Krea 2 Raw Training Slowdown on RTX 5080: Potential AI Toolkit Bug

NexusFred · reddit · 2026-08-23

A user reported severe performance issues when training Krea 2 Raw LoRA on an RTX 5080 16GB. Despite enabling Low VRAM and gradient checkpointing, speed degraded from 21s/iter to 40s/iter with near-max VRAM usage. In contrast, the same model trains much faster on OneTrainer, suggesting a potential bug related to AI Toolkit issue #990.

Original post →

More from coding & agent

coding & agent channel →