do you consider construct agents to consume arbitrary APIs, here APIs conformant to the OpenAPI/Swagger specification.

Just wanted to say you're a hero.

I was hesitant to use Guice or any other dependency injection framework(I actually tried but deleted it later on) as I am afraid the decision would enforce every user who is using our library to use Guice or Spring.
The consequence without DI is that components are tightly coupled.
: (
This makes me feel worse.

After research, it turns out it shouldn't matter.
There's https://github.com/spring-projects/spring-guice that could help convert the Guice module to the corresponding part in Spring.

Even without the repo above, we can also do this: https://stackoverflow.com/questions/18826308/integrate-guice-component-into-spring-application

As we are switching to the dependency injection way, some code will need to be changed. For example, we would write less static factory methods that we did a lot. It makes no sense to create it if we are instantiating via the injection.

Adding async support to the capability unit and the graph

Every public class or protected class should have java doc comments.
Every class should have at least one unit test.

I will look for potential github app that would comment on PR if not done so.

Hi KnowlyAI

I really like your library - it is a most useful contribution, so thank you.
However it cannot be used legally without a license :)
Please add one so I know what I can and cannot do with your code.

Cheers, and keep up the great work.

Seamus

Documentation: https://huggingface.co/docs/hub/api

Langtorch hub provides users a single entry point to all components.

Agent can construct thoughts and take action till meeting the goal.

Hey guys, I hope you are all doing well!

I'm wondering how can I create a trivia game with this?

For example, I want the template for the gpt generation to follow:
{$question}
{$optionA}
{$optionB}
{$optionC}
{$optionD}
{$answer}

Thanks in advance!

exponential backoff retry

can add embedding support for cohere, similar to openai

I hope to join the development of langchain-java, but I found that the Discord invitation link has expired.

Currently, the Conversation Memory Class can store an arbitrary number of strings with no maximum length limitation.

However, the language model has a fixed prompt length.

Feature Requirements:
Add a parameter for the maximum token length in the Memory Class. If not specified, the Memory Class should default to the current behavior with no maximum string length.

If a string is added to the Memory Class that exceeds this maximum length, the Memory Class should truncate the messages to fit within the maximum length. [remove the oldest messages]

Design a composable extension framework that can easily swap and equip with different extensions.

For example,

CapabilityUnit baseCapabilityUnit = CapabilityUnit.create(...)

CapabilityUnit unitWithLoggingExtension = LoggingExtension.create(baseCapabilityUnit)

CapabilityUnit unitWithInputParser = InputParserExtension.create(unitWithLoggingExtension)

Design is not finalized yet.

Hello,

I tried to import the release "0.0.10" in a Spring Boot project, but on resolving dependencies I got a failure:

I think using openai-java library is bit awkward here as the library doesn't support async or future-based style but in reactive way.

https://openai.com/blog/function-calling-and-other-api-updates

https://github.com/pgvector/pgvector-java