[HN Gopher] Show HN: Agentflow - Run Complex LLM Workflows from ...
___________________________________________________________________
Show HN: Agentflow - Run Complex LLM Workflows from Simple JSON
So, it feels like this should exist. But I couldn't find it. So I
tried to build it. Agentflow lets you run complex LLM workflows
from a simple JSON file. This can be as little as a list of tasks.
Tasks can include variables, so you can reuse workflows for
different outputs by providing different variable values. They can
also include custom functions, so you can go beyond text generation
to do anything you want to write a function for. Someone might
say: "Why not just use ChatGPT?" Among other reasons, I'd say that
you can't template a workflow with ChatGPT, trigger it with
different variable values, easily add in custom functions, or force
the use of custom functions for steps in the workflow. Someone
might also say: "Then why not use Auto-GPT or BabyAGI?" Among other
reasons, I'd say you can't if you want consistency because these
tools operate autonomously, creating and executing their own tasks.
Agentflow, on the other and, lets you define a step-by-step
workflow to give you more control. I'd like to do more with this,
including adding more custom functions, and more examples, and more
ways to trigger workflows (such as in response to events). But
first, I want to make sure I'm not wasting my time! For starters,
if something like this already exists, please tell me.
Author : simonmesmith
Score : 36 points
Date : 2023-08-08 17:57 UTC (5 hours ago)
(HTM) web link (github.com)
(TXT) w3m dump (github.com)
| dchuk wrote:
| Looks awesome. Any plans to allow for it to use local LLMs (like
| llama) instead of openai APIs?
| simonmesmith wrote:
| Great question, and something I've thought about.
|
| The main issue right now is that I'm relying on the OpenAI
| API's function-calling capabilities to enable the use of
| functions in workflows.
|
| If we switch to other LLMs, we'll need to create some
| additional wrapping around them to allow such function calling.
| (As far as I know. If any open source LLMs already have
| function calling built in, let me know.)
|
| LangChain (which I'm not using due to personally finding it
| overcomplicated and too obfuscating) does have function
| calling, but it uses approaches like REACT (if I remember
| correctly) that aren't as reliable as OpenAI's approach.
| gremlinsinc wrote:
| If ever there was a place where a standard would be nice,
| it's with LLM API's.
| jmarbach wrote:
| Thank you for creating this. I've been looking for something that
| provides a more lightweight alternative to LangChain.
| simonmesmith wrote:
| Thank you for saying so! It's very gratifying that I'm not the
| only one who finds this useful :)
| hk1337 wrote:
| Seems cool. Seems a little like an open source version of Step
| Functions you only need python to run.
|
| I did a similar, rudimentary, version where I save the JSON in
| DynamoDB and the tasks indicate how to transform tabular (Excel,
| CSV) data. From renaming columns, adding columns, and transposing
| columns into new rows.
| simonmesmith wrote:
| Perhaps we could give people the option of saving workflows in
| different places. I went with JSON as I want to make it as easy
| as possible for people to create new workflows. Plus, they also
| get versioned easily this way.
| rexreed wrote:
| I'm trying to understand what this does and how it works. Can you
| provide more examples with different use cases? How does this
| work with / compare to LangChain?
| simonmesmith wrote:
| For sure.
|
| Let's use an example, like this:
| https://github.com/simonmesmith/agentflow/blob/main/agentflo...
|
| This is a workflow for coming up with a product idea and
| illustrating it with an image.
|
| This workflow has 9 steps, starting with brainstorming ideas,
| and ending with saving an HTML file containing the product
| name, description, and image.
|
| It also has two variables, {market} for the target market, and
| {price_point} for the price point.
|
| To run this workflow, you simply enter this in the command
| line: python -m run --flow=example_with_variables --variables
| 'market=college students' 'price_point=$50'
|
| You don't need to write any code with LangChain.
|
| You simply specify your workflow in a JSON file, and execute
| it.
|
| Does that help to clarify?
| rexreed wrote:
| Thanks - this does help. Curious about the function calls,
| especially around image generation.
|
| Also, can you clarify what you mean by "You don't need to
| write any code with LangChain."?
| simonmesmith wrote:
| Sure!
|
| In Agentflow, you write functions by inheriting from the
| BaseFunction class. You need to provide the definition in
| JSON that GPT-3.5/4 uses to understand how to call a
| function, and also the function logic itself. This just
| means creating a get_definition() function that returns a
| JSON Schema object, and an execute() function that performs
| your logic and returns a string. Once you have those, you
| can then just use the function in your workflow by adding
| "function_call": "your_function". The application does the
| rest. Here's the create_image function, for example, which
| uses the Dall-e API: https://github.com/simonmesmith/agentf
| low/blob/main/agentflo...
|
| What I mean by "you don't need to write any code with
| LangChain" is that you don't need to write any Python at
| all to use Agentflow, unless you want to create a new
| function. Creating workflows just involves creating JSON
| files. It's not like LangChain, for which you'd have to
| chain together multiple prompts in Python.
|
| Does that help clarify?
|
| PS: You'll notice heavy documentation in the link above. I
| want to experiment with automatically generating
| documentation using Sphinx, so I documented everything with
| Sphinx formatting. It might be overkill.
| rexreed wrote:
| Thanks! I'll check this out!
| toomuchtodo wrote:
| Y'all should team up with the Magic Loop folks:
| https://news.ycombinator.com/item?id=36958731
|
| And if you want to talk to a bunch of APIs:
| https://news.ycombinator.com/item?id=37020783
| simonmesmith wrote:
| Yeah, I saw the Magic Loop on HN and was like, "oh no, time to
| dump my project!"
|
| But I think there are some key differences. I'm not sure if
| Magic Loop is open source, for example, so I don't know how
| they're currently building their workflows.
|
| Automating API calls is neat, too, but I get a bit anxious with
| too much automation when you want to run a workflow with
| consistent results. My gut says you want to have the functions
| here pretty locked down and tested, rather than rely on
| automating API calls. But maybe I'll be proven wrong.
| whalesalad wrote:
| kinda reminds me of dagster or langchain. do you anticipate
| building a huge library of functions like 'save_file' that would
| add up to a library or is that intended to be left to the reader?
| if the latter, the fact that this is based on json feels kinda
| moot.
| simonmesmith wrote:
| I definitely plan on adding more functions, and hopefully
| having others do the same. And I think these should be
| versatile things like get_url(), which I added today, versus
| the very specialized plugins that seem to dominate something
| like ChatGPT's plugin space. I think we want reliable,
| versatile building blocks as functions.
___________________________________________________________________
(page generated 2023-08-08 23:02 UTC)