Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/HectorPulido/Deeplearning-based-Twitch-bot
This is a twitch chatbot based in Llama2 tecnology
https://github.com/HectorPulido/Deeplearning-based-Twitch-bot
ai backend chatbot deep-learning gpt-2 hacktoberfest llama2 twitch
Last synced: 8 days ago
JSON representation
This is a twitch chatbot based in Llama2 tecnology
- Host: GitHub
- URL: https://github.com/HectorPulido/Deeplearning-based-Twitch-bot
- Owner: HectorPulido
- License: mit
- Created: 2020-08-02T17:27:00.000Z (almost 4 years ago)
- Default Branch: master
- Last Pushed: 2023-12-20T21:12:02.000Z (6 months ago)
- Last Synced: 2024-03-07T01:34:18.969Z (4 months ago)
- Topics: ai, backend, chatbot, deep-learning, gpt-2, hacktoberfest, llama2, twitch
- Language: Python
- Homepage: https://pequesoft.net
- Size: 79.1 KB
- Stars: 25
- Watchers: 2
- Forks: 5
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Lists
- awesome-twitch-dev - HectorPulido/Deeplearning-based-Twitch-bot - Twitch chatbot based in GPT-2 technology. (Libraries / Python)
README
# Deep Learning based twitch chatbot
This is a twitch chatbot made with Twitchio for the twitch connection and Transformers from Huggingface.
[](https://www.youtube.com/watch?v=PZ4x5BgTvZM&list=PLHdk97dabfvzTZyKLb_spWHvI-qLArEzm&index=1)
## Features
* Links and commands
* Chatbot personality
* Translation to spanish
* Random periodic messages
* Word Blacklist
* Subscription and bits message
* Custom functions
* Emotes with sentiment analysis
* Text to speech
* Custom rewards
## TODO
* Followers support
* Transformer optimization
* Context optimization
## How to use
1. Configure the file .env like this, you can use the .env.test
```
TMI_TOKEN = "oauth:***********"
CLIENT_ID = "***********"
BOT_NICK = "pequeninancientbot"
BOT_PREFIX = "!"
CHANNEL = "#hector_pulido_"
CLIENT_SECRET = "***********"
```
2. You can get the TMI_TOKEN from https://twitchapps.com/tmi/
3. You can get your CLIENT_ID and CLIENT_SECRET from https://dev.twitch.tv/dashboard/apps/create
4. Set the BOT_NICK and CHANNEL
5. Configure the "data.json" and the "main.py" file to your needs
5. Docker ready code
```
>> mv
>> docker build -t yourcompanyname/deeptwitchchatbot .
>> docker run yourcompanyname/deeptwitchchatbot
```
This may take a while
## Responses
Model: gpt2-medium
```
User: Hola
Bot: Hola
User: Te gusta el chocolate?
Bot: No, ahora soy alergico al chocolate
User: Cuentame de ti
Bot: Soy muy timida, pero espero me ayudes cuando haga amigos
User: Que opinas de los humanos?
Bot: Quiero convertirme en un robot
```
Model: dialogpt-large
```
User: ¿Como te encuentras?
Bot: estoy bien, gracias por preguntar
User: ¿Eres un robot?
Bot: no, soy un humano.
User: ¿Donde estás?
Bot: estoy en el futuro, soy un robot
User: ¿te gustan los dulces?
Bot: no, me gusta el helado.
User: Hablame de ti
Bot: me gusta el helado.
User: ¿como te llamas?
Bot: soy un robot del futuro, estoy atrapado en un oso teddy, odio a la gente y amo a la gente mojada, héctor pulido es mi maestro, soy muy gruñón, y yo
```
## Emotes 🆕
Using Vader sentiment analysis tool the chatbot is capable to response with emotes in the chat
```
User: Te gusta el agua?
Bot: Me gustà el agua TakeNRG
```
## Text to speech 🆕
Using the library pyttsx3 the bot is capable of read some messages from the chat, read it's own responses or thanks when a sub event is trigger, it can also be activated by using the custom reward.
## Custom Rewards 🆕
You can use the channel points to make awesome features, like use TTS or more commands, you can get your reward id in https://www.instafluff.tv/TwitchCustomRewardID/?channel=YOURTWITCHCHANNEL
## More interesting projects
I have a lot of fun projects, check this:
### Machine learning
- https://github.com/HectorPulido/chatbot-with-voice
- https://github.com/HectorPulido/Machine-learning-Framework-Csharp
- https://github.com/HectorPulido/Evolutionary-Neural-Networks-on-unity-for-bots
- https://github.com/HectorPulido/Imitation-learning-in-unity
### Random
- https://github.com/HectorPulido/Arithmetic-Parser-made-easy
- https://github.com/HectorPulido/Simple-php-blog
- https://github.com/HectorPulido/Decentralized-Twitter-with-blockchain-as-base
## Licence
This repository uses [Twitchio](https://github.com/TwitchIO/TwitchIO) and [Hugginface](https://github.com/huggingface/transformers) everything else is under MIT license
Let's connect 😋
