* Fixing test base image renamed to master (@vuillaume)
* rename glearn into glearnenv for clarity (@vuillaume)
* message to force build (@vuillaume)
* force build docker env master (@vuillaume)
* gitlab docker as separate file for clarity and sequential building (@vuillaume)
* force mamba build (@vuillaume)
* move docker dev to docker master (@vuillaume)
* read camera geometry_0 in BaseLSTDataset (@vuillaume)
* Cityscenes (@michael.dellaiera)
* Add dockerfiles for mamba and glearn containers (@vuillaume)
* Variational Auto Encoder (@michael.dellaiera)
* fix issue with setting libmamba solver (@vuillaume)
* Deepjdot (@michael.dellaiera)
* Fix torchmetrics version to be lower than 1.8 (@vuillaume)
* Data loading refactoring (@michael.dellaiera)
* Multigpus hotfix (@michael.dellaiera)
* Base mae (@michael.dellaiera)
* Prevent from testing in Multigpu mode (@jacquemont)
* Loss_balancing as a LightningModule attribute (@jacquemont)
* Reduce CI duration (@jacquemont)
* Fix metrics (@jacquemont)
* Quick fix for Gitlab runner memory issue (@jacquemont)
* Implementation of Masked AutoEncoder (@jacquemont)
* Allow mixing MC and LST1 data at train time (@jacquemont)
* libmamba solver (@vuillaume)
* Checkpointing options (@jacquemont)
* Allow to use MNIST, USPS and SVHN datasets (@michael.dellaiera)
* Incremental changes to merge deepjdot in the future (@michael.dellaiera)
* fix metric (@michael.dellaiera)
* Fresh docker (@vuillaume)
* log precision (@michael.dellaiera)
* Replace GPUStatsMonitor by DeviceStatsMonitor (pytorch lightning) (@jacquemont)
* numpy < 1.23 and fresh install in unit tests (@vuillaume)
* roll back from DeviceStatsMonitor to GPUStatsMonitor (@jacquemont)
* Update codemeta.json for latest OSSR requirements (@vuillaume)
* Create Inference mode (@jacquemont)
* Addition of pointing. (@michael.dellaiera)
* renamed exp setting domain adaptation -> dann + move loss commputation out of... (@michael.dellaiera)
* Vitorch (@michael.dellaiera)
* Replace mamba by conda in CI (@jacquemont)
* Draft: Auto merge (@michael.dellaiera)
* Fix environment file (@jacquemont)