OpenChat: Opensource chatting framework for generative models

Overview

OpenChat: Opensource chatting framework for generative models

    ____   ____   ______ _   __   ______ __  __ ___   ______
   / __ \ / __ \ / ____// | / /  / ____// / / //   | /_  __/
  / / / // /_/ // __/  /  |/ /  / /    / /_/ // /| |  / /   
 / /_/ // ____// /___ / /|  /  / /___ / __  // ___ | / /    
 \____//_/    /_____//_/ |_/   \____//_/ /_//_/  |_|/_/     
  • OpenChat is opensource chatting framework for generative models.
  • You can talk with AI with only one line of code.



Installation

pip install openchat



Model Supports

  • DialoGPT : [small, medium, large]
  • BlenderBot : [small, medium, large, xlarge]
  • Coming Soon...



Usage

1. Terminal environment

  • Chatting on terminal.
  • Type .exit if you want to exit dialogue.
  • Type .clear if you want to clear all histories.
  • You can add additional user keywords. (Details are described below.)
>>> from openchat import OpenChat
>>> OpenChat(model="blenderbot", size="large")
Chat with facebook/blenderbot-1B-distill !
user : Hi. It is nice day.
bot : It is a nice day out. I am going to go for a walk. What are you up to?
user : I'll play computer game. Do you like playing game?
bot : I love playing video games. What kind of games do you like to play on your computer?
user : I love racing game. It is so exited. How about you?
bot : I love racing games as well. I like to play on the Playstation 4. What about you?
user : Usually I like to play on PC. I have a question. Who is CEO of Apple?
bot : The CEO of Apple is Steve Jobs. He founded the company in 1976.
user : That's right. I have an IPhone. It is my favorite phone. What is your phone?
bot : I have a Samsung Galaxy S8. I love it. It's the best phone I've ever had.
user : .exit
bot : good bye.

  • If you want to use GPU, use argument device="cuda"
>>> from openchat import OpenChat
>>> OpenChat(model="blenderbot", size="large", device="cuda")

2. Your own environment (not terminal)

  • You can use user id to manage user-specific history.
  • This can be useful when deployed on Facebook messenger or WhatsApp.
  • There is a web demo implementation in the /demo folder.


2.1. Write your own environment class

  • Make your own environment class inherited from BaseEnv
  • And implement your own run(model: BaseModel) method like below.
from typing import Dict
from flask import Flask, render_template
from flask_cors import CORS
from openchat.envs import BaseEnv
from openchat.models import BaseModel


class WebDemoEnv(BaseEnv):

    def __init__(self):
        super().__init__()
        self.app = Flask(__name__)
        CORS(self.app)

    def run(self, model: BaseModel):

        @self.app.route("/")
        def index():
            return render_template("index.html", title=model.name)

        @self.app.route('/send//', methods=['GET'])
        def send(user_id, text: str) -> Dict[str, str]:

            if text in self.keywords:
                # Format of self.keywords dictionary
                # self.keywords['/exit'] = (exit_function, 'good bye.')

                _out = self.keywords[text][1]
                # text to print when keyword triggered

                self.keywords[text][0](user_id, text)
                # function to operate when keyword triggered

            else:
                _out = model.predict(user_id, text)

            return {"output": _out}

        self.app.run(host="0.0.0.0", port=8080)

2.2. Start to run application.

from openchat import OpenChat
from demo.web_demo_env import WebDemoEnv

OpenChat(model="blenderbot", size="large", env=WebDemoEnv())



3. Additional Options

3.1. Add custom Keywords

  • You can add new manual keyword such as .exit, .clear,
  • call the self.add_keyword('.new_keyword', 'message to print', triggered_function)' method.
  • triggered_function should be form of function(user_id:str, text:str)
from openchat.envs import BaseEnv


class YourOwnEnv(BaseEnv):
    
    def __init__(self):
        super().__init__()
        self.add_keyword(".new_keyword", "message to print", self.function)

    def function(self, user_id: str, text: str):
        """do something !"""
        



3.2. Modify generation options

  • You can modify max_context_length (number of input history tokens, default is 128).
>>> OpenChat(size="large", device="cuda", max_context_length=256)

  • You can modify generation options ['num_beams', 'top_k', 'top_p'].
>>> model.predict(
...     user_id="USER_ID",
...     text="Hello.",
...     num_beams=5,
...     top_k=20,
...     top_p=0.8,
... )



3.3. Check histories

  • You can check all dialogue history using self.histories
from openchat.envs import BaseEnv


class YourOwnEnv(BaseEnv):
    
    def __init__(self):
        super().__init__()
        print(self.histories)
{
    user_1 : {'user': [] , 'bot': []},
    user_2 : {'user': [] , 'bot': []},
    ...more...
    user_n : {'user': [] , 'bot': []},
}

3.4. Clear histories

  • You can clear all dialogue histories
from flask import Flask
from openchat.envs import BaseEnv
from openchat.models import BaseModel

class YourOwnEnv(BaseEnv):
    
    def __init__(self):
        super().__init__()
        self.app = Flask(__name__)

    def run(self, model: BaseModel):
        
        @self.app.route('/send//', methods=['GET'])
        def send(user_id, text: str) -> Dict[str, str]:
            
            self.clear(user_id, text)
            # clear all histories ! 



License

Copyright 2021 Hyunwoong Ko.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
Owner
Hyunwoong Ko
Co-Founder and Research Engineer at @tunib-ai. previously @kakaobrain.
Hyunwoong Ko
Spokestack is a library that allows a user to easily incorporate a voice interface into any Python application with a focus on embedded systems.

Welcome to Spokestack Python! This library is intended for developing voice interfaces in Python. This can include anything from Raspberry Pi applicat

Spokestack 133 Sep 20, 2022
Official code repository of the paper Linear Transformers Are Secretly Fast Weight Programmers.

Linear Transformers Are Secretly Fast Weight Programmers This repository contains the code accompanying the paper Linear Transformers Are Secretly Fas

Imanol Schlag 77 Dec 19, 2022
💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants

Rasa Open Source Rasa is an open source machine learning framework to automate text-and voice-based conversations. With Rasa, you can build contextual

Rasa 15.3k Dec 30, 2022
NL. The natural language programming language.

NL A Natural-Language programming language. Built using Codex. A few examples are inside the nl_projects directory. How it works Write any code in pur

2 Jan 17, 2022
Idea is to build a model which will take keywords as inputs and generate sentences as outputs.

keytotext Idea is to build a model which will take keywords as inputs and generate sentences as outputs. Potential use case can include: Marketing Sea

Gagan Bhatia 364 Jan 03, 2023
Code for EMNLP20 paper: "ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-training"

ProphetNet-X This repo provides the code for reproducing the experiments in ProphetNet. In the paper, we propose a new pre-trained language model call

Microsoft 394 Dec 17, 2022
Multiple implementations for abstractive text summurization , using google colab

Text Summarization models if you are able to endorse me on Arxiv, i would be more than glad https://arxiv.org/auth/endorse?x=FRBB89 thanks This repo i

463 Dec 26, 2022
Nested Named Entity Recognition for Chinese Biomedical Text

CBio-NAMER CBioNAMER (Nested nAMed Entity Recognition for Chinese Biomedical Text) is our method used in CBLUE (Chinese Biomedical Language Understand

8 Dec 25, 2022
DeBERTa: Decoding-enhanced BERT with Disentangled Attention

DeBERTa: Decoding-enhanced BERT with Disentangled Attention This repository is the official implementation of DeBERTa: Decoding-enhanced BERT with Dis

Microsoft 1.2k Jan 03, 2023
Using Bert as the backbone model for lime, designed for NLP task explanation (sentence pair text classification task)

Lime Comparing deep contextualized model for sentences highlighting task. In addition, take the classic explanation model "LIME" with bert-base model

JHJu 2 Jan 18, 2022
A program that uses real statistics to choose the best times to bet on BloxFlip's crash gamemode

Bloxflip Smart Bet A program that uses real statistics to choose the best times to bet on BloxFlip's crash gamemode. https://bloxflip.com/crash. THIS

43 Jan 05, 2023
Natural language computational chemistry command line interface.

nlcc Install pip install nlcc Must have Open-AI Codex key: export OPENAI_API_KEY=your key here then nlcc key bindings ctrl-w copy to clipboard (Note

Andrew White 37 Dec 14, 2022
Unet-TTS: Improving Unseen Speaker and Style Transfer in One-shot Voice Cloning

Unet-TTS: Improving Unseen Speaker and Style Transfer in One-shot Voice Cloning English | 中文 ❗ Now we provide inferencing code and pre-training models

164 Jan 02, 2023
Machine translation models released by the Gourmet project

Gourmet Models Overview The Gourmet project has released several machine translation models to translate low-resource languages. This repository conta

Edinburgh NLP 5 Dec 08, 2021
Research code for the paper "Fine-tuning wav2vec2 for speaker recognition"

Fine-tuning wav2vec2 for speaker recognition This is the code used to run the experiments in https://arxiv.org/abs/2109.15053. Detailed logs of each t

Nik 103 Dec 26, 2022
A collection of Korean Text Datasets ready to use using Tensorflow-Datasets.

tfds-korean A collection of Korean Text Datasets ready to use using Tensorflow-Datasets. TensorFlow-Datasets를 이용한 한국어/한글 데이터셋 모음입니다. Dataset Catalog |

Jeong Ukjae 20 Jul 11, 2022
Official PyTorch implementation of "Dual Path Learning for Domain Adaptation of Semantic Segmentation".

Dual Path Learning for Domain Adaptation of Semantic Segmentation Official PyTorch implementation of "Dual Path Learning for Domain Adaptation of Sema

27 Dec 22, 2022
This is my reading list for my PhD in AI, NLP, Deep Learning and more.

This is my reading list for my PhD in AI, NLP, Deep Learning and more.

Zhong Peixiang 156 Dec 21, 2022
Use PaddlePaddle to reproduce the paper:mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer

MT5_paddle Use PaddlePaddle to reproduce the paper:mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer English | 简体中文 mT5: A Massively

2 Oct 17, 2021