| MiniMax H3 is licensed under the MiniMax H3 Community License Agreement, Copyright © 2026 MiniMax. All Rights Reserved. | |
| This repository contains a modified MiniMax H3 Model Derivative: a four-step student LoRA trained using Audio:Video loss weights 2:1, 50% authored audio-description prompts, and teacher audio latent endpoint regression. Spectral supervision is disabled. Base-model weights are not included. | |