Minor updates and improvements in request and response models

Request Model Updates

Duplicate biomarker detection - The input request model for scoring and generating the Health Report now detect duplicate biomarkers in the request and return a validation message where a duplicate is found. In the case of duplicates, the first occurrence of a biomarker will be used for the scoring with the validation message containing the biomarkers that were skipped. Key things to note are:
1. We detect duplicate biomarkers by the observation code (LOINC) that's sent regardless of the value. For time series scoring and health report generation, we detect duplicates in a single health event and for batch scoring we will detect duplicates per user.
2. A validation message will be sent in the response when a duplicate is detected regardless of the scoring mode you have selected. You will still see a message if you have selected score as the mode.

Response Model Updates

New fields in scoring predictors - The scoring_predictor object will now contain the following 2 additional fields:
1. predictor_id - This is the ID field of the predictor / biomarker that was sent as part of the response. We match this back to allow you to track these. Where a predictor / biomarker wasn't sent, we generate an ID and return it (UUID).
2. health_event_date - This is the event date sent in the request for time series scoring. We already return the health_event_id as part of the scoring_predictors object to help identify which event the predictor was taken from for scoring; however, we will now also include the date to make it easier to track which data was used in generating scores.