LoraHub
What it is
LoraHub is a self-hosted LoRA (Low-Rank Adaptation) training workbench that unifies multiple training backends (like kohya-ss, diffusion-pipe, and ai-toolkit) into a single platform. It provides a unified Web UI, CLI, and TUI for managing training tasks, dataset processing, and model downloading.
The problem it solves
Fragmented training tools and complex configuration files for different LoRA backends, requiring manual handling of datasets and multi-GPU setups.
What you could use it for
- Fine-tuning Stable Diffusion or Flux models using multiple backends from a single interface.
- Managing large-scale LoRA training jobs with automated scheduling and multi-GPU support.
- Automated dataset auditing, WD14/JoyTag labeling, and image workbenching for AI model training.
- Distributed training across multiple GPUs with support for DeepSpeed and FSDP.
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.