Instructions to use modilify/Modilify-Mk2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use modilify/Modilify-Mk2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="modilify/Modilify-Mk2", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("modilify/Modilify-Mk2", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use modilify/Modilify-Mk2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "modilify/Modilify-Mk2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "modilify/Modilify-Mk2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/modilify/Modilify-Mk2

SGLang

How to use modilify/Modilify-Mk2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "modilify/Modilify-Mk2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "modilify/Modilify-Mk2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "modilify/Modilify-Mk2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "modilify/Modilify-Mk2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use modilify/Modilify-Mk2 with Docker Model Runner:
```
docker model run hf.co/modilify/Modilify-Mk2
```

Modilify-Mk2

File size: 935 Bytes

4bbff32

# Attribution

Modilify Mk2 uses the text backbone of Google DeepMind's [DiffusionGemma 26B-A4B-it](https://huggingface.co/google/diffusiongemma-26B-A4B-it), whose model card identifies its license as Apache 2.0. The vision tower is not included in this text-only release.

The Modilify architecture, rolling-canvas generation, trajectory memory, and trained dense/expert adapters modify the upstream model. This release is the saved step 4400 checkpoint, converted to a self-contained Transformers model. Modilify's contributions use the [Modilify Non-Commercial Model License 1.0](LICENSE). This does not replace upstream rights and terms.

The code integrates with Hugging Face Transformers (Copyright 2018–, The Hugging Face team; Apache 2.0) and uses PyTorch and Safetensors as installed dependencies. Those libraries retain their own licenses. The Apache license text is reproduced in [LICENSE.apache-2.0](LICENSE.apache-2.0).