kairawal/Gemma-3-4B-IT-ES-SynthDolly-r16alpha32-E3-S73

VISIONConcurrency Cost:1Model Size:4.3BQuant:BF16Ctx Length:32kPublished:May 14, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The kairawal/Gemma-3-4B-IT-ES-SynthDolly-r16alpha32-E3-S73 is a 4.3 billion parameter instruction-tuned language model developed by kairawal, finetuned from unsloth/gemma-3-4b-it. This model was trained using Unsloth and Huggingface's TRL library, achieving 2x faster finetuning. It is designed for general instruction-following tasks, leveraging the Gemma architecture for efficient performance.

Loading preview...

Model Overview

The kairawal/Gemma-3-4B-IT-ES-SynthDolly-r16alpha32-E3-S73 is a 4.3 billion parameter instruction-tuned language model. Developed by kairawal, this model is finetuned from the unsloth/gemma-3-4b-it base model, leveraging the Gemma architecture.

Key Capabilities

  • Efficient Finetuning: This model was finetuned using Unsloth and Huggingface's TRL library, resulting in a 2x speed improvement during the training process.
  • Instruction Following: As an instruction-tuned model, it is designed to understand and execute commands or prompts given in natural language.
  • Gemma Architecture: Built upon the Gemma family, it benefits from Google's open and lightweight large language model architecture.

Good For

  • Applications requiring a compact yet capable instruction-following model.
  • Developers looking for a model that has undergone efficient finetuning processes.
  • General natural language understanding and generation tasks where the Gemma architecture is preferred.