Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Various] Proposed Functionalities / Roadmap (from some guy) #85

Open
3 of 6 tasks
zudsniper opened this issue Jun 21, 2023 · 4 comments
Open
3 of 6 tasks

[Various] Proposed Functionalities / Roadmap (from some guy) #85

zudsniper opened this issue Jun 21, 2023 · 4 comments

Comments

@zudsniper
Copy link

zudsniper commented Jun 21, 2023

Hello kurtosis tech team!

I have various thoughts about this project, all of which are simply ways to improve -- at least as I see it -- the usability, execution speed, functional application space, or simply the quality of life. This is to say, I think the current state of this project is already great.

BUT...

Here is my vision for how this package could be used which would, for me at least, supercharge usage of this project.

💬 Suggestions

These are not provided in any particular order, and it may well be that some should definitely be implemented before others. I leave that up to the more knowledgeable.

I guess you can't have scrollTo id links to Headers in issues, or I'm being dumb... Either way sorry, the links above are, save the first, bogus.

🧱 Supporting Local/Unpublished/Private Auto-GPT Plugins

I think it would be a very useful feature to allow users who are actively developing, working with a proprietary codebase, or for any other reason are deploying autogpt with currently unpublished forks of Auto-GPT-Plugin-Template to attach their local code to an instance of autogpt-package.

As for the implementation of this, I am less clear on how it would work, as I don't fully understand kurtosis under the hood. However, I do think it is possible. I also always think that... but I digress

📦 Adding autogpt Branch/Version Specification at Runtime

A method with which we could specify the version of autogpt we want to use would be very helpful, especially given the drastic changes version to version -- such as the scrapping of the MEMORY_BACKEND system in Auto-GPT v0.4.0.
Currently, I would like to be able to use autogpt version 0.3.1, which can be located under the branch stable-0.3.1 in the Auto-GPT repository.

📑 Creating a Wiki / Nested Documentation

While you guys have definitely documented things well enough to use your project through just README.md, I think it's time you have a full-on wiki. I think the easiest, most user-friendly way to do this would be to utilize GitHub repository Wikis, but there are various possible ways. I would be happy to help create such documentation, if you would like.

⏸ "Pausing" & "Unpausing" Instances for Prolonged Use / Execution

Finding a way to save the state of your autogpt-package instance, write it to a file or files, and then return to it at a later date, would make utilizing this tool much nicer in terms of quality of life (in my opinion of course)

This could also potentially be a benefit when it comes to fault-tolerance / error recovery, which would be a great thing to support, both implemented by you guys, but also opening up a way for us to write our own conditions for "failure" and hook in our own subsequent handling systems.

👑 Constructing a "build" System for Configuration / Local Caching / Pausable Instances / etc

This one is by far the biggest ask. (save the best for last or something)

I am personally tired of having to provide all my arguments to the project in the form of a single CLI string. Even escaping special characters aside, growth while ultimately hampered by this constraint adds in my opinion unnecessary overhead.
As an alternative, I propose a build system not unlike that of npm or setuptools for Node.js and Python respectively, involving a json1 file, maybe by convention called agpkg_config.json2, which would facilitate multiple executions of the same instance -- especially if pausing is supported. (somehow)

Here's my suggestion, in terms of file examples.

agpkg_config.jsonc3

// =============================================================================== //  
/*
               __                    __                         __                  
  ____ ___  __/ /_____  ____ _____  / /_      ____  ____ ______/ /______ _____ ____ 
 / __ `/ / / / __/ __ \/ __ `/ __ \/ __/_____/ __ \/ __ `/ ___/ //_/ __ `/ __ `/ _ \
/ /_/ / /_/ / /_/ /_/ / /_/ / /_/ / /_/_____/ /_/ / /_/ / /__/ ,< / /_/ / /_/ /  __/
\__,_/\__,_/\__/\____/\__, / .___/\__/     / .___/\__,_/\___/_/|_|\__,_/\__, /\___/ 
                     /____/_/             /_/                          /____/       
*/
// =============================================================================== // 
// VERSION: ${VERSION}
// <insert project credits, authors, disclaimers, whatever else> 
{
"log_level": "DEBUG" // simple stuff like log level can go here
"model_type":"GPT4ALL", // "OpenAI", "GPT4ALL", "llama.cpp", etc (potentially C:)  
"autogpt": {
   "branch":"stable-0.3.1", 
   /* https://github.com/Significant-Gravitas/Auto-GPT/blob/master/.env.template 
      This file (.env) includes all necessary configuration for autogpt, even removing the need to pass
      OpenAI API Keys to `kurtosis` in plain-text. */
   "env":"path/to/a/.env",
   "settings": "path/to/ai_settings.yaml"  
   /* ...etc */
  }, 
// if "model_type" is GPT4ALL... (or if you have some other want from the config process)
"gpt4all": {
  "local": false, // if true, "model" should be a local filepath instead of a URL.  
  /* if "local": true & "model" typeof URL, then a download of the model could be performed to a standard default location -- probably just ${CWD}/models/... -- wherein after the download is finished, the URL value of "model" is replaced with the local filepath to the newly downloaded model. (this is a bit much to ask, but would be very, very cool!) */  
  "model": "https://gpt4all.io/models/ggml-gpt4all-l13b-snoozy.bin",  
  "temperature": 0.28
  /* other LLM settings */  
  },
"network": {
   "proxies": [
     {
       "url":"http://localhost",
       "port": 80
     }
   ]
  // idk what else, VPN via OpenVPN support? or user-agent swapping? seems like it would be
  // specified in the autogpt specifications
  }
}

With this concept, a .env file which is exactly the same format as that which is used in standard autogpt instances can be utilized off-the-shelf by a new user intending to try out autogpt-package. Same idea with ai_settings.yaml.

Along with that, all the specification for other things, such as gpt4all configuration, or networking, or any other custom stuff you decide to put on top of autogpt in this package to increase its usefulness, could all be specified from a central file, and we wouldn't need to parse and merge .env files into json data then escape it. It just seems like this would be a more scalable approach in my eyes, but I could be wrong.

I also understand this would NOT be as simple as I have put it out to be -- it would be a somewhat major undertaking for the codebase -- but I wanted to put my suggestion out here. Feel free to use it as you see fit; I will not be offended. C:


🎀 Summary

Here is the current README.md's method of initiating autogpt-package with a redis memory backend.4

$ kurtosis run github.com/kurtosis-tech/autogpt-package '{"GPT_4ALL": true, "MODEL_URL": "https://gpt4all.io/models/ggml-gpt4all-l13b-snoozy.bin"}'

Here is how I would have things work, in a perfect world for me -- which is obviously not everyone's perfect world.

$ kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt `/path/to/agpkg_config.json`  

Alright, this is insanely long, and I apologize in advance! Hopefully something in here will inspire somebody. Thank you to everyone at kurtosis-tech again for making this package. It was the first way that I got Auto-GPT to work at all when I was first playing with it, and it was as seamless as it purported to be.
But then I wanted more... hah.

Cheers,
Jason, or zod

Footnotes

  1. Or yaml, or even just xml if you don't like us very much. Format doesn't really matter here

  2. This should be able to be overwritten by a CLI flag, perhaps --build-configuration <filename.json>

  3. This is just a simple extension of json which supports multi line, single line, and inline comments using js syntax. Here is a parser from Microsoft written for Node.js This is not required by any means; simple json should be used, & developers can precompile our jsonc or jsonl into appropriate json conforming structures on our own time.

  4. I realize that the README.md document is, while replete with information about the actual use of the repository, a bit scattered. I think you guys should add either just a docs/ directory with various .md files, or a full-on GitHub Pages Wiki -- but I will make a separate issue for that

@h4ck3rk3y
Copy link
Collaborator

h4ck3rk3y commented Jun 21, 2023

Hey! Thank you for being so thorough with the ticket & list of features :) Addressing a few of these for now -

📦 Adding autogpt Branch/Version Specification at Runtime

Kurtosis runs of Docker Images that are published to the DockerHub (or that are available locally in your Docker). I have added support for changing the version of AutoGPT we spin up using the AUTOGPT_IMAGE flag. Last code snippet in this section.

📑 Creating a Wiki / Nested Documentation

Would love for you to start doing this! You should be able to do so now!

👑 Constructing a "build" System for Configuration / Local Caching / Pausable Instances / etc

This I am trying to understand a bit more! Say you a have a file called args.json in your current working directory and say it looks like

{"GPT_4ALL": true, "MODEL_URL": "https://gpt4all.io/models/ggml-gpt4all-l13b-snoozy.bin"}

You should be able to pass it to Kurtosis using

kurtosis run github.com/kurtosis-tech/autogpt-package --enclave autogpt "$(cat args.json)"

Does that work for you?

🧱 Supporting Local/Unpublished/Private Auto-GPT Plugins

I can think of some ways of supporting this. I can definitely add support for repositories that aren't in the main.star but are still public and available on GitHub.

⏸ "Pausing" & "Unpausing" Instances for Prolonged Use / Execution

This we are working on as a team as an addition to core Kurtosis where users should be able to have persistent environments that they can stop and restart. This might take a while!

@zudsniper
Copy link
Author

📦 Adding autogpt Branch/Version Specification at Runtime

Kurtosis runs of Docker Images that are published to the DockerHub (or that are available locally in your Docker). I have added support for changing the version of AutoGPT we spin up using the AUTOGPT_IMAGE flag. Last code snippet in this section.

Sweet, and thank you for pointing me to this.

📑 Creating a Wiki / Nested Documentation

Would love for you to start doing this! You should be able to do so now!

Sure, though it might be sporadic due to my obligations at work and at home. I do write markdown really, really hard though... whatever that means.

👑 Constructing a "build" System for Configuration / Local Caching / Pausable Instances / etc

This I am trying to understand a bit more! Say you a have a file called args.json in your current working directory and say it looks like
[etc]

Yes... I could definitely just do this... don't mind me, I was tired or something oops!

🧱 Supporting Local/Unpublished/Private Auto-GPT Plugins

I can think of some ways of supporting this. I can definitely add support for repositories that aren't in the main.star but are still public and available on GitHub.

Adding support for public repositories is definitely the right direction to go in, but ideally I think being able to develop proprietary ones, or just develop offline, would make this project usable as the only development tool necessary for autogpt. If you guys think it's the right thing to do or not: that's up to you! haha

⏸ "Pausing" & "Unpausing" Instances for Prolonged Use / Execution

This we are working on as a team as an addition to core Kurtosis where users should be able to have persistent environments that they can stop and restart. This might take a while!

That's ok! It's cool to know I stumbled upon a feature under active development. It's certainly a difficult problem, especially when addressed from the scale of the entire kurtosis project. I hope to see it developed one day -- I'll be waiting C:

Thanks for the thorough response to my thorough suggestions. I will be sharing this project as much as possible to other autogpt enthusiasts, especially with the GPT4ALL addition - I would say that is a game changer.

@mieubrisse
Copy link
Member

Hey @zudsniper , catching up on this thread! As I understand, the two potential outstanding pieces of work are:

  1. Supporting Local/Unpublished/Private Auto-GPT Plugins
    2."Pausing" & "Unpausing" Instances for Prolonged Use / Execution

For no. 2, we're tracking something sort of similar here: kurtosis-tech/kurtosis#705 (though I think we'll also need the persistent data that I mentioned in #84 ).

For no. 1, would you mind opening a new ticket for just that piece of work, and what you had in mind, so we can track it individually? That way we can close this issue and track the individual followups

@zudsniper
Copy link
Author

Hey @zudsniper , catching up on this thread! As I understand, the two potential outstanding pieces of work are:

  1. Supporting Local/Unpublished/Private Auto-GPT Plugins
    2."Pausing" & "Unpausing" Instances for Prolonged Use / Execution

For no. 2, we're tracking something sort of similar here: kurtosis-tech/kurtosis#705 (though I think we'll also need the persistent data that I mentioned in #84 ).

For no. 1, would you mind opening a new ticket for just that piece of work, and what you had in mind, so we can track it individually? That way we can close this issue and track the individual followups

Sure, I will make an issue just for it @mieubrisse

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants