unslothai/unsloth

How can I increase the GPU utilization while training / finetune an LLM with unsloth

Open

#3,957 opened on Feb 1, 2026

View on GitHub
 (1 comment) (0 reactions) (0 assignees)Python (5,658 forks)batch import
help wanted

Repository metrics

Stars
 (64,271 stars)
PR merge metrics
 (Avg merge 3d 15h) (525 merged PRs in 30d)

Description

I was fine-tuning a gemma3 4b model with unsloth with a custom dataset in Colab, but when I observed the GPU utilization, it was very low. Is there a way to fully utilize it, maybe use multiple processes?

Here is the Colab notebook I am using: https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Gemma3_(4B).ipynb

Here is a sc I got while training: https://drive.google.com/file/d/1o3P-ORRJPACbGgOglfkbkoj1fV4VahZu/view?usp=sharing

Any help on this would be appreciated Thanks Dinusha

Contributor guide

Research direction
Investigate ways to increase GPU utilization during unsloth fine tuning by adjusting batch size, gradient accumulation steps, using mixed precision (e.g., bfloat16), and considering multi GPU setups. Check the official unsloth documentation for performance optimization tips.
Tech stack
pythonpytorch
Domain
machine learning
Issue type
Research
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Fresh
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Mostly clear
Prerequisites
PythonGPU training basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
60