Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -22,12 +22,32 @@ configs:
|
|
22 |
path: "NER/test.json"
|
23 |
---
|
24 |
|
25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
```bibtex
|
27 |
@inproceedings{AdaptLLM,
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
}
|
33 |
-
```
|
|
|
22 |
path: "NER/test.json"
|
23 |
---
|
24 |
|
25 |
+
# Adapting Large Language Models via Reading Comprehension
|
26 |
+
|
27 |
+
This repo contains the evaluation datasets for our paper [Adapting Large Language Models via Reading Comprehension](https://arxiv.org/pdf/2309.09530.pdf)
|
28 |
+
|
29 |
+
We explore **continued pre-training on domain-specific corpora** for large language models. While this approach enriches LLMs with domain knowledge, it significantly hurts their prompting ability for question answering. Inspired by human learning via reading comprehension, we propose a simple method to **transform large-scale pre-training corpora into reading comprehension texts**, consistently improving prompting performance across tasks in **biomedicine, finance, and law domains**. Our 7B model competes with much larger domain-specific models like BloombergGPT-50B. Moreover, our domain-specific reading comprehension texts enhance model performance even on general benchmarks, indicating potential for developing a general LLM across more domains.
|
30 |
+
|
31 |
+
## GitHub repo:
|
32 |
+
https://github.com/microsoft/LMOps
|
33 |
+
|
34 |
+
## Domain-specific LLMs:
|
35 |
+
Our models of different domains are now available in Huggingface: [Biomedicine-LLM](https://huggingface.co/AdaptLLM/medicine-LLM), [Finance-LLM](https://huggingface.co/AdaptLLM/finance-LLM) and [Law-LLM](https://huggingface.co/AdaptLLM/law-LLM), the performances of our AdaptLLM compared to other domain-specific LLMs are:
|
36 |
+
|
37 |
+
<p align='center'>
|
38 |
+
<img src="./comparison.png" width="700">
|
39 |
+
</p>
|
40 |
+
|
41 |
+
## Domain-specific Tasks:
|
42 |
+
To easily reproduce our results, we have uploaded the filled-in zero/few-shot input instructions and output completions of each domain-specific task: [biomedicine-tasks](https://huggingface.co/datasets/AdaptLLM/medicine-tasks), [finance-tasks](https://huggingface.co/datasets/AdaptLLM/finance-tasks), and [law-tasks](https://huggingface.co/datasets/AdaptLLM/law-tasks).
|
43 |
+
|
44 |
+
|
45 |
+
## Citation:
|
46 |
```bibtex
|
47 |
@inproceedings{AdaptLLM,
|
48 |
+
title={Adapting Large Language Models via Reading Comprehension},
|
49 |
+
author={Daixuan Cheng and Shaohan Huang and Furu Wei},
|
50 |
+
url={https://arxiv.org/abs/2309.09530},
|
51 |
+
year={2023},
|
52 |
}
|
53 |
+
```
|