How can I increase the GPU utilization while training / finetune an LLM with unsloth · unslothai/unsloth#3957

(1 评论) (0 反应) (0 负责人)Python (64,271 star) (5,658 fork)batch import

help wanted

描述

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

贡献者指南

技术栈: pythonpytorch
领域: machine learning
议题类型: performance
难度: 3
预计时间: under 1 hour
活动状态: needs maintainer response
清晰度: mostly clear
前置要求: familiarity with PyTorchbasic understanding of LLM fine tuning
新手友好度: 30
研究方向: Review the linked Colab notebook to identify potential bottlenecks like small batch sizes or data loading issues. Suggest increasing batch size or using gradient accumulation. Check unsloth documentation for performance optimization tips.