README.md

---
language:
- en
- zh
license: apache-2.0
tags:
- vision
- image-text-to-text
datasets:
- lmms-lab/LLaVA-OneVision-Data
pipeline_tag: image-text-to-text
inference: false
arxiv: 2408.03326
---
# LLaVA-Onevision Model Card

![image/png](llava_onevision_arch.png)

Below is the model card of 72B LLaVA-Onevision model which is copied from the original LLaVA-Onevision model card that you can find [here](https://huggingface.co/lmms-lab/llava-onevision-qwen2-72b-si).



## Model details

**Model type:**
LLaVA-Onevision is an open-source multimodal LLM trained by fine-tuning Qwen2 on GPT-generated multimodal instruction-following data.
LLaVA-OneVision is the first single model that can simultaneously push the performance boundaries of open LMMs in three important computer
vision scenarios: single-image, multi-image, and video scenarios. Importantly, the design of LLaVA-OneVision allows strong transfer learning
across different modalities/scenarios, yielding new emerging capabilities. In particular, strong video understanding and cross-scenario
capabilities are demonstrated through task transfer from images to videos.

**Model date:**
LLaVA-Onevision-72b-si was added in August 2024.

**Paper or resources for more information:**
https://llava-vl.github.io/

- **Architecture:** SO400M + Qwen2
- **Pretraining Stage:** LCS-558K, 1 epoch, projector
- **Mid Stage:** A mixture of 4.7M high-quality synthetic data, 1 epoch, full model
- **Final-Image Stage:** A mixture of 3.6M single-image data, 1 epoch, full model
- **OneVision Stage:** A mixture of 1.6M single-image/multi-image/video data, 1 epoch, full model
- **Precision:** bfloat16


## How to use the model

First, make sure to have `transformers` installed from source or `transformers >= 4.45.0`. 
The model supports multi-image and multi-prompt generation. Meaning that you can pass multiple images in your prompt. Make sure also to follow the correct prompt template by applyong chat template:

### Using `pipeline`:

Below we used [`"llava-hf/llava-onevision-qwen2-72b-si-hf"`](https://huggingface.co/llava-hf/llava-onevision-qwen2-72b-si-hf) checkpoint.

```python
from transformers import pipeline
from PIL import Image    
import requests

model_id = "llava-hf/llava-onevision-qwen2-72b-si-hf"
pipe = pipeline("image-to-text", model=model_id)
url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/ai2d-demo.jpg"
image = Image.open(requests.get(url, stream=True).raw)

# Define a chat history and use `apply_chat_template` to get correctly formatted prompt
# Each value in "content" has to be a list of dicts with types ("text", "image") 
conversation = [
    {

      "role": "user",
      "content": [
          {"type": "text", "text": "What does the label 15 represent? (1) lava (2) core (3) tunnel (4) ash cloud"},
          {"type": "image"},
        ],
    },
]
prompt = processor.apply_chat_template(conversation, add_generation_prompt=True)

outputs = pipe(image, prompt=prompt, generate_kwargs={"max_new_tokens": 200})
print(outputs)
>>> {"generated_text": "user\n\nWhat does the label 15 represent? (1) lava (2) core (3) tunnel (4) ash cloud\nassistant\nLava"}
```

### Using pure `transformers`:

Below is an example script to run generation in `float16` precision on a GPU device:

```python
import requests
from PIL import Image

import torch
from transformers import AutoProcessor, LlavaNextForConditionalGeneration

model_id = "llava-hf/llava-onevision-qwen2-72b-si-hf"
model = LlavaNextForConditionalGeneration.from_pretrained(
    model_id, 
    torch_dtype=torch.float16, 
    low_cpu_mem_usage=True, 
).to(0)

processor = AutoProcessor.from_pretrained(model_id)

# Define a chat history and use `apply_chat_template` to get correctly formatted prompt
# Each value in "content" has to be a list of dicts with types ("text", "image") 
conversation = [
    {

      "role": "user",
      "content": [
          {"type": "text", "text": "What are these?"},
          {"type": "image"},
        ],
    },
]
prompt = processor.apply_chat_template(conversation, add_generation_prompt=True)

image_file = "http://images.cocodataset.org/val2017/000000039769.jpg"
raw_image = Image.open(requests.get(image_file, stream=True).raw)
inputs = processor(prompt, raw_image, return_tensors='pt').to(0, torch.float16)

output = model.generate(**inputs, max_new_tokens=200, do_sample=False)
print(processor.decode(output[0][2:], skip_special_tokens=True))
```

### Model optimization

#### 4-bit quantization through `bitsandbytes` library

First make sure to install `bitsandbytes`, `pip install bitsandbytes` and make sure to have access to a CUDA compatible GPU device. Simply change the snippet above with: 

```diff
model = LlavaNextForConditionalGeneration.from_pretrained(
    model_id, 
    torch_dtype=torch.float16, 
    low_cpu_mem_usage=True,
+   load_in_4bit=True
)
```

#### Use Flash-Attention 2 to further speed-up generation

First make sure to install `flash-attn`. Refer to the [original repository of Flash Attention](https://github.com/Dao-AILab/flash-attention) regarding that package installation. Simply change the snippet above with: 

```diff
model = LlavaNextForConditionalGeneration.from_pretrained(
    model_id, 
    torch_dtype=torch.float16, 
    low_cpu_mem_usage=True,
+   use_flash_attention_2=True
).to(0)
```

# Citation
```
@misc{li2024llavaonevisioneasyvisualtask,
      title={LLaVA-OneVision: Easy Visual Task Transfer}, 
      author={Bo Li and Yuanhan Zhang and Dong Guo and Renrui Zhang and Feng Li and Hao Zhang and Kaichen Zhang and Yanwei Li and Ziwei Liu and Chunyuan Li},
      year={2024},
      eprint={2408.03326},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2408.03326}, 
}
```