🔥Minimax H3
A MiniMax H3 LoRA trained primarily on high quality Rule34 / 3DCG adult content.
The main focus of this LoRA is the visual style, but strong emphasis was also put on animation quality, natural poses, and fluid movement.
Works with both T2VA and I2VA; for best results in pure Text-to-Video (T2VA), it is highly recommended to stack this LoRA with other NSFW / anatomy-focused LoRAs.
This LoRA alone does not have perfect anatomical knowledge.
Training details
100 clips
20,000 steps
All training data was in landscape orientation → generations work best in landscape
Recommended strength
0.4 – 0.9
Because of the training volume, the optimal strength can vary depending on your prompt and other LoRAs.
Start around 0.6–0.9 and adjust until you find the sweet spot for your specific use case.
Trigger
zy_3dcg