Instructions to use aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain with Transformers:

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

pipe = pipeline("image-text-to-text", model="aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain")

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain

SGLang

How to use aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain 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 "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain" \
    --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": "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain",
		"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 "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain" \
        --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": "aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain with Docker Model Runner:
```
docker model run hf.co/aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain
```

Model Card: LLaVA_MORE-llama_3_1-8B-pretrain

LLaVA-MORE enhances the well-known LLaVA architecture by integrating the use of LLaMA 3.1 as the language model. We are publicly releasing the checkpoints for stages one and two for the first model with 8B parameters.

In this model space, you will find the stage one (pretrain) weights of LLaVA-MORE LLaMA 3.1 8B.

For more information, visit our LLaVA-MORE repository.

Inference

You can try our LLaVA-MORE in the Image-To-Text task by cloning our repository and running the following script.

python -u llava/eval/run_llava.py

Citation

If you make use of our work, please cite our repo:

@article{cocchi2025llava,
      title={{LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning}},
      author={Cocchi, Federico and Moratelli, Nicholas and Caffagni, Davide and Sarto, Sara and Baraldi, Lorenzo and Cornia, Marcella and Cucchiara, Rita},
      journal={arXiv preprint arXiv:2503.15621},
      year={2025}
}

Downloads last month: 8

Inference Providers NEW

Image-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Dataset used to train aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain

Collection including aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain

LLaVA-MORE

Collection

LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning • 12 items • Updated Jul 20, 2025 • 2

Paper for aimagelab/LLaVA_MORE-llama_3_1-8B-pretrain

LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction Tuning

Paper • 2503.15621 • Published Mar 19, 2025