Math-LLaVA: Bootstrapping Mathematical Reasoning for Multimodal Large Language Models
Paper • 2406.17294 • Published • 11
How to use from
SGLangUse 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 "Zhiqiang007/Math-LLaVA" \
--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": "Zhiqiang007/Math-LLaVA",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'Quick Links
Math-LLaVA-13B Model Card
Model details
Model type: Math-LLaVA is an open-source MLLM by fine-tuning LLaVA-1.5-13B on selected and GPT4-Vision-assisted synthesized MathV360K data.
Model date: Math-LLaVA-13B was trained in June 2024.
Paper or resources for more information: [Paper] [Code]
License
Llama 2 is licensed under the LLAMA 2 Community License, Copyright (c) Meta Platforms, Inc. All Rights Reserved.
Intended use
Primary intended uses: The primary use of Math-LLaVA is research on multimodal large language models, multimodal reasoning and question answering.
Primary intended users: The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.
Training dataset
- MathV360K instruction-tuning data
Evaluation dataset
A collection of 3 benchmarks, including 2 multimodal mathematical reasoning benchmarks and 1 benchmark for multi-discipline multimodal reasoning.
- Downloads last month
- 43
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Install from pip and serve model
# Install SGLang from pip: pip install sglang# Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Zhiqiang007/Math-LLaVA" \ --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": "Zhiqiang007/Math-LLaVA", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'