Magentic
Dev tools
python llm library
SHOW HN2023
How HN took it
14 positive, 8 neutral or asking, 1 mixed, 4 critical, out of 27 comments about the project. Thread score +0.40.
Design: 9 for, 2 againstIdea: 4 for, 2 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
Clean, intuitive decorator-based API design
From 4 comments
The API looks very clean. Today I learned about "..." in Python
smilingemoji on HNReally like how this is implemented with decorators. Everything just feels really smooth
bogtog on HNJust wanna say, that's pretty great API design :)
pedrovhb on HNSuper cool! Looks quite intuitive, especially for function calls.
retrovrv on HN
Fun examples and humor
From 1 comment
This looks really useful. Langchain is not my idea of a fun time. Love the examples too. Low-effort humor is the best: > create_superhero("Garden Man") > # Superhero(name='Garden…
jstarfish on HN
Useful for prototyping LLM logic in code
From 2 comments
This is neat. It makes it easy to prototype, and then you can just remove the decorator and write a specific implementation if you need to.
pphysch on HNI am amazed that `...` is a valid syntax in Python, not a pseudo grammar. This library is impressive, I appreciate it and I will apply it to my project.
hitori on HN
Objections
Only supports OpenAI-compatible chat models
From 2 comments
Is it really LLMs (plural) when you only have OpenAPI integration?
ElectricalUnion on HNDo you support custom LLMs?
czyhandsome on HN
Misuse of Ellipsis instead of pass in function bodies
From 1 comment
It's a misuse of the Python Ellipsis, though PEP has no opinion on it. The Ellipsis is "Special value used mostly in conjunction with extended slicing syntax for user-defined…
politelemon on HN
Framework may hinder chain-of-thought prompting
From 1 comment
I've personally found frameworks like this to get in the way of quality COT: It's rare for a prompt that takes great advantage of the LLM's reasoning to fit in the format these…
BoorishBears on HN
Requests
Support for other LLM backends
From 2 comments
Is it really LLMs (plural) when you only have OpenAPI integration?
ElectricalUnion on HNDo you support custom LLMs?
czyhandsome on HN
System prompt and chat message support
From 1 comment
Does this do System vs. Assistant vs. User prompting?
quickthrower2 on HN
Few-shot examples in prompts
From 1 comment
Awesome job with the simplicity, gonna play with it. Have you tried using yaml as the format with the models instead of JSOn? Feel like you'll use far fewer tokens to describe…
ramraj07 on HN
Questions
Does it return actual code output or call OpenAI each time
From 1 comment
Looks super cool! A few questions: 1) Can you get the actual code output or will this end up calling OpenAI each function call? 2) What latency does it add? What about token…
conor_f on HN
Latency and token usage
From 1 comment
Looks super cool! A few questions: 1) Can you get the actual code output or will this end up calling OpenAI each function call? 2) What latency does it add? What about token…
conor_f on HN
Why a decorator instead of a function call
From 1 comment
Curious as to why you chose to do it as a decorator instead of just a function call?
jedberg on HN
Alternatives named
Marvin
From 2 comments
Seems a lot like https://github.com/PrefectHQ/marvin? The prompting you do seems awfully similar to: https://www.askmarvin.ai/prompting/prompt_function/
cosmonoot on HNAre you familiar with https://github.com/PrefectHQ/marvin? This looks very similar
bl00p on HN
Guidance
From 2 comments
Looks great! I don't normally like these LLM libraries but this one sparks joy. I'll try it out on my next experiment. Could you highlight how you're parsing to structured…
Difwif on HNI tried out guidance. Encountered endless bugs
AmazingTurtle on HN
LMQL
From 1 comment
This is also similar in spirit to LMQL https://github.com/eth-sri/lmql
lachlan_gray on HN
LocalAI
From 1 comment
There's also LocalAI[0] which allows the use of local LLMs with an OpenAI compatible API. [0] https://github.com/go-skynet/LocalAI
michaelmior on HN