HighNoon LLM Training Update & Future Plans

[versoindustries]

We're excited to share the latest progress on HighNoon LLM, our open-source AI project at Verso Industries. Here's what's happening with the model, upcoming releases, and our broader vision!

Model Training Update

HighNoon LLM is currently training on the Alpaca dataset. As training progresses, we'll release gradient checkpoints to the GitHub repository. These checkpoints will include:

• Loss values for transparency.
• Status updates on training progress.
• Insights into what to expect from future releases.

Note: Datasets are trained sequentially. The first checkpoint will represent the first epoch and gradient checkpoint of the Alpaca dataset. Stay tuned for announcements as we upload these checkpoints starting July 2025!

Inference GUI in Development

We’re also working on an inference-end GUI to make HighNoon LLM easier to use. This GUI is not yet released, as our single machine is currently handling all model training. Once we secure additional funding, we’ll:

• Dedicate a separate system for training.
• Free up resources to focus on coding and testing the inference GUI.

This will allow us to polish and release the GUI for community testing. We’re excited to bring you a user-friendly way to interact with HighNoon LLM!

Verso Backend: Expanding Our Open-Source Ecosystem

In addition to HighNoon LLM, we maintain the Verso Backend, an open-source Python-based backend hosted under our GitHub profile. We’re exploring the possibility of developing it into a robust, open-source foundation for broader use. Its future depends on the performance and community feedback for our open-source PyQt5-based inference GUI.

How You Can Help

HighNoon LLM and Verso Backend are ambitious projects, and we need your support to grow! Here’s how you can get involved:

• Contribute: Join our open-source community by coding, testing, or sharing feedback. See our Contributing Guide.
• Sponsor: Help us fund dedicated training hardware and development. Check out our sponsorship tiers starting at just $5/month.
• Spread the Word: Share this update on X or other platforms with #HighNoonLLM.

What’s Next?

• Checkpoint Releases: Starting July 2025, expect regular updates with model checkpoints and detailed training logs.
• GUI Development: A polished inference GUI once training is offloaded to dedicated hardware.
• Verso Backend Expansion: Potential growth into a standalone open-source tool based on community needs.

Join us on Discord or follow our GitHub for real-time updates. Let’s build AI that’s open, efficient, and accessible to all!

#HighNoonLLM #OpenSource #AI

[FabianaCampanari]

⠀.　　　　　　　　　　⠀⠀⠀✦ ⠀ ⠀　　　　　　　　　　　　　　⠀⠀⠀⠀⠀* ⠀⠀⠀.　　　　　　　　　　. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀✦⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀ ⠀⠀⠀⠀⠀⠀.　　　　　　　　　　　　　.　　　ﾟ .　　　　　　　　　　　　　. 　　　　　　　　　　　　　　　✦ 　　　　　,　　　　　　　.⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀　　　　　　*　　　　　　　　　　　..　　　　　　　　　　　　　. 　　✦⠀　   　　　,　　　　　　　　　*　　　　　⠀　　　　⠀　　,⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀.　　　　　 　　⠀　　　⠀.　　　˚　　　⠀　 🌕 .⠀  　　,　　　　　　.　　　　　　　　　　　　　.　　　　　　*⠀　　⠀  　　　　　⠀✦⠀　　　　　　　*　　　　　　　　　　　　　　　　　　.　　　　.　　　　.　　　⠀　　　　　　　　　　　.　　　　　　　　　　˚　　　　　　　　ﾟ　　　　　.　.⠀　　⠀‍⠀‍⠀‍⠀‍⠀‍⠀‍⠀‍⠀‍⠀‍⠀‍⠀,　　　*　　⠀.　　　　　.　　　　　　　　　　⠀✦　˚　　　　　　　　　　　　　　.⠀ 　　　　　　　　　　.　　　　　　　　.　　　　　✦⠀　   　　　,　　    　　　　　　　　.⠀⠀⠀⠀⠀⠀.　　　　　　　　　　⠀ ✨ * ⠀⠀✦ ⠀ ⠀　　　　　　　　　　　　　　⠀⠀⠀⠀⠀ ⠀⠀⠀.　　　　　　　　　　. ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀✦⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀✨ * ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀ ⠀⠀⠀⠀⠀⠀.　　　　　　　　　　　　　.　　　ﾟ .　　　　　　　　　　　　　. 　　　　　　　　　　　　　　　✦ 　　　　　,　　　　　　　.⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀　　　　　　*　　　　　　　　　　　..　　　　　　　　　　　　　. 　　✦⠀　   　　　,　　　　　　　　　*　　　　　⠀　　　　⠀　　,⠀⠀ * ⠀⠀⠀⠀⠀⠀⠀.　　　　　 　　⠀　　　⠀.　　　˚　　　⠀　⠀  ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀ ⠀⠀⠀⠀⠀⠀.　　　　　　　　　　　　　.　　　ﾟ .　　　　　　　　　　　　　. 　　　　　　　　　　　　　　　✦ 　　　　　,　　　　　　　.
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
　　　　　　*　　　　　　　　　　　. ✨ *
.　　　　　　　　　　　　　. 　　✦⠀　   　　　,　　　　　　　　　*
　　　　　⠀　　　　⠀　　,
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀.　　　　　 　　⠀　　　⠀.　
　　˚　　　⠀　⠀  　　,　　　　　　.
　　　　　　　　　　　　　.
　　　　　　*⠀　　⠀  　　　　　⠀✦⠀　
　　　　　　*　　　　　　　　　　　　　　　　　　.
　　　　.　　　　.　　　⠀
　

⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀

Hey Michel, thanks for sharing, HighNoon’s shaping up to be seriously cool.
I’ll be ready to jump in and follow more closely starting July. Would love to learn from the way you’re approaching it.

If you’re up for it, I’d really appreciate a bit of guidance along the way 👀

Let’s sync soon!

⊹ ࣪ ﹏𓊝﹏𓂁﹏⊹ ࣪ ˖