license: apache-2.0
datasets:
- AIDC-AI/Ovis-dataset
library_name: transformers
tags:
- MLLM
pipeline_tag: image-text-to-text
Introduction
Ovis is a novel Multimodal Large Language Model (MLLM) architecture, designed to structurally align visual and textual embeddings. For a comprehensive introduction, please refer to Ovis paper and Ovis GitHub.
Model
As always, Ovis1.5 remains fully open-source: we release the training datasets, training & inference codes, and model weights for reproducible transparency and community collaboration.
| MiniCPM-Llama3-V2.5 | Ovis1.5-Llama3-8B | |
|---|---|---|
| Training scripts | - | Github |
| ViT | Siglip-400M | Siglip-400M |
| LLM | Llama3-8B-Instruct | Llama3-8B-Instruct |
| MMTBench-VAL | 57.6 | 60.7 |
| MMBench-EN-V1.1 | 74 | 78.2 |
| MMBench-CN-V1.1 | 70.1 | 75.2 |
| MMStar | 51.8 | 57.2 |
| MMMU-Val | 45.8 | 48.6 |
| MathVista-Mini | 54.3 | 62.4 |
| HallusionBenchAvg | 42.4 | 44.5 |
| AI2D | 78.4 | 82.5 |
| OCRBench | 725 | 743 |
| MMVet | 52.8 | 52.2 |
| RealWorldQA | 63.5 | 64.6 |
Usage
Below is a code snippet to run Ovis with multimodal inputs. For additional usage instructions, including inference wrapper and Gradio UI, please refer to Ovis GitHub.
pip install torch==2.1.0 transformers==4.42.4 deepspeed==0.14.0 pillow==10.3.0
import torch
from PIL import Image
from transformers import AutoModelForCausalLM
# load model
model = AutoModelForCausalLM.from_pretrained("AIDC-AI/Ovis1.5-Llama3-8B",
torch_dtype=torch.bfloat16,
multimodal_max_length=8192,
trust_remote_code=True).cuda()
text_tokenizer = model.get_text_tokenizer()
visual_tokenizer = model.get_visual_tokenizer()
conversation_formatter = model.get_conversation_formatter()
# enter image path and prompt
image_path = input("Enter image path: ")
image = Image.open(image_path)
text = input("Enter prompt: ")
query = f'<image>\n{text}'
prompt, input_ids = conversation_formatter.format_query(query)
input_ids = torch.unsqueeze(input_ids, dim=0).to(device=model.device)
attention_mask = torch.ne(input_ids, text_tokenizer.pad_token_id).to(device=model.device)
pixel_values = [visual_tokenizer.preprocess_image(image).to(
dtype=visual_tokenizer.dtype, device=visual_tokenizer.device)]
# generate output
with torch.inference_mode():
gen_kwargs = dict(
max_new_tokens=1024,
do_sample=False,
top_p=None,
top_k=None,
temperature=None,
repetition_penalty=None,
eos_token_id=model.generation_config.eos_token_id,
pad_token_id=text_tokenizer.pad_token_id,
use_cache=True
)
output_ids = model.generate(input_ids, pixel_values=pixel_values, attention_mask=attention_mask, **gen_kwargs)[0]
output = text_tokenizer.decode(output_ids, skip_special_tokens=True)
print(f'Output: {output}')
Citation
If you find Ovis useful, please cite the paper
@article{lu2024ovis,
title={Ovis: Structural Embedding Alignment for Multimodal Large Language Model},
author={Shiyin Lu and Yang Li and Qing-Guo Chen and Zhao Xu and Weihua Luo and Kaifu Zhang and Han-Jia Ye},
year={2024},
journal={arXiv:2405.20797}
}
License
The project is licensed under the Apache 2.0 License and is restricted to uses that comply with the license agreements of Qwen, Llama3, Clip, and Siglip.