kairawal/Gemma-3-4B-IT-HI-SynthDolly-r16alpha128-E8-S73

Hugging Face
VISIONConcurrency Cost:1Model Size:4.3BQuant:BF16Ctx Length:32kPublished:May 23, 2026License:apache-2.0Architecture:Transformer Open Weights Warm

The kairawal/Gemma-3-4B-IT-HI-SynthDolly-r16alpha128-E8-S73 is a 4.3 billion parameter instruction-tuned Gemma model developed by kairawal, fine-tuned from unsloth/gemma-3-4b-it. This model was trained using Unsloth and Huggingface's TRL library, emphasizing faster training. It is designed for general instruction-following tasks, leveraging its 32768 token context length for processing longer inputs.

Loading preview...

Model Overview

The kairawal/Gemma-3-4B-IT-HI-SynthDolly-r16alpha128-E8-S73 is an instruction-tuned language model based on the Gemma-3-4B architecture, developed by kairawal. It has 4.3 billion parameters and supports a substantial context length of 32768 tokens, allowing it to handle complex and lengthy prompts.

Key Characteristics

  • Base Model: Fine-tuned from unsloth/gemma-3-4b-it.
  • Training Efficiency: This model was trained with a focus on speed, utilizing Unsloth and Huggingface's TRL library, resulting in 2x faster training compared to standard methods.
  • Instruction Following: Optimized for understanding and executing instructions, making it suitable for a variety of conversational and task-oriented applications.

When to Use This Model

This model is particularly well-suited for developers looking for:

  • Efficiently Trained Models: If training speed and resource optimization are critical, the Unsloth-accelerated training process is a key differentiator.
  • General Instruction-Following: For applications requiring a model to accurately follow user prompts and generate relevant responses.
  • Long Context Processing: Its 32768 token context window makes it effective for tasks involving extensive documents or detailed conversations.