You gave me plenty to work with! Thanks for your help and kindness, stranger!

Could I trouble you for some advice? I’ve not been developing for very long and I do want to make this is as good as I can. Could I ask what was concerning so I can work to correct?

Of course! Here’s a link to the repo and an example of some of my backend code.

github.com/Thrasher-Int...

Not sure what you mean really, I use AI with tools like Zed to help me code all the time.

This is a really cool project! I’ve been looking for something new to get into, but you know how crazy the dev life is and with a 9-5 stacked on top and so many new games all the time, it’s hard to know what I might like to try.

Tell me again that AI isn't changing the game. #buildinpublic #AIDev @zed.dev

GitHub - Thrasher-Intelligence/sigil: A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. - Thrasher-Intelligence/sigil

github.com/Thrasher-Int...

It’s been a second since I’ve posted but I wanted to share my progress on something I’m working on! Sigil is a Python + React based LLM runner for models downloaded from Hugging Face. I’ve recently added tabs, enhanced theme support, direct downloading from HF. Link below!

#localai #buildinpublic

We got there. Updated chat and instruction modes as well. Enjoy!

Rigging up theme switching was easy. Making good looking themes, another challenge yet.

Theme support does not work yet, but it will soon. Stay tuned!

GitHub - Thrasher-Intelligence/sigil: A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. - Thrasher-Intelligence/sigil

Fresh update to Sigil! Inference is much smoother and chat is actually enjoyable now. I'd love for you to check it out and let me know what you think!

GitHub - Thrasher-Intelligence/sigil: A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. A lightweight, local-first LLM development studio. Build, test, and customize inference workflows with your own models — no cloud, totally local. - Thrasher-Intelligence/sigil

github.com/Thrasher-Int...

Been working on a React version for the frontend of my inference engine and ended being a little more than that. I'd love if you can check out Sigil and see how it can help with your projects! Link in replies. #ai #localllm #llm #oss #opensource

A little demo of inference using the Prometheus engine I built. The backend is less than 100 lines and easy to learn with or to just drop into your projects. Repo's in my bio, check it out! #AI #OSS #Python

Is there anything that you feel like you're missing with the inference frameworks you're using now? Are there any features you think you'd enjoy that you currently don't? I've been working on that sort of thing lately and would love to hear from the community. #ai #llm #llmdevs

GitHub - Thrasher-Intelligence/prometheus: Run LLMs locally with a minimal FastAPI + curses UI setup. No frills, v barebones. Run LLMs locally with a minimal FastAPI + curses UI setup. No frills, v barebones. - Thrasher-Intelligence/prometheus

Just open sourced Prometheus, a simple Python-based inference engine you can run locally. Plug in a model, start chatting, and crack it open to learn how it works.

This isn’t a polished app, it’s a starting point. If you’ve been meaning to mess with LLMs under the hood, this is for you. #ai