Show HN: Anything To JSON – a language model for structured extraction

Re:Show HN: Anything To JSON – a language model for structur... by yeldarb on Wednesday March 20, 2024 @03:16PM (#39770818)

Nice, makes sense to chain models so you don't waste the attention of the big smart model on the grunt work of JSON structure.

I bet there are some good post-processing heuristics you could also apply for hallucination with a flag for "this should be in the text verbatim" & then string matching whether the answer it outputted was a string from the text or not.

[ Parent | on HN ]

Re:Show HN: Anything To JSON – a language model for structur... by edunteman on Wednesday March 20, 2024 @03:20PM (#39770866)

interesting idea! yeah next big challenge for us is hallucination in null fields (IE: if you ask for a "name" from text that doesn't have a name you usually get "John") so need to add more sampling heuristics to doublecheck against the input blob for a sense of if it's an accurate extraction

[ Parent | on HN ]

Re:Show HN: Anything To JSON – a language model for structur... by 1997 on Wednesday March 20, 2024 @03:50PM (#39771231)

Awesome, agree on the nulls! Any plans to add support for image inputs?

[ Parent | on HN ]

Re:Show HN: Anything To JSON – a language model for structur... by edunteman on Wednesday March 20, 2024 @04:35PM (#39771715)

we have no plans for image or other modalities. I'd like to keep it just text-to-text so it can be as sharp of a tool as possible

[ Parent | on HN ]