Skip to content

GitLab

Closed
Created by Elizaveta Lutsenko@LizLutsenkoOwner

Error with Pipeline import/export in several examples

Created by: gkirgizov

Example pipeline_import_export.py and examples/simple/time_series_forecasting/clstm.py fail with the following, something related to serialization of operations:

Traceback (most recent call last):
  File "/home/hades/projects/FEDOT/examples/simple/pipeline_import_export.py", line 85, in <module>
    run_import_export_example(pipeline_path='import_export', pipeline=regression_ransac_pipeline())
  File "/home/hades/projects/FEDOT/examples/simple/pipeline_import_export.py", line 77, in run_import_export_example
    pipeline_from_dict.load(dict_pipeline, dict_fitted_operations)
  File "/home/hades/projects/FEDOT/fedot/core/pipelines/pipeline.py", line 280, in load
    self.template.import_pipeline(source, dict_fitted_operations)
  File "/home/hades/projects/FEDOT/fedot/core/pipelines/template.py", line 217, in import_pipeline
    self.convert_to_pipeline(self.link_to_empty_pipeline, path, dict_fitted_operations)
  File "/home/hades/projects/FEDOT/fedot/core/pipelines/template.py", line 267, in convert_to_pipeline
    f.write(dict_fitted_operations['preprocessing'])
TypeError: a bytes-like object is required, not '_io.BytesIO'