Use ML.net to recognize patterns (as trained into model with multivariate time series CSV) on a testing CSV (same format as training model)

Question

I have train'g & test'g CSVs with a simple stock market format: DateTime,Open,Close,Up,Down,Mov_Avg, ... ,RecognizablePattern. Using all of the attributes as features and the RecognizablePattern (manually entered "1" for patterns to be recognized, otherwise "0") as the label (column to forecast), I used Model Builder to create the model (creates JSON [that represents data format, algorithms test score, etc.], some C# code, and I believe a binary file). Somehow, I believe a zip file is to be created and consumed by the application (to appropriately use the output of the Model Builder). I have made several (manual) attempts to load a zip file with componentry needed to satisfy MLContest's LoadFromTextFile