This model was trained for 2400 steps with 8 batches, and this LoRA is the final version.
luwanq_000002400.safetensors