chore(dvc): ModelSize: 10M instead of 4M

chore(dvc): First experiment

docs(dvc): Best explanation on experiment

feat(dvc): Change plots to simple ...

It's easier to see the points in the plot.

fix(dvc): Only evaluate trained areas

chore(dvc): Test modelsize 4M

fix(dvc): Metrics where in an array ...

DVC couldn't read interesting values

feat(dvc): Add metrics ...

Statistics on precision values of all geographic areas.

feat: Add evaluation and plots

docs: Update dag

feat: Add maximum size to models

feat: Add train-areas ...

And run it

feat: Add fasttext 0.9.2 binary

fix(assign-to-areas): One area seemed to be undefined

chore(dvc): Integrate test-to-areas into split stage

chore(dvc): Add stage test-to-areas

feat: Add assign-to-areas ...

only to existing areas

style(address-to-area): Remove dead code, improve comments

docs: Add diagram

fix(split-addresses): Remove area from addresses

feat(dvc.yaml): Reorganize dvc.yaml ...

To prepare train AND test

docs: À propos des aires géographiques

feat(split): Reduce the number of areas

feat(create-tree): Create areas directory ...

And subdirectories

feat(extract-areas): Add extract-areas to dvc.yaml

docs(dvc.yaml): Add a description to the stage

feat(address-to-area): Use params.yaml

feat(params): Add getParam

chore(dvc): Generate area-address.tsv

chore(dvc): Use the parameter