FLUX.1 is a high-level model supported by AI Toolkit by Ostris, and its training has strict hardware requirements. According to the tool description, running FLUX.1 model training requires an Nvidia GPU with at least 24GB of video memory, a requirement that stems from the complex architecture and large-scale parameter computation needs of the FLUX.1 model.
To accommodate different hardware environments, the toolset offers flexible configuration options:
- In the case of limited video memory, the configuration file can be set to
low_vram: trueIn order to make the model perform quantization on the CPU - For cloud-based training, an A40 graphics card with 48GB of video memory is recommended
- Supports cloud-based platforms such as RunPod and Modal for distributed training
This stringent hardware requirement reflects the advanced nature of FLUX.1 models and demonstrates AI Toolkit's focus on professional-grade model training.
This answer comes from the articleAI Toolkit by Ostris: Stable Diffusion with FLUX.1 Model Training ToolkitThe
May not be reproduced without permission:AI productivity tools " FLUX.1 model training requires Nvidia GPU support with at least 24GB of video memory
English 
简体中文 
日本語 
Deutsch 
Português do Brasil