A robust, all-in-one GPT3 interface for Discord. Chat just like ChatGPT right inside Discord! Generate beautiful AI art using DALL-E 2! Automatically moderate your server using AI! Upload documents, videos, and files to get AI-assisted insights! A thorough integration with permanent conversation memory, automatic request retry, fault tolerance and reliability for servers of any scale, and much more.
- **CUSTOM INDEXES** - This is a huge update. You can now upload files to your server and use them as custom context when asking GPT3 questions. You can also use webpage links as context, images, full documents, csvs, powerpoints, audio files, and even **youtube videos**! Read more in the 'Custom Indexes' section below.
- **Have long term, permanent conversations with the bot, just like chatgpt, with `/gpt converse`** - Conversations happen in threads that get automatically cleaned up!
- **DALL-E Image Generation** - Generate DALL-E AI images right in discord with `/dalle draw <prompt>`! It even supports multiple image qualities, multiple images, creating image variants, retrying, and saving images.
- **DALL-E Image Prompt Optimization** - Given some text that you're trying to generate an image for, the bot will automatically optimize the text to be more DALL-E friendly! `/dalle optimize <prompt>`
- **Redo Requests** - A simple button after the GPT3 response or DALL-E generation allows you to redo the initial prompt you asked. You can also redo conversation messages by just editing your message!
- Change and view model parameters such as temp, top_p, and etc directly within discord.
- Tracks token usage automatically
- Automatic pagination and discord support, the bot will automatically send very long message as multiple messages, and is able to send discord code blocks and emoji, gifs, etc.
- A low usage mode, use a command to automatically switch to a cheaper and faster model to conserve your tokens during times of peak usage.
- Prints debug to a channel of your choice, so you can view the raw response JSON
- Ability to specify a limit to how long a conversation can be with the bot, to conserve your tokens.
These commands are grouped, so each group has a prefix but you can easily tab complete the command without the prefix. For example, for `/gpt ask`, if you type `/ask` and press tab, it'll show up too.
`/gpt edit <instruction> <input> <temp> <top_p> <codex>` Use the bot to edit text using the given instructions for how to do it, currently an alpha openai feature so results might vary. Codex uses a model trained on code. Editing is currently free
-`opener:<opener text>` - Start a conversation with the bot, with a custom opener text (this is useful if you want it to take on a custom personality from the start).
- Loads files from the `/openers` folder, has autocomplete support so files in the folder will show up. Added before the `opener` as both can be used at the same time
- Can use .json files in the `{"text": "your prompt", "temp":0, "top_p":0,"frequency_penalty":0,"presence_penalty":0}` format to include permanent overrides
This bot supports per-user custom indexes. This means that users can upload files of their choosing, such as PDFs and ask GPT to answer questions based on those files.
`/index add file:<file> or link:<link>` - Use a document or use a link to create/add to your indexes. If you provide a youtube link, the transcript of the video will be used. If you provide a web url, the contents of the webpage will be used, if you provide an image, the image text will be extracted and used!
`/index query query:<prompt>` - Query your current index for a given prompt. GPT will answer based on your current document/indedx
`/index load index:<index>` - Load a previously created index to query
`/system settings <setting> <value>` - Change a model setting to a new value. Has autocomplete support, certain settings will have autocompleted values too.
`/system usage` Estimate current usage details (based on davinci)
`/system settings low_usage_mode True/False` Turn low usage mode on and off. If on, it will use the curie-001 model, and if off, it will use the davinci-003 model.
`/system delete-conversation-threads` - Delete all threads related to this bot across all servers.
`/system local-size` - Get the size of the local dalleimages folder
`/system clear-local` - Clear all the local dalleimages.
- Set the moderation thresholds of the bot for the specific type of moderation (`warn` or `delete`). You can view the thresholds by typing just `/mod config type:<warn/delete>` without any other parameters. You don't have to set all of them, you can just set one or two items if you want. For example, to set the hate threshold for warns, you can type `/mod config type:warn hate:0.2`
- Lower values are more strict, higher values are more lenient. There are default values that I've fine tuned the service with for a general server.
The bot needs Administrative permissions for this, and you need to set `MODERATIONS_ALERT_CHANNEL` to the channel ID of a desired channel in your .env file if you want to receive alerts about moderated messages.
This uses the OpenAI Moderations endpoint to check for messages, requests are only sent to the moderations endpoint at a MINIMUM request gap of 0.5 seconds, to ensure you don't get blocked and to ensure reliability.
The bot uses numerical thresholds to determine whether a message is toxic or not, and I have manually tested and fine tuned these thresholds to a point that I think is good, please open an issue if you have any suggestions for the thresholds!
There are two thresholds for the bot, there are instances in which the bot will outright delete a message and an instance where the bot will send a message to the alert channel notifying admins and giving them quick options to delete and timeout the user (check out the screenshots at the beginning of the README to see this).
If you'd like to help us test and fine tune our thresholds for the moderation service, please join this test server: https://discord.gg/CWhsSgNdrP. You can let off some steam in a controlled environment ;)
Pinecone is a vector database. The OpenAI Ada embeddings endpoint turns pieces of text into embeddings. The way that this feature works is by embedding the user prompts and the GPT3 responses, storing them in a pinecone index, and then retrieving the most relevant bits of conversation whenever a new user prompt is given in a conversation.
**You do NOT need to use pinecone, if you do not define a `PINECONE_TOKEN` in your `.env` file, the bot will default to not using pinecone, and will use conversation summarization as the long term conversation method instead.**
To enable permanent memory with pinecone, you must define a `PINECONE_TOKEN` in your `.env` file as follows (along with the other variables too):
To get a pinecone token, you can sign up for a free pinecone account here: https://app.pinecone.io/ and click the "API Keys" section on the left navbar to find the key. (I am not affiliated with pinecone).
After signing up for a free pinecone account, you need to create an index in pinecone. To do this, go to the pinecone dashboard and click "Create Index" on the top right.
Permanent memory using pinecone is still in alpha, I will be working on cleaning up this work, adding auto-clearing, and optimizing for stability and reliability, any help and feedback is appreciated (**add me on Discord Kaveen#0001 for pinecone help**)! If at any time you're having too many issues with pinecone, simply remove the `PINECONE_TOKEN` line in your `.env` file and the bot will revert to using conversation summarizations.
This bot supports per-user custom indexes. This means that users can upload files of their choosing, such as PDFs and ask GPT to answer questions based on those files. We also support using URLs for indexes.
Indexes can be combined with other indexes through a composition. To combine indexes, you can run the `/index compose` command, and select the indexes that you want to combine together. You should only combine relevant indexes together, combining irrelevant indexes together will result in poor results (for example, don't upload a math textbook and then upload a large set of poems and combine them together). When creating a composition, you will be given the option to do a "Deep" composition, deep compositions are more detailed and will give you better results, but are incredibly costly and will sometimes take multiple minutes to compose.
You can also compose a singular index with itself with "Deep Compose", this will give you a more detailed version of the index, but will be costly and will sometimes take multiple minutes to compose. **Deep compositions are useless for very short documents!**
This bot supports and uses DeepL for translations (optionally). If you want to enable the translations service, you can add a line in your `.env` file as follows:
```
DEEPL_TOKEN="your deepl token"
```
You can get a DeepL token by signing up at https://www.deepl.com/pro-api?cta=header-pro-api/ and clicking on the *free plan* to start. The DeepL translation service unlocks some new commands for your bot:
`/translate <text> <language>` - Translate any given piece of text into the language that you provide
`/languages` - See a list of all supported languages
Using DeepL also adds a new app menu button (when you right click a message) to the bot which allows you to quickly translate any message in a channel into any language you want:
This bot supports multi-user tenancy in regards to API keys. This means that, if you wanted, you could make it such that each user needs to enter their own API key in order to use commands that use GPT3 and DALLE.
To enable this, add the following line to the end of your `.env` file:
```env
USER_INPUT_API_KEYS="True"
```
Then, restart the bot, and it will set up the system for everyone to input their own API keys.
The bot will use SQLite to store API keys for the users, each user's key will be saved with a USER_ID <> API_KEY mapping in SQLite, and will be persistent across restarts. All the data will be saved in a file called `user_key_db.sqlite` in the current working directory of the bot.
With this feature enabled, any attempt to use a GPT3 or DALL-E command without a valid API key set for the user will pop up the following modal for them to enter their API key:
Once the user enters their key, the bot will send a small test request to OpenAI to validate that the key indeed works, if not, it will tell the user to try again and tell them why it did not work.
After the user's key is validated, they will be able to use GPT3 and DALLE commands.
The Moderations service still uses the main API key defined in the `.env` file. Pinecone and discord-tokens are also per-host tokens, not per-user.
If you follow the link above, you will now get to detailed step-by-step guides that will help you to install and set up your GPT3Discord bot quickly and easily. If you still run into problems or have suggestions for improving the guides, you can join the [**Discord-Server**](https://discord.gg/WvAHXDMS7Q) and we will try to help you. Keep in mind that the maintainers are volunteers and will try to help you on their schedule.
All the model parameters are configurable inside discord. Type `/system settings` to view all the configurable parameters, and use `/system settings <param> <value>` to set parameters.
For example, if I wanted to change the number of images generated by DALL-E by default to 4, I can type the following command in discord: `/system settings num_images 4`
You can learn how to add the discord bot to your server via https://www.ionos.co.uk/digitalguide/server/know-how/creating-discord-bot/
Both the OpenAI API key and the Discord bot token needed to be loaded into a .env file in the same local directory as the bot file.
You also need to add a DEBUG_GUILD id and a DEBUG_CHANNEL id, the debug guild id is a server id, and the debug channel id is a text-channel id in Discord. Your final .env file should look like the following:
You also need to add the allowed guilds that the bot can operate on, this is the `ALLOWED_GUILDS` field, to get a guild ID, right click a server and click "Copy ID".
You also need to add the roles that can use the bots various features, scroll down a bit to "Permissions", and check out the sample environment file below.
WELCOME_MESSAGE = "Hi There! Welcome to our Discord server. We hope you'll enjoy our server and we look forward to engaging with you!" # This is a fallback message if gpt3 fails to generate a welcome message.
USER_INPUT_API_KEYS="False" # If True, users must use their own API keys for OpenAI. If False, the bot will use the API key in the .env file.
# Moderations Service alert channel, this is where moderation alerts will be sent as a default if enabled
MODERATIONS_ALERT_CHANNEL = "977697652147892304"
# User API key db path configuration. This is where the user API keys will be stored.
As mentioned in the comments of the sample environment file, there are three permission groups that you can edit in the environment (`.env`) file. `ADMIN_ROLES` are roles that allow users to use `/system` and `/mod` commands. `GPT_ROLES` are roles that allow users to use `/gpt` commands, and `DALLE_ROLES` are roles that allow users to use `/dalle` commands.
If for a command group you want everybody to be able to use those commands, just don't include the relevant line in the `.env` file. For example, if you want everyone to be able to use GPT3 commands, you can just omit `the GPT_ROLES="...."` line.
First, you want to get a server, for this guide, I will be using DigitalOcean as the host.
For instructions on how to get a server from start to finish, they are available on DigitalOcean's website directly from the community, available here: https://www.digitalocean.com/community/tutorials/how-to-set-up-an-ubuntu-20-04-server-on-a-digitalocean-droplet. Ignore the part about setting up an "ssh key", and just use a password instead.
After you set up the server, the DigitalOcean GUI will give you an IP address, copy this IP address. Afterwards, you will need to SSH into the server. This can be done using a program such as "PuTTy", or by using your commandline, if it's supported. To login to the server, your username will be "root", your password will be the password that you defined earlier when setting up the droplet, and the IP address will be the IP address you copied after the droplet was finished creation.
To connect with ssh, run the following command in terminal:
`ssh root@{IP ADDRESS}`
It will then prompt you for your password, which you should enter, and then you will be logged in.
After login, we need to install the various dependencies that the bot needs. To do this, we will run the following commands:
# Copy the sample.env file into a regular .env file. `DEBUG_GUILD` and the ID for `ALLOWED_GUILDS` can be found by right-clicking your server and choosing "Copy ID". Similarly, `DEBUG_CHANNEL` can be found by right-clicking your debug channel.
This bot now supports having overrides be permanent in an entire conversation if you use an opener file which includes them. The new opener files should be .json files formatted like this. `text` corresponds to what you want the conversational opener to be and the rest map 1:1 to the appropriate model settings. An example .json file is included by the name of `english_translator.json` in the `openers` folder
- Make sure you have updated your .env file with valid values for `DEBUG_GUILD`, `DEBUG_CHANNEL` and `ALLOWED_GUILDS`, otherwise the bot will not work. Guild IDs can be found by right clicking a server and clicking `Copy ID`, similarly, channel IDs can be found by right clicking a channel and clicking `Copy ID`.
The bot has the ability to launch a HTTP endpoint at `<host>:8181/` that will return a json response of the bot's status and uptime. This is especially useful if you want to run this bot on cloud application containers, like Azure App Service.
To enable this, add `HEALTH_SERVICE_ENABLED="True"` to your `.env` file.
The health check endpoint will then be present in your bot's console when it is starting up, it will look like this, the possible HTTP urls for your health endpoint will be visible near the bottom:
