DEV Community

Gemma

Gemma is a collection of lightweight, state-of-the-art open models built from the same technology that powers our Gemini models.

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Which Gemini Model Should You Use? Neon, Argon, Hydrogen, Carbon, Flash or Gemma. It Depends.

Which Gemini Model Should You Use? Neon, Argon, Hydrogen, Carbon, Flash or Gemma. It Depends.

7
Comments
10 min read
Thinking On or Off? Gemma 4 E4B on an 8 GB RTX 3050

Thinking On or Off? Gemma 4 E4B on an 8 GB RTX 3050

10
Comments 1
9 min read
llama.cpp on a Cloud TPU: An Unofficial ggml Backend Through XLA

llama.cpp on a Cloud TPU: An Unofficial ggml Backend Through XLA

2
Comments
10 min read
Five Serving Engines on a Lambda GPU: Gemma 4 E2B Under vLLM, SGLang, llama.cpp, PyTorch and JAX

Five Serving Engines on a Lambda GPU: Gemma 4 E2B Under vLLM, SGLang, llama.cpp, PyTorch and JAX

2
Comments
11 min read
Gemma 4 on a Hosted Colab TPU: The QAT Repack, Served in Two Minutes

Gemma 4 on a Hosted Colab TPU: The QAT Repack, Served in Two Minutes

22
Comments
12 min read
Serverless Gemma 4: E2B on a Lambda GPU Function, with Google's QAT Weights and Drafter

Serverless Gemma 4: E2B on a Lambda GPU Function, with Google's QAT Weights and Drafter

24
Comments 2
10 min read
On-Device AI in Flutter: Run Gemma 2 & Local LLMs

On-Device AI in Flutter: Run Gemma 2 & Local LLMs

1
Comments
7 min read
Gemma 4 on AWS Lambda with a GPU: E2B on an NVIDIA L4, Day One

Gemma 4 on AWS Lambda with a GPU: E2B on an NVIDIA L4, Day One

11
Comments
10 min read
Hands-On with Diffusion Gemma: Another Gemma with a Diffusion Mechanism

Hands-On with Diffusion Gemma: Another Gemma with a Diffusion Mechanism

Comments
12 min read
Thinking On or Off? Gemma 4 E2B on an 8 GB RTX 3050

Drafters double speed while thinking adds nothing

Thinking On or Off? Gemma 4 E2B on an 8 GB RTX 3050

25
Comments 7
9 min read
Weight Space Programming: Giving Gemma 4 E2B a Space Needle Obsession with Gemini 4

Weight Space Programming: Giving Gemma 4 E2B a Space Needle Obsession with Gemini 4

Comments
6 min read
NoteForge: Offline-First Intelligent Learning Platform

NoteForge: Offline-First Intelligent Learning Platform

Comments
3 min read
Withdrawing, and Sharing Everything We Measured: 1,800+ Trials of a 4B Quantized SWE Agent

Withdrawing, and Sharing Everything We Measured: 1,800+ Trials of a 4B Quantized SWE Agent

1
Comments 7
6 min read
High or Low Thinking? Gemini 4 and Gemma 4 on Hard Math, Cost First

High thinking cost 4x more for zero gain

High or Low Thinking? Gemini 4 and Gemma 4 on Hard Math, Cost First

27
Comments 11
11 min read
Hidden Objects: an escape room where the narrator would rather you stayed

Hacktoberfest Open-Source AI Challenge: Ghost in the Machine Submission 👻🌼

Hidden Objects: an escape room where the narrator would rather you stayed

5
Comments
5 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.