[script.*] Implement F1-score a priori where possible; Improve plotting of precision-recall curves with better legends, more consistence; Remove redundant plotting code; Update tests

[engine.trainer] Move creation of output_folder into engine; Call train argument output-path -> output-folder to make it more explicit

[script.train] Add option to drop incomplete last batch;  Improve documentation of options to explain batch processing

[checkpointer] Make it flexible to directory renames; Fix saving of trainlog in case we are interrupted

[docs,scripts] Separate evaluation script; Improve trainer engine code; Reset network model names to match resource names