workcell 0.0.29

Turn python function into microservice.

Workcell

Instantly turn your Python function into production-ready microservice.

English | 简体中文

Getting Started • License • Changelog

Instantly turn your Python function into delightful app and production-ready microservice, with lightweight UI to interact with.

---

Highlights

• 🔌 Instantly turn functions into microservices within seconds.
• 📈 Automatically generate user-friendly UI for interaction.
• 🤗 One-click deployment to Hugging Face Spaces.
• ☁️ Develop locally, deploy to the cloud.
• 🧩 Empower development and analysis with scalable components.
• 🦄 Get inspired by the open-source community.

Status

Status	Stability	Goal
✅	Alpha	We are testing Workcell with a closed set of customers
🚧	Public Alpha	Anyone can sign up over at weanalyze.co. But go easy on us, there are a few kinks.
❌	Public Beta	Stable enough for most non-enterprise use-cases
❌	Public	Production-ready

We are currently in: Alpha.

Requirements

Python 3.8+

Installation

To get started with Workcell, you can install it using pip:

Recomended: First activate your virtual environment, with your favourite system. For example, we like poetry and conda!

pip install workcell

Getting Started

Here is an example of a simple Workcell-compatible function:

from pydantic import BaseModel

class Input(BaseModel):
    message: str

class Output(BaseModel):
    message: str

def hello_workcell(input: Input) -> Output:
    """Returns the `message` of the input data."""
    return Output(message=input.message)

💡Note: A workcell-compatible function must have an input parameter and return value based on Pydantic models. The input and output models are specified using type hints.

To start a Workcell app, follow these steps:

1. Copy the above code to a file named app.py.

2. Create a folder, e.g. hello_workcell, and place the app.py inside.

3. Open your terminal and navigate to the folder hello_workcell.

4. Start the Workcell app using the following command:

workcell serve app:hello_workcell

💡Note: The output will display the location where the API is being served on your local machine.

Alternatively, you can import the workcell package and serve your function using an ASGI web server such as Uvicorn or FastAPI. Simply wrap your function with workcell.create_app like this:

from pydantic import BaseModel
import workcell

class Input(BaseModel):
    message: str

class Output(BaseModel):
    message: str

def hello_workcell(input: Input) -> Output:
    """Returns the `message` of the input data."""
    return Output(message=input.message)

app = workcell.create_app(hello_workcell)

Finally, run the app using the following command:

uvicorn app:app --host 0.0.0.0 --port 7860

💡Note: The output will display the location where the API is being served on your local machine.

Workcell deployment

🤗 You can deploy your workcell to Hugging Face Spaces in 1-click! You'll be able to access your workcell from anywhere and share it with your team and collaborators.

Prepare your Hugging Face account

First you need a Hugging Face account, and prepare your Hugging Face username and User Access Tokens, then set environment variables like below:

export HUGGINGFACE_USERNAME={huggingface_username}
export HUGGINGFACE_TOKEN={hf_XXX}

Replace {huggingface_username} with your actual Hugging Face username, and {hf_XXX} with your actual User Access Token. You can also store these environment variables in a .env file in your project folder for convenience.

Deploy Workcell

1. Wrap your function with workcell.create_app like example above.

2. In your project folder, package your Workcell app using the following command:

workcell pack app:hello_workcell

💡Note: workcell pack {file_name}:{create_app_name} will package your function code with a Dockerfile template into .workcell folder in your project folder.

3. Once packaged, deploy your Workcell app using the following command:

workcell deploy

Voila! The deployment process will start, and within a few minutes, your workcell will be available on Hugging Face Spaces, accessible by a unique URL.

More details

You can monitor the deployment process and the logs in your terminal, and the deployment status will be shown in your Hugging Face Spaces repo.

You can deploy multiple workcells, and they will be listed in your Hugging Face Spaces account, you can manage and remove them from there.

You can also configure various deployment options like environment variables, system requirements, custom domain, etc., by using command line options or a workcell.yaml from .workcell dir in your project folder.

Please stay tuned, as a comprehensive guide will be available soon to provide further explanation.

Examples

🎮 Get inspired and learn more about Workcell by exploring our examples:

• https://github.com/weanalyze/workcell-examples

🏆 We also have a curated list for you to check out, feel free to contribute!

• https://github.com/weanalyze/best-of-workcell

Roadmap

🗓️ Missing a feature? Have a look at our public roadmap to see what the team is working on in the short and medium term. Still missing it? Please let us know by opening an issue!

Contacts

❓ If you have any questions about the workcell or weanalyze , feel free to email us at: [e-mail]

🙋🏻 If you want to say hi, or are interested in partnering with us, feel free to reach us at: [e-mail]

😆 Feel free to share memes or any questions at Discord: https://discord.weanalyze.co

License

Apache-2.0 License.