This is an experiment to approach #106

1 - Removed the keyword argument "extractor_file" from Extractor.train
so we remove the model persistence from it
2 - Create the method Extractor.save to handle the persistence task
3 - Make tools.extractor save the model.
3 jobs for big-issue-106 in 20 minutes and 24 seconds (queued for 2 seconds)
latest
Status Job ID Name Coverage
  Build
failed #170269
docker
build_linux_36

00:20:24

failed #170273
macosx
build_macosx_36

00:10:24

77.0%
failed #170270
macosx
build_macosx_36

00:00:02

 
Name Stage Failure
failed
build_macosx_36 Build
TOTAL                                                  4155    938    77%
----------------------------------------------------------------------
Ran 73 tests in 51.514s

FAILED (errors=13, failures=2)
Tests failed for bob.bio.base-4.0.2b0-py36ha2d65cb_30.tar.bz2 - moving package to /Users/gitlab/builds/f6f95b82/0/bob/bob.bio.base/miniconda/conda-bld/broken
WARNING conda_build.build:tests_failed(2188): Tests failed for bob.bio.base-4.0.2b0-py36ha2d65cb_30.tar.bz2 - moving package to /Users/gitlab/builds/f6f95b82/0/bob/bob.bio.base/miniconda/conda-bld/broken
TESTS FAILED: bob.bio.base-4.0.2b0-py36ha2d65cb_30.tar.bz2
ERROR: Job failed: exit status 1
failed
build_linux_36 Build
    raise CondaMultiError(exceptions)
conda.CondaMultiError: Downloaded bytes did not match Content-Length
url: https://repo.anaconda.com/pkgs/main/linux-64/mkl-2019.3-199.tar.bz2
target_path: /scratch/builds/bob/bob.bio.base/miniconda/pkgs/mkl-2019.3-199.tar.bz2
Content-Length: 213214191
downloaded bytes: 11470516


ERROR: Job failed: exit code 1