AI Toolkit MiniMax H3 LoRA Training: T2V and I2V

ComfyUI Wikinews

Train MiniMax H3 LoRAs with Ostris AI Toolkit: T2V and I2V support, NVFP4 quantization and VAE gradient checkpointing for consumer GPUs.

Ostris AI Toolkit added support for MiniMax H3 text-to-video and image-to-video training on August 3, letting you fine-tune LoRAs for the open-weights H3 model on consumer-grade GPUs.

The support landed in commit 8502a84 ("Add support for MiniMax H3 T2V and I2V training"), which ships a new minimax_h3 extension with the full training pipeline: packed-token transformer, video/audio VAE, text encoder, and NVFP4 quantization utilities.

What's included

  • T2V and I2V LoRA training for both H3 checkpoints (FL2VA and Ref2VA)
  • NVFP4 quantization support (nvfp4_quant.py) for reduced VRAM during training
  • VAE gradient checkpointing (commit 602306d) to lower memory usage
  • Joint audio-video handling — the H3 pipeline denoises video and stereo audio latents together, and the toolkit preserves that during training
  • Image support in the data pipeline (commit 88ac27f) for I2V training pairs
  • Faster quantization processing (commit d3a3f70) tuned for H3's layer structure

Model support

MiniMax H3 is now listed in AI Toolkit's supported models as MiniMaxAI/MiniMax-H3. H3 is the 33.1B omni-modal transformer behind MiniMax's Hailuo line, open-sourced on August 3 with native ComfyUI support (read the open-weights story).

Getting started

AI Toolkit runs as a GUI or CLI on consumer hardware. Install it from the ostris/ai-toolkit repo, then create a training config pointing at the H3 base model. The minimax_h3 extension is built in — no separate install needed.

Comments

Sign in with GitHub to join the discussion.

Loading comments…
AI Toolkit MiniMax H3 LoRA Training: T2V and I2V | ComfyUI Wiki