Add some information in the header in respect on how the document was procesed

[lfoppiano]

I find sometimes difficult to detect whether certain processes has been performed by the grobid engine, one for example is the sentence segmentation, which would require to pass through the entire, or part of the document, to detect it.

I would suggest to add somewhere (in the header), maybe in the part related to the application, which type of processing has been used to produce the output TEI, and this would help understand part of the structure of the underlying TEI.

Some of the information that could be included:

• sentence segmentation
• git revision (the version alone might not be enough since the patch-released are not so frequent - although this would be more relevant for development / testing)
• consolidation was used
• used models architecture

In case of consolidation, for example, could be useful to avoid re-running on subsequent processes. E.g. DataStet processing TEI that are already consolidated would not need to repeat the process.