In this video I want to take an initial look at the new DiffusionGemma from Google. A new technique in token generation, I will run it through some rudimentary tests of my own as well as discuss the pros and cons of this new technique.
The model is running on a local AI PC I have built with 16GB VRAM and 32GB DDR4 RAM.
If you’re interested in local LLMs, AI and homelabs from the perspective of a software engineer with many years of professional experience working with LLMs in production – feel free to subscribe!
Google Blog: https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/
Model: https://huggingface.co/unsloth/diffusiongemma-26B-A4B-it-GGUF
GitHub: https://github.com/lukesdevlab/youtube
Patreon: https://www.patreon.com/cw/LukesDevLab
#localllm #localai #homelab #llamacpp #homelab #gemma4 #quantization #qat #26b #diffusiongemma #gemma #diffusion
Chapters:
0:00 Intro
0:06 DiffusionGemma Info
1:59 Model Info
2:34 System Specs
2:51 Reasoning
5:18 Simulated Agency
6:33 My Thoughts
source
Leave a Reply