AI & Tech Developments - Apr 07

I'm open-sourcing an MLX port of the ESM-2 protein model family so more people can tinker with AI & biology on Apple silicon! (1/5)

— @josephjojoe

You can now fine-tune Gemma 4 (and 500 other open source models) in a free Google Colab

— @itsPaulAi

I think people may overlooking why we would want to quantize to 1-bit/1.58-bit in the first place.

— @leloykun

llm-wiki version 0.0.10 is out!

— @nvk

Stop scrolling. This might be one of the most important thing happening on Hyperliquid right now and almost nobody is talking about it.

— @Yaugourt

TurboQuant+ update. New build dropped from @spiritbuun with a lot of changes.

— @Bent302

Okay, I have no clue how he did it but this is the best 27B distill I've used to date.

— @JoelDeTeves

During the last week I executed very long autonomous sessions of Claude Code Opus 4.6 and Codex GPT 5.4.

— @antirez

Open-sourcing an MLX port of the ESM-2 protein model family to enhance AI and biology integration.

— @josephjojoe

A new method allows fine-tuning of Gemma 4 and other models in Google Colab.

— @itsPaulAi

Discussion on the benefits of quantizing models to 1-bit/1.58-bit for efficiency.

— @leloykun

The latest version of llm-wiki is out, enhancing AI tools.

— @nvk

A significant development in Hyperliquid with an independent client achieving block hash parity.

— @Yaugourt

TurboQuant+ update introduces support for new architecture, enhancing performance.

— @Bent302

A new distillation of the Hermes Agent shows improved performance with Hermes models.

— @JoelDeTeves

Insights on executing long autonomous sessions with Claude Code and Codex models.

— @antirez

OpenAI seems to be working on an upgraded agent builder, internally named "Hermes," which will likely be integrated directly into ChatGPT. And "Pluto Model".

— @testingcatalog

Hermes Agent: This open-source beast actually grows with you — it learns from experience, builds its own skills, improves over time, and keeps persistent memory. Feels like having a truly self-improving AI teammate. Huge thanks to the team!

— @HaseebMir91