Cappy Large

🚀 Cappy-Large

Cappy是一个预训练的小型评分器，旨在提升多任务大语言模型（LLM）的性能和效率。它能根据指令对候选回复进行评分，以此评估回复的正确性。仅3.6亿参数的Cappy，既可以独立完成分类任务，也能作为辅助组件提升LLM的性能。此外，它还能在不微调LLM或访问其参数的情况下，高效整合下游监督信号。同时，Cappy还能与其他LLM适配方法（如微调、上下文学习和提示调优等）灵活协作，进一步提升性能。

🚀 快速开始

• 代码仓库：https://github.com/tanyuqian/cappy
• 论文链接：arxiv.org/abs/2311.06720

Cappy是一个预训练的小型评分器，旨在提升多任务大语言模型（LLM）的性能和效率。它接收指令和候选回复作为输入，输出一个0到1之间的分数，用于表示回复相对于指令的预估正确性。仅3.6亿参数的Cappy，既可以独立用于分类任务，也可以作为大语言模型（LLM）的辅助组件，提升其性能。此外，Cappy能够在无需微调LLM或访问其参数的情况下，高效整合下游监督信号。而且，Cappy可以灵活地与其他LLM适配方法（包括微调、上下文学习和提示调优等）协作，进一步提升性能。

💻 使用示例

基础用法

Cappy可以加载为Jax/Flax模型或PyTorch模型。

Jax/Flax

from transformers import AutoTokenizer, FlaxAutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained('btan2/cappy-large')
cappy = FlaxAutoModelForSequenceClassification.from_pretrained('btan2/cappy-large')

instruction = """
What label best describes this news article?
Carlyle Looks Toward Commercial Aerospace (Reuters) Reuters - Private investment firm Carlyle Group,\which has a reputation for making well-timed and occasionally\controversial plays in the defense industry, has quietly placed\its bets on another part of the market.
"""
response = 'Business'

inputs = tokenizer([(instruction, response), ], return_tensors='pt')
score = cappy(**inputs).logits[0][0].item()

PyTorch

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained('btan2/cappy-large')
cappy = AutoModelForSequenceClassification.from_pretrained('btan2/cappy-large')

instruction = """
What label best describes this news article?
Carlyle Looks Toward Commercial Aerospace (Reuters) Reuters - Private investment firm Carlyle Group,\which has a reputation for making well-timed and occasionally\controversial plays in the defense industry, has quietly placed\its bets on another part of the market.
"""
response = 'Business'

inputs = tokenizer([(instruction, response), ], return_tensors='pt')
score = cappy(**inputs).logits[0][0].item()

📚 详细文档

评估

我们通过一系列与预训练任务不同的保留任务对Cappy进行了验证。整体性能如图1和图2所示。具体而言，在从PromptSource选取的11个语言理解任务中，仅有3.6亿参数的Cappy显著优于OPT - IML - 30B和OPT - 175B，并达到了此前多任务大语言模型中的最佳水平。此外，在来自BIG - Bench的45个不同复杂任务中，Cappy大幅提升了先进多任务大语言模型FLAN - T5的性能。而且，当与微调或上下文学习结合使用时，Cappy还能进一步提升性能。后续的消融实验证明了我们提出的预训练和数据增强策略的重要性。

软件依赖

Cappy的预训练使用了Red Coast（一个用于自动化分布式训练的轻量级工具包）中此示例的代码。

引用信息

@inproceedings{
tan2023cappy,
title={Cappy: Outperforming and Boosting Large Multi-Task {LM}s with a Small Scorer},
author={Bowen Tan and Yun Zhu and Lijuan Liu and Eric Xing and Zhiting Hu and Jindong Chen},
booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
year={2023},
url={https://openreview.net/forum?id=Srt1hhQgqa}
}

📄 许可证

本项目采用Apache-2.0许可证。