GPT-JSON
AI
llm output parsing library
SHOW HN2023
How HN took it
5 positive, 11 neutral or asking, 1 mixed, 10 critical, out of 27 comments about the project. Thread score -0.14.
Idea: 3 for, 7 againstExecution: 0 for, 2 againstDesign: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Pydantic-based prompt-to-schema approach is useful for chaining GPT calls
From 1 comment
You might not! Depends on what you're looking for. I've been finding this library most helpful in places where I have a lot of GPT calls in pipelines, so having typehinted schema…
icyfox on HN
Objections
Prompt-only JSON instructions already work about 99% of the time
From 2 comments
I've been interfacing with GPT programmatically for a little while now, leveraging it's "soft and fuzzy" interface to produce hard / machine-readable results. JSON was the format…
btbuildem on HNIt works fine 99% of the time by just using a small amount of extra instruction in the actual prompt. The method GP describes works in any language with just the basic building…
verdverm on HN
A library is unnecessary overhead for this task
From 2 comments
It works fine 99% of the time by just using a small amount of extra instruction in the actual prompt. The method GP describes works in any language with just the basic building…
verdverm on HNyeah, don't disagree, however this idea is better a function I use within gpt-index or langchain. There is a horse race and we're all making our bets about who's going to win
verdverm on HN
Retry loops are questioned as unreliable
From 1 comment
I dont see how these infinite loops are a good idea... You never sure if you actually getting a good result ? What is the failure rate?
TeeWEE on HN
Requests
Support for YAML or markdown output instead of JSON
From 2 comments
In the last 24 hours I've seen a bunch of projects doing LLM -> JSON. I think we want to be focusing on markdown instead. An intuition I have developed is that an ideal prompt…
killthebuddha on HNI'm having success with simple YAML schema. One thing that's very helpful for the prompt is to include "description" and "example": ``` - column name: salary_max format: number…
mmaia on HN
Include examples in the schema prompt
From 1 comment
I'm having success with simple YAML schema. One thing that's very helpful for the prompt is to include "description" and "example": ``` - column name: salary_max format: number…
mmaia on HN
Questions
Why are retry errors appended as system messages rather than user messages
From 1 comment
Very cool. Out of curiosity, for the retries why are the errors appended as system messages as opposed to appending to the user message? And in either case, would it help to…
SeanA208 on HN
Is CUE validation better than pydantic or zod
From 1 comment
Pretty good pitch from ChatGPT if you ask me! I haven't used cue outside of test projects before, so a few questions: 1. Is cue's validation a material improvement from something…
icyfox on HN
Alternatives named
pydantic-chatcompletion
From 1 comment
Here's 40 lines of python code that I've found to be unreasonably effective at accomplishing something similar:…
wskish on HN
Guardrails
From 1 comment
Have you considered Guardrails, https://shreyar.github.io/guardrails/, it’s like Pydantic but for LLMs?
armcat on HN
Langchain output parser
From 1 comment
Langchain also has a built in Pydantic output parser https://python.langchain.com/en/latest/modules/prompts/output_parsers/examples/pydantic.html However I've found it to be…
lharries on HN
zod-chatgpt
From 1 comment
I built zod-chatgpt https://github.com/olup/zod-chatgpt a couple of weeks ago to serve same purpose in typescript
olup on HN
jsonformer
From 1 comment
Hey HN, I've been using GPT a lot lately in some side projects around data generation and benchmarking. During the course of prompt tuning I ended up with a pretty complicated…
icyfox on HN
struct-gpt
From 1 comment
I'm curious what you might think of https://github.com/knowsuchagency/struct-gpt
knowsuchagency on HN