Model CoverageDiffusion

Qwen-Image

View as Markdown

Qwen-Image is Alibaba Cloud’s text-to-image diffusion transformer. NeMo AutoModel supports Qwen-Image training via its flow-matching pipeline with a dedicated qwen_image adapter, enabling FSDP2 parallelization, multiresolution bucketed dataloading and LoRA-style fine-tuning.

TaskText-to-Image
ArchitectureDiT (Flow Matching)
HF OrgQwen

Available Models

  • Qwen-Image

Task

  • Text-to-Image (T2I)

Example HF Models

ModelHF ID
Qwen-ImageQwen/Qwen-Image

Example Recipes

RecipeDescription
qwen_image_t2i_flow.yamlFine-tune — Qwen-Image with flow matching
qwen_image_t2i_flow.yamlPretrain — Qwen-Image with flow matching

Try with NeMo AutoModel

1. Clone and install from source (full instructions):

$git clone https://github.com/NVIDIA-NeMo/Automodel.git
$cd Automodel
$uv sync --locked --all-groups --all-extras --extra diffusion-media

2. Run the recipe from inside the repo:

$uv run torchrun --nproc-per-node=8 \
>  examples/diffusion/finetune/finetune.py \
>  -c examples/diffusion/finetune/qwen_image_t2i_flow.yaml

1. Pull the container and mount a checkpoint directory:

$docker run --gpus all -it --rm \
>  --shm-size=8g \
>  -v $(pwd)/checkpoints:/opt/Automodel/checkpoints \
>  nvcr.io/nvidia/nemo-automodel:26.06.00

2. Navigate to the AutoModel directory (where the recipes are):

$cd /opt/Automodel

3. Install media dependencies (not installed in the container by default):

$uv pip install ".[diffusion-media]"

4. Run the recipe:

$torchrun --nproc-per-node=8 \
>  examples/diffusion/finetune/finetune.py \
>  -c examples/diffusion/finetune/qwen_image_t2i_flow.yaml

See the Installation Guide and Diffusion Training and Fine-Tuning Guide.

Fine-Tuning

See the Diffusion Training and Fine-Tuning Guide.

Hugging Face Model Cards