Lately I have been studying Julien Simon’s videos for autotrain.
Each time I try it with a new dataset I get stuck in errors during data prep however its not easy to fix whatever data expectation is for the input…
I am curious if you could maybe make three easy learning examples for a small autotrain model (maybe a couple from text and tabular with a small CSV file)?
The reason they are needed is alot of people have been trying autotrain but they stop after the first few errors which usually occur with a null data error or key ID problem potentially with non unique identifiers.
I think if you were to prepare a simple set of examples complete with their own shortened dataset to show running end to end it would be better to start successfully.
e.g. If this dataset is used:
How could I produce a model that would score a text input and give descending matches across the ParentName field to get all LoincName values with their corresponding score on closeness to input sentence (e.g. sentence similarity between input matched to text content fields.
Sample search using Spaxes and “LOINC Panels and Forms” search:
