How to Use
July 1, 2026 ยท View on GitHub
MiniT2I uses a MiniT2I diffusion transformer and google/flan-t5-large as the text encoder.
Download weights
- Download MiniT2I diffusion model
- safetensors: https://huggingface.co/MiniT2I/MiniT2I/tree/main/minit2i-b-16/transformer (
diffusion_pytorch_model.safetensors)
- safetensors: https://huggingface.co/MiniT2I/MiniT2I/tree/main/minit2i-b-16/transformer (
- Download flan-t5-large text encoder
- safetensors: https://huggingface.co/google/flan-t5-large/tree/main (
model.safetensors)
- safetensors: https://huggingface.co/google/flan-t5-large/tree/main (
Examples
Mac Metal
./bin/sd-cli \
--backend metal \
--diffusion-model ../models/minit2i/diffusion_pytorch_model.safetensors \
--t5xxl ../models/flan-t5-large/model.safetensors \
--prompt "a cat" \
--steps 100 \
--cfg-scale 6 \
--width 512 \
--height 512 \
--seed 42 \
--sampling-method euler \
--rng cpu \
--output minit2i_metal.png \
--threads 8
CUDA with diffusion flash attention
./bin/sd-cli \
--diffusion-model ../models/minit2i/diffusion_pytorch_model.safetensors \
--t5xxl ../models/flan-t5-large/model.safetensors \
--prompt "a cat" \
--steps 100 \
--cfg-scale 6 \
--width 512 \
--height 512 \
--seed 42 \
--sampling-method euler \
--diffusion-fa \
--output minit2i_cuda.png