This project is a FastAPI-based application that generates AI-generated responses using the Poe API Wrapper (https://github.com/ading2210/poe-api).
The use of the Poe Server and the api wrapper is at your own risk. We are not responsible for any damages or losses that may occur as a result of your use of the server or API. This includes, but is not limited to, damage to your computer, loss of data, or violation of the Terms of Service (TOS) for Poe (Quora).
We do not endorse or promote the use of the Poe server or API wrapper in any way that violates the TOS. If you use the server or API in a way that violates the TOS, you may be subject to disciplinary action by Quora .This may include, but is not limited to, loss of your account or proxy provider.
Please use the server and API responsibly and in accordance with the TOS for the server and API. If you have any questions or concerns, please contact the developers of the server or API by opening an issue on this Github repo.
Due to recent changes, our project may experience occasional downtimes linked to issues with a crucial API wrapper it depends on. This could lead to periods where the project does not function as expected.
Thankfully, you don't have to be in the dark about this. We're actively tracking these potential disruptions and will immediately update the operational status of the API wrapper on this README. This way, you can quickly check here to know if our project is currently up and running or facing some downtime.
We appreciate your understanding during this time. Rest assured, we're working diligently to ensure consistent functionality. Remember to check this space for the latest status updates.
Thank you for your patience.
last checked version:
poe-api 0.4.17
last time last checked:
03:39 AM (UTC +0)
- Token rotational
- Proxy rotational
- Add Streaming Support and make it compatible with Bettergpt.chat (Much thanks PaniniCo!!)
- Add OPEN AI format of api like (/v1/chat/completions) use serverof.py for this
- Work on Stability so not overloaded easily (rate limiting added)
- Add a retry logic and switch to a different token if a model has reached daily limit
- Add Anse Support (https://github.com/anse-app/anse)
- Add One-Api Support (https://github.com/songquanpeng/one-api)
- Add Embeddings to OPEN AI format (/api/v1/embedding)
- Add AI.LS Support (https://ai.ls/)
To install the project dependencies, run the following command:
pip install -r requirements.txt
- Clone the repository:
git clone https://github.com/greengeckowizard/poe-server
cd poe-server
- Install the dependencies:
pip install -r requirements.txt
- Define Proxy and Tokens inside the proxies.txt for the proxies and for the tokens put inside the tokens.txt file (Proxy and Tokens are required to operate this Project WE DO NOT PROVIDE ANY PROXIES OR TOKENS UNDER ANY CIRCUMSTANCES)
if you don't define a proxy or token you will have errors. Note:Any of the proxies that work inside the poe-api will work in this project.
- Running the FastAPI Server Locally
You have two options for running the server: Using Python directly, or using Docker.
To run the server with Python, execute the following command in your shell:
python server.py
another
To run the server with Docker, you need to build and start the Docker container. Run the following commands:
CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "8000"]
then rebuild the dockerfile and you can docker-compose up. For running in Daemon mode use -d flag docker-compose up -d
docker-compose build
docker-compose up
If you want the server to run in the background (daemonized), add the -d
flag:
docker-compose up -d
After running these commands, the server will start and listen on the following URL:
http://localhost:8000
Important: When interacting with the server locally via Docker, it's recommended to use the IP address assigned to your device instead of
http://localhost:8000
. For instance, if your local device's IP is192.168.0.45
, the appropriate address to use would behttp://192.168.0.45:8000
.
To find your local IP address, run the following command:
-
On Linux:
ifconfig
-
On Windows:
ipconfig
-
On MacOS:
ifconfig | grep "inet " | grep -v 127.0.0.1
This command will display a list of all network interfaces. The grep "inet "
command filters this list to include only entries that have an IP address (inet stands for internet). The second grep -v 127.0.0.1
command excludes the loopback address, which is typically 127.0.0.1. The IP address you're likely interested in will most likely be the one associated with en0
or en1
- these are usually the primary Ethernet and Wi-Fi interfaces, respectively.
Remember, IP addresses that start with 192.168 or 10.0 are local addresses - they're only valid on your local network. If you're trying to give someone your IP address to connect to over the internet, you'll need your public IP address, which you can find by searching 'what is my IP' on Google or any other search engine.
"assistant": "capybara",
"claude-instant": "a2",
"claude-2-100k": "a2_2",
"claude-instant-100k": "a2_100k",
"gpt-3.5-turbo-0613": "chinchilla",
"gpt-3.5-turbo": "chinchilla",
"gpt-3.5-turbo-16k-0613": "agouti",
"gpt-3.5-turbo-16k": "agouti",
"gpt-4": "beaver",
"gpt-4-0613": "beaver",
"gpt-4-32k": "vizcacha",
"chat-bison-001": "acouchy",
"llama-2-70b": "llama_2_70b_chat",
You can make a request like this use the server.py to use this type request
import openai
openai.api_key = "anyrandomstring"
openai.api_base = "http://localhost:8000/v1"
response = openai.ChatCompletion.create(
model='gpt-3.5-turbo',
messages=[
{'role': 'user', 'content': "Hello"},
]
)
print(response)
You can configure the project by providing the following (set inside the text files):
- Inside the
tokens.txt
: A new line for each Poe API token. - Inside the
proxies.txt
: A new line for each proxy server.
If you want to contribute to this project, please follow these steps:
- Fork the repository.
- Create a new branch.
- Make your changes.
- Test your changes.
- Commit your changes.
- Push your changes to your forked repository.
- Create a pull request.
This project uses the [GNU General Public License v3.0] see the LICENSE file for more details.
Thank you to PaniniCo for adding streaming support and model corresponding to names to Poe Models!!
Thank you UltGamer for adding better stability with checking bad tokens and increasing better load capacity support!!