Instructions to use moot20/paligemma2-28b-mix-224-MLX-8bits with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moot20/paligemma2-28b-mix-224-MLX-8bits with Transformers:

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

pipe = pipeline("image-text-to-text", model="moot20/paligemma2-28b-mix-224-MLX-8bits")

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("moot20/paligemma2-28b-mix-224-MLX-8bits")
model = AutoModelForMultimodalLM.from_pretrained("moot20/paligemma2-28b-mix-224-MLX-8bits", device_map="auto")

MLX

How to use moot20/paligemma2-28b-mix-224-MLX-8bits with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("moot20/paligemma2-28b-mix-224-MLX-8bits")
config = load_config("moot20/paligemma2-28b-mix-224-MLX-8bits")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

vLLM

How to use moot20/paligemma2-28b-mix-224-MLX-8bits with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moot20/paligemma2-28b-mix-224-MLX-8bits"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moot20/paligemma2-28b-mix-224-MLX-8bits",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/moot20/paligemma2-28b-mix-224-MLX-8bits

SGLang

How to use moot20/paligemma2-28b-mix-224-MLX-8bits 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 "moot20/paligemma2-28b-mix-224-MLX-8bits" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moot20/paligemma2-28b-mix-224-MLX-8bits",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "moot20/paligemma2-28b-mix-224-MLX-8bits" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moot20/paligemma2-28b-mix-224-MLX-8bits",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use moot20/paligemma2-28b-mix-224-MLX-8bits with Docker Model Runner:
```
docker model run hf.co/moot20/paligemma2-28b-mix-224-MLX-8bits
```
Atomic Chat

moot20/paligemma2-28b-mix-224-MLX-8bits

This model was converted to MLX format from google/paligemma2-28b-mix-224 using mlx-vlm version 0.1.13. Refer to the original model card for more details on the model.

Use with mlx

pip install -U mlx-vlm

python -m mlx_vlm.generate --model moot20/paligemma2-28b-mix-224-MLX-8bits --max-tokens 100 --temp 0.0 --prompt "Describe this image." --image <path_to_image>