<div align="center">
<img src="https://z3.ax1x.com/2021/11/11/IwED0K.png" width="350px">
**An Open-Source Framework for Prompt-learning.**
------
<p align="center">
<a href="#Overview">Overview</a> •
<a href="#installation">Installation</a> •
<a href="#use-openprompt">How To Use</a> •
<a href="https://thunlp.github.io/OpenPrompt/">Docs</a> •
<a href="https://arxiv.org/abs/2111.01998">Paper</a> •
<a href="#citation">Citation</a> •
<a href="https://github.com/thunlp/OpenPrompt/tree/main/results/">Performance</a> •
</p>
</div>
## What's New?
- ❗️ April 2023: $\color{red}{\normalsize{\textbf{Want to build your Chat AI?}}}$ **We are releasing [UltraChat](https://github.com/thunlp/UltraChat), use OpenPrompt and UltraChat to conduct supervised instruction tuning, see** [`./tutorial/9_UltraChat.py`](./tutorial/9_UltraChat.py).
- Aug 2022: Thanks to contributor [zhiyongLiu1114](https://github.com/zhiyongLiu1114), OpenPrompt now supports [ERNIE 1.0](https://github.com/thunlp/OpenPrompt/tree/main/tutorial/7_ernie_paddlepaddle) in PaddlePaddle.
- July 2022: OpenPrompt supports OPT now.
- June 2022: OpenPrompt wins ACL 2022 Best Demo Paper Award.
- Mar 2022: We add a [tutorial](https://github.com/thunlp/OpenPrompt/blob/main/tutorial/6.1_chinese_dataset_uer_t5.py) as the response to [issue 124](https://github.com/thunlp/OpenPrompt/issues/124), which uses a customized tokenizer_wrapper to perform tasks that are not in the default configuration of OpenPrompt (e.g., Bert tokenizer+T5 model）.
- Feb 2022: Check out our sister repo [OpenDelta](https://github.com/thunlp/OpenDelta)!
- Dec 2021: `pip install openprompt`
- Dec 2021: [SuperGLUE performance](https://github.com/thunlp/OpenPrompt/tree/main/results) are added
- Dec 2021: We support **generation paradigm for all tasks** by adding a new verbalizer:[GenerationVerbalizer](https://github.com/thunlp/OpenPrompt/blob/main/openprompt/prompts/generation_verbalizer.py) and a [tutorial: 4.1_all_tasks_are_generation.py](https://github.com/thunlp/OpenPrompt/blob/main/tutorial/4.1_all_tasks_are_generation.py)
- Nov 2021: Now we have released a paper [OpenPrompt: An Open-source Framework for Prompt-learning](https://arxiv.org/abs/2111.01998).
- Nov 2021 PrefixTuning supports t5 now.
- Nov 2021: We made some major changes from the last version, where a flexible template language is newly introduced! Part of the docs is outdated and we will fix it soon.
## Overview
**Prompt-learning** is the latest paradigm to adapt pre-trained language models (PLMs) to downstream NLP tasks, which modifies the input text with a textual template and directly uses PLMs to conduct pre-trained tasks. This library provides a standard, flexible and extensible framework to deploy the prompt-learning pipeline. OpenPrompt supports loading PLMs directly from [huggingface transformers](https://github.com/huggingface/transformers). In the future, we will also support PLMs implemented by other libraries. For more resources about prompt-learning, please check our [paper list](https://github.com/thunlp/PromptPapers).
<div align="center">
<img src="https://z3.ax1x.com/2021/11/03/IAdT3D.png" width="85%" align="center"/>
</div>
## What Can You Do via OpenPrompt?
- Use the implementations of current prompt-learning approaches.* We have implemented various of prompting methods, including templating, verbalizing and optimization strategies under a unified standard. You can easily call and understand these methods.
- *Design your own prompt-learning work.* With the extensibility of OpenPrompt, you can quickly practice your prompt-learning ideas.
## Installation
**Note: Please use Python 3.8+ for OpenPrompt**
### Using Pip
Our repo is tested on Python **3.8+** and PyTorch **1.8.1+**, install OpenPrompt using pip as follows:
```shell
pip install openprompt
```
To play with the latest features, you can also install OpenPrompt from the source.
### Using Git
Clone the repository from github:
```shell
git clone https://github.com/thunlp/OpenPrompt.git
cd OpenPrompt
pip install -r requirements.txt
python setup.py install
```
Modify the code
```
python setup.py develop
```
## Use OpenPrompt
### Base Concepts
A `PromptModel` object contains a `PLM`, a (or multiple) `Template` and a (or multiple) `Verbalizer`, where the `Template` class is defined to wrap the original input with templates, and the `Verbalizer` class is to construct a projection between labels and target words in the current vocabulary. And a `PromptModel` object practically participates in training and inference.
### Introduction by a Simple Example
With the modularity and flexibility of OpenPrompt, you can easily develop a prompt-learning pipeline.
#### Step 1: Define a task
The first step is to determine the current NLP task, think about what’s your data looks like and what do you want from the data! That is, the essence of this step is to determine the `classes` and the `InputExample` of the task. For simplicity, we use Sentiment Analysis as an example. tutorial_task.
```python
from openprompt.data_utils import InputExample
classes = [ # There are two classes in Sentiment Analysis, one for negative and one for positive
"negative",
"positive"
]
dataset = [ # For simplicity, there's only two examples
# text_a is the input text of the data, some other datasets may have multiple input sentences in one example.
InputExample(
guid = 0,
text_a = "Albert Einstein was one of the greatest intellects of his time.",
),
InputExample(
guid = 1,
text_a = "The film was badly made.",
),
]
```
#### Step 2: Define a Pre-trained Language Models (PLMs) as backbone.
Choose a PLM to support your task. Different models have different attributes, we encourge you to use OpenPrompt to explore the potential of various PLMs. OpenPrompt is compatible with models on [huggingface](https://huggingface.co/transformers/).
```python
from openprompt.plms import load_plm
plm, tokenizer, model_config, WrapperClass = load_plm("bert", "bert-base-cased")
```
#### Step 3: Define a Template.
A `Template` is a modifier of the original input text, which is also one of the most important modules in prompt-learning.
We have defined `text_a` in Step 1.
```python
from openprompt.prompts import ManualTemplate
promptTemplate = ManualTemplate(
text = '{"placeholder":"text_a"} It was {"mask"}',
tokenizer = tokenizer,
)
```
#### Step 4: Define a Verbalizer
A `Verbalizer` is another important (but not necessary) in prompt-learning,which projects the original labels (we have defined them as `classes`, remember?) to a set of label words. Here is an example that we project the `negative` class to the word bad, and project the `positive` class to the words good, wonderful, great.
```python
from openprompt.prompts import ManualVerbalizer
promptVerbalizer = ManualVerbalizer(
classes = classes,
label_words = {
"negative": ["bad"],
"positive": ["good", "wonderful", "great"],
},
tokenizer = tokenizer,
)
```
#### Step 5: Combine them into a PromptModel
Given the task, now we have a `PLM`, a `Template` and a `Verbalizer`, we combine them into a `PromptModel`. Note that although the example naively combine the three modules, you can actually define some complicated interactions among them.
```python
from openprompt import PromptForClassification
promptModel = PromptForClassification(
template = promptTemplate,
plm = plm,
verbalizer = promptVerbalizer,
)
```
#### Step 6: Define a DataLoader
A ``PromptDataLoader`` is basically a prompt version of pytorch Dataloader, which also includes a ``Tokenizer``, a ``Template`` and a ``TokenizerWrapper``.
```python
from openprompt import PromptDataLoader
data_loader = PromptDataLoader(
dataset = dataset,
tokenizer = tokenizer,
template = promptTemplate,
tokenizer_wrapper_class=WrapperClass,
)
```
#### Step 7: Train and inference
Done! We can conduct training and inference the same as other processes in Pytorch.
```python
import torch
# making zero-shot inference using pretrained MLM with prompt
promptModel.eval()
with torch.no_grad():
for batch in data_loader:
logits = promptModel(batch)
preds = torch.argmax(logits, dim = -1)
print(classes[preds])
# predictions would be 1, 0 for classes 'positive', 'negative'
```
Please refer to our [tutorial scripts](https://github.com/thunlp/OpenPrompt/tree/main/tutorial), and [documentation](https://thunlp.github.io/OpenPrompt/) for more details.
## Datasets
We provide a series of download scripts in the `dataset/` folder, feel free to use them to download benchmarks.
## Performance Report
There are too many possible combinations powered by OpenPrompt. We are trying our best
to test the performance of different methods as soon as possible. The performance will be constantly updated into the [Tables](https://github.com/thunlp/OpenPrompt/tree/main/results/).
We also encourage the users to find the best hyper-parameters for their own tasks and report the results by making pull request.
## Known Issues
Major improvement/enhancement in future.
- We made some major changes from the last version, so part of the docs is outdated. We will fix it soon.
## Citation
Please cite our paper if you use OpenPrompt in your work
```bibtex
@article{ding2021openprompt,
title={OpenPrompt: An Open-source Framework for Prompt-learning},
author={Ding, Ning and Hu, Shengding and Zhao, Weilin and Chen, Yulin and Liu, Zhiyuan and Zheng, Hai-Tao and Sun, Maosong},
journal={arXiv preprint arXiv:2111.01998},
year={2021}
}
```
## Contributors
<!-- Copy-paste in your Readme.md file -->
<a href="https://github.com/thunlp/OpenPrompt/graphs/contributors">
<img src="https://contrib.rocks/image?repo=thunlp/OpenPrompt" />
</a>
We thank all the contributors to this project, more contributors are welcome!
Raw data
{
"_id": null,
"home_page": "https://github.com/9p4/OpenPrompt-updated",
"name": "openprompt-updated",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.11.0",
"maintainer_email": null,
"keywords": "PLM, prompt, AI, NLP",
"author": "Ning Ding, Shengding Hu, Weilin Zhao, Yulin Chen",
"author_email": "dingn18@mails.tsinghua.edu.cn",
"download_url": "https://files.pythonhosted.org/packages/49/12/e2e581bc9ac0b2fcf4694afe90fc634ba0b0fbce0fa978618c9177be5e73/openprompt_updated-1.0.1.tar.gz",
"platform": null,
"description": "\n\n<div align=\"center\">\n\n<img src=\"https://z3.ax1x.com/2021/11/11/IwED0K.png\" width=\"350px\">\n\n**An Open-Source Framework for Prompt-learning.**\n\n------\n\n<p align=\"center\">\n <a href=\"#Overview\">Overview</a> \u2022\n <a href=\"#installation\">Installation</a> \u2022\n <a href=\"#use-openprompt\">How To Use</a> \u2022\n <a href=\"https://thunlp.github.io/OpenPrompt/\">Docs</a> \u2022\n <a href=\"https://arxiv.org/abs/2111.01998\">Paper</a> \u2022\n <a href=\"#citation\">Citation</a> \u2022\n <a href=\"https://github.com/thunlp/OpenPrompt/tree/main/results/\">Performance</a> \u2022\n</p>\n\n</div>\n\n\n\n\n\n\n## What's New?\n\n \n\n- \u2757\ufe0f April 2023: $\\color{red}{\\normalsize{\\textbf{Want to build your Chat AI?}}}$ **We are releasing [UltraChat](https://github.com/thunlp/UltraChat), use OpenPrompt and UltraChat to conduct supervised instruction tuning, see** [`./tutorial/9_UltraChat.py`](./tutorial/9_UltraChat.py).\n- Aug 2022: Thanks to contributor [zhiyongLiu1114](https://github.com/zhiyongLiu1114), OpenPrompt now supports [ERNIE 1.0](https://github.com/thunlp/OpenPrompt/tree/main/tutorial/7_ernie_paddlepaddle) in PaddlePaddle.\n- July 2022: OpenPrompt supports OPT now.\n- June 2022: OpenPrompt wins ACL 2022 Best Demo Paper Award.\n- Mar 2022: We add a [tutorial](https://github.com/thunlp/OpenPrompt/blob/main/tutorial/6.1_chinese_dataset_uer_t5.py) as the response to [issue 124](https://github.com/thunlp/OpenPrompt/issues/124), which uses a customized tokenizer_wrapper to perform tasks that are not in the default configuration of OpenPrompt (e.g., Bert tokenizer+T5 model\uff09.\n- Feb 2022: Check out our sister repo [OpenDelta](https://github.com/thunlp/OpenDelta)!\n- Dec 2021: `pip install openprompt`\n- Dec 2021: [SuperGLUE performance](https://github.com/thunlp/OpenPrompt/tree/main/results) are added\n- Dec 2021: We support **generation paradigm for all tasks** by adding a new verbalizer:[GenerationVerbalizer](https://github.com/thunlp/OpenPrompt/blob/main/openprompt/prompts/generation_verbalizer.py) and a [tutorial: 4.1_all_tasks_are_generation.py](https://github.com/thunlp/OpenPrompt/blob/main/tutorial/4.1_all_tasks_are_generation.py)\n- Nov 2021: Now we have released a paper [OpenPrompt: An Open-source Framework for Prompt-learning](https://arxiv.org/abs/2111.01998).\n- Nov 2021 PrefixTuning supports t5 now.\n- Nov 2021: We made some major changes from the last version, where a flexible template language is newly introduced! Part of the docs is outdated and we will fix it soon.\n\n\n## Overview\n\n**Prompt-learning** is the latest paradigm to adapt pre-trained language models (PLMs) to downstream NLP tasks, which modifies the input text with a textual template and directly uses PLMs to conduct pre-trained tasks. This library provides a standard, flexible and extensible framework to deploy the prompt-learning pipeline. OpenPrompt supports loading PLMs directly from [huggingface transformers](https://github.com/huggingface/transformers). In the future, we will also support PLMs implemented by other libraries. For more resources about prompt-learning, please check our [paper list](https://github.com/thunlp/PromptPapers).\n\n\n\n<div align=\"center\">\n\n\n<img src=\"https://z3.ax1x.com/2021/11/03/IAdT3D.png\" width=\"85%\" align=\"center\"/>\n\n</div>\n\n\n\n## What Can You Do via OpenPrompt?\n\n\n\n\n\n\n\n- Use the implementations of current prompt-learning approaches.* We have implemented various of prompting methods, including templating, verbalizing and optimization strategies under a unified standard. You can easily call and understand these methods.\n- *Design your own prompt-learning work.* With the extensibility of OpenPrompt, you can quickly practice your prompt-learning ideas.\n\n\n\n\n## Installation\n\n**Note: Please use Python 3.8+ for OpenPrompt**\n\n### Using Pip\n\nOur repo is tested on Python **3.8+** and PyTorch **1.8.1+**, install OpenPrompt using pip as follows:\n\n\n\n\n```shell\npip install openprompt\n```\n\nTo play with the latest features, you can also install OpenPrompt from the source.\n\n### Using Git\n\nClone the repository from github:\n\n```shell\ngit clone https://github.com/thunlp/OpenPrompt.git\ncd OpenPrompt\npip install -r requirements.txt\npython setup.py install\n```\n\nModify the code\n\n```\npython setup.py develop\n```\n\n\n\n## Use OpenPrompt\n\n### Base Concepts\n\nA `PromptModel` object contains a `PLM`, a (or multiple) `Template` and a (or multiple) `Verbalizer`, where the `Template` class is defined to wrap the original input with templates, and the `Verbalizer` class is to construct a projection between labels and target words in the current vocabulary. And a `PromptModel` object practically participates in training and inference.\n\n### Introduction by a Simple Example\n\nWith the modularity and flexibility of OpenPrompt, you can easily develop a prompt-learning pipeline.\n\n#### Step 1: Define a task\n\nThe first step is to determine the current NLP task, think about what\u2019s your data looks like and what do you want from the data! That is, the essence of this step is to determine the\u00a0`classes`\u00a0and the\u00a0`InputExample`\u00a0of the task. For simplicity, we use Sentiment Analysis as an example.\u00a0tutorial_task.\n\n```python\nfrom openprompt.data_utils import InputExample\nclasses = [ # There are two classes in Sentiment Analysis, one for negative and one for positive\n \"negative\",\n \"positive\"\n]\ndataset = [ # For simplicity, there's only two examples\n # text_a is the input text of the data, some other datasets may have multiple input sentences in one example.\n InputExample(\n guid = 0,\n text_a = \"Albert Einstein was one of the greatest intellects of his time.\",\n ),\n InputExample(\n guid = 1,\n text_a = \"The film was badly made.\",\n ),\n]\n```\n\n\n\n#### Step 2: Define a Pre-trained Language Models (PLMs) as backbone.\n\nChoose a PLM to support your task. Different models have different attributes, we encourge you to use OpenPrompt to explore the potential of various PLMs. OpenPrompt is compatible with models on\u00a0[huggingface](https://huggingface.co/transformers/).\n\n```python\nfrom openprompt.plms import load_plm\nplm, tokenizer, model_config, WrapperClass = load_plm(\"bert\", \"bert-base-cased\")\n```\n\n\n\n#### Step 3: Define a Template.\n\nA\u00a0`Template`\u00a0is a modifier of the original input text, which is also one of the most important modules in prompt-learning.\u00a0\nWe have defined `text_a` in Step 1.\n\n```python\nfrom openprompt.prompts import ManualTemplate\npromptTemplate = ManualTemplate(\n text = '{\"placeholder\":\"text_a\"} It was {\"mask\"}',\n tokenizer = tokenizer,\n)\n```\n\n\n\n#### Step 4: Define a Verbalizer\n\nA\u00a0`Verbalizer`\u00a0is another important (but not necessary) in prompt-learning,which projects the original labels (we have defined them as\u00a0`classes`, remember?) to a set of label words. Here is an example that we project the\u00a0`negative`\u00a0class to the word\u00a0bad, and project the\u00a0`positive`\u00a0class to the words\u00a0good,\u00a0wonderful,\u00a0great.\n\n```python\nfrom openprompt.prompts import ManualVerbalizer\npromptVerbalizer = ManualVerbalizer(\n classes = classes,\n label_words = {\n \"negative\": [\"bad\"],\n \"positive\": [\"good\", \"wonderful\", \"great\"],\n },\n tokenizer = tokenizer,\n)\n```\n\n\n\n#### Step 5: Combine them into a PromptModel\n\nGiven the task, now we have a\u00a0`PLM`, a\u00a0`Template`\u00a0and a\u00a0`Verbalizer`, we combine them into a\u00a0`PromptModel`. Note that although the example naively combine the three modules, you can actually define some complicated interactions among them.\n\n```python\nfrom openprompt import PromptForClassification\npromptModel = PromptForClassification(\n template = promptTemplate,\n plm = plm,\n verbalizer = promptVerbalizer,\n)\n```\n\n#### Step 6: Define a DataLoader\n\nA ``PromptDataLoader`` is basically a prompt version of pytorch Dataloader, which also includes a ``Tokenizer``, a ``Template`` and a ``TokenizerWrapper``.\n\n```python\nfrom openprompt import PromptDataLoader\ndata_loader = PromptDataLoader(\n dataset = dataset,\n tokenizer = tokenizer,\n template = promptTemplate,\n tokenizer_wrapper_class=WrapperClass,\n)\n```\n\n#### Step 7: Train and inference\n\nDone! We can conduct training and inference the same as other processes in Pytorch.\n\n\n```python\nimport torch\n\n# making zero-shot inference using pretrained MLM with prompt\npromptModel.eval()\nwith torch.no_grad():\n for batch in data_loader:\n logits = promptModel(batch)\n preds = torch.argmax(logits, dim = -1)\n print(classes[preds])\n# predictions would be 1, 0 for classes 'positive', 'negative'\n```\n\nPlease refer to our [tutorial scripts](https://github.com/thunlp/OpenPrompt/tree/main/tutorial), and [documentation](https://thunlp.github.io/OpenPrompt/) for more details.\n\n## Datasets\n\nWe provide a series of download scripts in the `dataset/` folder, feel free to use them to download benchmarks.\n\n## Performance Report\nThere are too many possible combinations powered by OpenPrompt. We are trying our best\nto test the performance of different methods as soon as possible. The performance will be constantly updated into the [Tables](https://github.com/thunlp/OpenPrompt/tree/main/results/).\nWe also encourage the users to find the best hyper-parameters for their own tasks and report the results by making pull request.\n\n## Known Issues\nMajor improvement/enhancement in future.\n\n- We made some major changes from the last version, so part of the docs is outdated. We will fix it soon.\n\n## Citation\nPlease cite our paper if you use OpenPrompt in your work\n\n```bibtex\n@article{ding2021openprompt,\n title={OpenPrompt: An Open-source Framework for Prompt-learning},\n author={Ding, Ning and Hu, Shengding and Zhao, Weilin and Chen, Yulin and Liu, Zhiyuan and Zheng, Hai-Tao and Sun, Maosong},\n journal={arXiv preprint arXiv:2111.01998},\n year={2021}\n}\n```\n## Contributors\n\n<!-- Copy-paste in your Readme.md file -->\n\n<a href=\"https://github.com/thunlp/OpenPrompt/graphs/contributors\">\n <img src=\"https://contrib.rocks/image?repo=thunlp/OpenPrompt\" />\n</a>\n\nWe thank all the contributors to this project, more contributors are welcome!\n\n\n",
"bugtrack_url": null,
"license": "Apache",
"summary": "An open source framework for prompt-learning, patched to work with newer versions of Python",
"version": "1.0.1",
"project_urls": {
"Homepage": "https://github.com/9p4/OpenPrompt-updated"
},
"split_keywords": [
"plm",
" prompt",
" ai",
" nlp"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "8f374c831921af6ee1f2a1c1f911eccf439e73ee26d9b9398231f310ee663c5b",
"md5": "3a70096b557bfac871d6b50f4ec1b9ef",
"sha256": "5c364a785bbc88e114adab1ba308ac17f01403159ad1aa8d99092ab908d5eebc"
},
"downloads": -1,
"filename": "openprompt_updated-1.0.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "3a70096b557bfac871d6b50f4ec1b9ef",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.11.0",
"size": 147617,
"upload_time": "2024-08-08T23:43:36",
"upload_time_iso_8601": "2024-08-08T23:43:36.020749Z",
"url": "https://files.pythonhosted.org/packages/8f/37/4c831921af6ee1f2a1c1f911eccf439e73ee26d9b9398231f310ee663c5b/openprompt_updated-1.0.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "4912e2e581bc9ac0b2fcf4694afe90fc634ba0b0fbce0fa978618c9177be5e73",
"md5": "7fa8b11184014a459fe5509ed6f881aa",
"sha256": "20ca4aadaff915c13ee55b6841df0be2e0da98d379c6f90972e33d3da3d0e16b"
},
"downloads": -1,
"filename": "openprompt_updated-1.0.1.tar.gz",
"has_sig": false,
"md5_digest": "7fa8b11184014a459fe5509ed6f881aa",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.11.0",
"size": 114660,
"upload_time": "2024-08-08T23:43:37",
"upload_time_iso_8601": "2024-08-08T23:43:37.397954Z",
"url": "https://files.pythonhosted.org/packages/49/12/e2e581bc9ac0b2fcf4694afe90fc634ba0b0fbce0fa978618c9177be5e73/openprompt_updated-1.0.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-08-08 23:43:37",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "9p4",
"github_project": "OpenPrompt-updated",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [],
"lcname": "openprompt-updated"
}
JSON
