---
title: instructblip-vicuna-13b
canonical_url: "https://www.modelscope.cn/models/AI-ModelScope/instructblip-vicuna-13b"
md_url: "https://www.modelscope.cn/models/AI-ModelScope/instructblip-vicuna-13b.md"
repository: AI-ModelScope/instructblip-vicuna-13b
last_updated: 2023-10-17
license: other
model_type:
  - instructblip
architectures:
  - InstructBlipForConditionalGeneration
library_name:
  - pytorch
  - transformer
frameworks:
  - pytorch
language:
  - en
downloads: 299
stars: 1
tags:
  - vision
  - image-captioning
---

# instructblip-vicuna-13b

> instructblip-vicuna-13b - AI-ModelScope 在 ModelScope 开源的模型。InstructBLIP model using Vicuna-13b as language model. InstructBLIP was introduced in the paper InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning by Dai et al.

AI-ModelScope/instructblip-vicuna-13b 是 ModelScope 魔搭社区上的机器学习模型，采用 other 许可。

- **Repository**: AI-ModelScope/instructblip-vicuna-13b
- **License**: other
- **Tags**: vision, image-captioning
- **Downloads**: 299
- **Stars**: 1
- **Last updated**: 2023-10-17

Source: https://www.modelscope.cn/models/AI-ModelScope/instructblip-vicuna-13b

---

# InstructBLIP model

InstructBLIP model using [Vicuna-13b](https://github.com/lm-sys/FastChat#model-weights) as language model. InstructBLIP was introduced in the paper [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://arxiv.org/abs/2305.06500) by Dai et al.

Disclaimer: The team releasing InstructBLIP did not write a model card for this model so this model card has been written by the Hugging Face team.

## Model description

InstructBLIP is a visual instruction tuned version of [BLIP-2](https://huggingface.co/docs/transformers/main/model_doc/blip-2). Refer to the paper for details.

![InstructBLIP architecture](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/instructblip_architecture.jpg)

## Intended uses & limitations

Usage is as follows:

```
from transformers import InstructBlipProcessor, InstructBlipForConditionalGeneration
import torch
from PIL import Image
import requests
from modelscope import snapshot_download

model_dir = snapshot_download("AI-ModelScope/instructblip-vicuna-13b")
model = InstructBlipForConditionalGeneration.from_pretrained(model_dir)
processor = InstructBlipProcessor.from_pretrained(model_dir)

device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)

url = "https://modelscope-open.oss-cn-hangzhou.aliyuncs.com/resources/aiyinsitan.jpg"
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")
prompt = "What is unusual about this image?"
inputs = processor(images=image, text=prompt, return_tensors="pt").to(device)

outputs = model.generate(
        **inputs,
        do_sample=False,
        num_beams=5,
        max_length=256,
        min_length=1,
        top_p=0.9,
        repetition_penalty=1.5,
        length_penalty=1.0,
        temperature=1,
)
generated_text = processor.batch_decode(outputs, skip_special_tokens=True)[0].strip()
print(generated_text)
```

### How to use

For code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/main/en/model_doc/instructblip).