AIbase
EN
Home
Gemma 3 4b It Qat 4bit
G

Gemma 3 4b It Qat 4bit

Developed by mlx-community
Gemma 3 4B IT QAT 4bit is a 4-bit quantized large language model trained with Quantization-Aware Training (QAT), based on the Gemma 3 architecture and optimized for the MLX framework.
Image-to-Text
Transformers
 OtherOpen Source License:Other#4-bit quantized inference#Multimodal understanding#Image caption generation
Downloads 607
Release Time : 4/15/2025

Model Overview

This model is a quantized version converted from Hugging Face format to MLX format, suitable for image-text-to-text tasks with multilingual support.

Model Features

4-bit quantization
The model is trained with Quantization-Aware Training (QAT) to operate at 4-bit precision, significantly reducing memory usage and computational requirements.
MLX optimization
Optimized for the MLX framework, delivering efficient inference performance.
Multilingual support
Supports multiple languages, making it suitable for international applications.

Model Capabilities

Image-text generation
Multilingual text processing
Efficient inference

Use Cases

Image understanding and captioning
Image caption generation
Generates detailed textual descriptions based on input images.
Produces accurate and coherent image captions.
Multilingual applications
Multilingual image labeling
Generates multilingual labels or descriptions for images.
Supports image annotation in multiple languages.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase