mPLUG-Owl🦉: Modularization Empowers Large Language Models with Multimodality

Qinghao Ye*, Haiyang Xu*, Guohai Xu*, Jiabo Ye, Ming Yan†, Yiyang Zhou, Junyang Wang, Anwen Hu, Pengcheng Shi, Yaya Shi, Chaoya Jiang, Chenliang Li, Yuanhong Xu, Hehong Chen, Junfeng Tian, Qian Qi, Ji Zhang, Fei Huang

阿里巴巴集团，达摩院

*Equal Contribution; † Corresponding Author

English | 简体中文

示例

亮点特色

一种面向多模态语言模型的模块化的训练范式。
能学习与语言空间相适应的视觉知识，并支持在多模态场景(支持图片、视频、文本输入)下进行多轮对话。
涌现多图关系理解，场景文本理解和基于视觉的文档理解等能力。
提出了针对视觉相关指令的测评集OwlEval，用以评估多模态语言模型的对带有视觉信息上下文的理解能力。
我们在模块化上的一些探索:
- E2E-VLP, mPLUG 和 mPLUG-2, 分别被ACL 2021, EMNLP 2022 and ICML 2023接收。
- mPLUG 首次在VQA Challenge上超越人类。
即将发布
- 多语言支持。
- 在多图片/视频数据上训练的模型
- 在HuggingFace Hub上发布。
- Huggingface 在线Demo
- 指令微调代码。
- 视觉相关指令的测评集OwlEval

与v0版本的兼容性

我们在main分支上将代码用huggingface的使用风格进行了完整重构，同时对模型中的一些错误进行了修复并重新训练，因此新的checkpoint和旧的代码是不兼容的。我们已经将旧的代码移到v0分支，你可以切换到v0分支以使用先前的checkpoints。

在线Demo

ModelScope

Hugging Face

模型权重 Huggingface Model Hub

Model	Phase	Download link
mPLUG-Owl 7B	Pre-training	下载链接
mPLUG-Owl 7B	Instruction tuning (LoRA)	下载链接
mPLUG-Owl 7B	Instruction tuning (FT)	下载链接
mPLUG-Owl 7B (多语言版)	Instruction tuning (LoRA)	Download link

OwlEval

我们所使用的评测集放在 ./OwlEval 中。

使用

安装依赖

创建conda环境

conda create -n mplug_owl python=3.10
conda activate mplug_owl

安装PyTorch

conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c nvidia

安装其它依赖

pip install -r requirements.txt

本地部署Demo

我们提供了一个易扩展的脚本来一键部署本地Demo，你可以根据自己的需求进行修改。

python -m serve.web_server --base-model 'your checkpoint directory' --bf16

推理

如果要实现自定义的推理，可以参考以下步骤。

构建model, tokenizer, processor

from pipeline.interface import get_model
model, tokenizer, processor = get_model(pretrained_ckpt='your checkpoint directory', use_bf16='use bf16 or not')

准备模型输入

# We use a human/AI template to organize the context as a multi-turn conversation.
# <image> denotes an image placehold.
prompts = [
'''The following is a conversation between a curious human and AI assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.
Human: <image>
Human: Explain why this meme is funny.
AI: ''']

# The image paths should be placed in the image_list and kept in the same order as in the prompts.
# We support urls, local file paths and base64 string. You can custom the pre-process of images by modifying the mplug_owl.modeling_mplug_owl.ImageProcessor
image_list = ['https://xxx.com/image.jpg',]

对于多张图片，由于这是模型涌现的能力（训练数据并不包含多图的样本），我们并不知道何种格式对模型而言是最佳的。下面是我们在实验中所使用的格式。大家也可以自行探索能让模型更好理解多图关系的格式或者prompt。

prompts = [
'''The following is a conversation between a curious human and AI assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.
Human: <image>
Human: <image>
Human: Do the shirts worn by the individuals in the first and second pictures vary in color? If so, what is the specific color of each shirt?
AI: ''']

image_list = ['https://xxx.com/image_1.jpg', 'https://xxx.com/image_2.jpg']

获取模型回复

# generate kwargs (the same in transformers) can be passed in the do_generate()
from pipeline.interface import do_generate
sentence = do_generate(prompts, image_list, model, tokenizer, processor, 
                       use_bf16=True, max_length=512, top_k=5, do_sample=True)

指令微调

训练样本存放在xxx.jsonl 中，格式如下:

{"text": "The following is a conversation between a curious human and AI assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.\nHuman: Construct a sentence using the given verb in the past tense\nshudder\nAI: She shuddered at the thought of being alone in the dark.", "task_type": "gpt4instruct_sft"}
{"image": ["xxx.jpg"], "text": "The following is a conversation between a curious human and AI assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.\nHuman: <image>\nHuman: Where is the man located in the image?\nAI: The man is located in a small restaurant, sitting and enjoying his meal.\nHuman: What type of food is the man eating in the image?\nAI: The man is eating a burrito in the image.\nHuman: What can be seen next to the man in the restaurant?\nAI: Next to the man in the restaurant, there is a fridge filled with cans of soda.\nHuman: How does the man feel as he is about to eat his meal?\nAI: The man looks happy and is smiling as he is about to take a bite of his meal.\nHuman: What unique aspect of the photo makes the scene appear distorted or unusual?\nAI: The unique aspect of this photo is that it was taken using a fish-eye lens, which creates a distorted, wide-angle view of the scene, making it appear different from a standard perspective.", "task_type": "llava_sft"}

其中 task_type 可以为 {'quora_chat_sft', 'sharegpt_chat_sft', 'llava_sft', 'gpt4instruct_sft'}。

准备好 train.jsonl 和 dev.jsonl 并修改 configs/v0.yaml 中的 data_files.

执行训练脚本。

PYTHONPATH=./ bash train_it.sh # 如果你想训练整个LLM，替换为 train_it_wo_lora.sh

性能比较

我们展示了50个单轮对话（左）和52个多轮对话（右）在人工评估指标下，mPLUG-Owl和基线方法的比较结果。A/B/C/D表示评分人员对每个回复的评级。

引用

如果我们的工作对你有帮助，可以考虑给我们的仓库点个star & 引用我们的论文。

@misc{ye2023mplugowl,
      title={mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality}, 
      author={Qinghao Ye and Haiyang Xu and Guohai Xu and Jiabo Ye and Ming Yan and Yiyang Zhou and Junyang Wang and Anwen Hu and Pengcheng Shi and Yaya Shi and Chaoya Jiang and Chenliang Li and Yuanhong Xu and Hehong Chen and Junfeng Tian and Qian Qi and Ji Zhang and Fei Huang},
      year={2023},
      eprint={2304.14178},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

README_zh.md

README_zh.md

mPLUG-Owl🦉: Modularization Empowers Large Language Models with Multimodality

示例

最新更新

亮点特色

与v0版本的兼容性

在线Demo

ModelScope

Hugging Face

模型权重 Huggingface Model Hub

OwlEval

使用

安装依赖

本地部署Demo

推理

指令微调

性能比较

相关项目

引用

Files

README_zh.md

Latest commit

History

README_zh.md

File metadata and controls

mPLUG-Owl🦉: Modularization Empowers Large Language Models with Multimodality

示例

最新更新

亮点特色

与v0版本的兼容性

在线Demo

ModelScope

Hugging Face

模型权重 Huggingface Model Hub

OwlEval

使用

安装依赖

本地部署Demo

推理

指令微调

性能比较

相关项目

引用