Simple One-Page LoRA Trainer for Anima (Portable, 6GB VRAM, Optimized Config)
#153
by ThetaCursed - opened
Most LoRA training tools are overloaded with tabs and settings. For beginners, this complexity is a massive barrier to entry. For experienced users, forgetting a single checkbox buried deep in the settings often means wasting hours of GPU time.
Anima TrainFlow ends this. It’s a zero-tab interface that brings all essential controls onto a single page, designed to be simple, intuitive, and focused.
GitHub Repository: https://github.com/ThetaCursed/Anima-TrainFlow
Reddit Discussion: https://www.reddit.com/r/StableDiffusion/comments/1tcxhoq/anima_trainflow_simple_onepage_lora_trainer_for/
(I’ve shared more details about the trainer there. You can check out the comments, where I've already answered some technical questions)
Why use it?
- Zero-Tab UI: Everything you need on one screen.
- Truly Portable: Pre-configured environment - just extract and run.
- Low VRAM Friendly: Optimized for 6GB+ NVIDIA GPUs.
- Live Previews: Built-in gallery that updates in real-time as samples are generated.
- Smart Dataset Analyzer: Auto-calculates optimal resolution and buckets.
- Prodigy Native: Pre-configured for intelligent learning rate handling.