Last released May 22, 2026
Blazing-fast LLM fine-tuning with minimal VRAM — multi-GPU, manual LoRA gradients, flash attention, 4-bit quant, web dashboard
Supported by
