GitHub - beam-cloud/beta9: Ultrafast serverless GPU inference, sandboxes, and background jobs

2 min read Original article ↗

Beam is a fast, open-source runtime for serverless AI workloads. It gives you a Pythonic interface to deploy and scale AI applications with zero infrastructure overhead.

Watch the demo

✨ Features

  • Fast Image Builds: Launch containers in under a second using a custom container runtime
  • Parallelization and Concurrency: Fan out workloads to 100s of containers
  • First-Class Developer Experience: Hot-reloading, webhooks, and scheduled jobs
  • Scale-to-Zero: Workloads are serverless by default
  • Volume Storage: Mount distributed storage volumes
  • GPU Support: Run on our cloud (4090s, H100s, and more) or bring your own GPUs

📦 Installation

⚡️ Quickstart

  1. Create an account here
  2. Follow our Getting Started Guide

Creating a sandbox

Spin up isolated containers to run LLM-generated code:

from beam import Image, Sandbox


sandbox = Sandbox(image=Image()).create()
response = sandbox.process.run_code("print('I am running remotely')")

print(response.result)

Deploy a serverless inference endpoint

Create an autoscaling endpoint for your custom model:

from beam import Image, endpoint
from beam import QueueDepthAutoscaler

@endpoint(
    image=Image(python_version="python3.11"),
    gpu="A10G",
    cpu=2,
    memory="16Gi",
    autoscaler=QueueDepthAutoscaler(max_containers=5, tasks_per_container=30)
)
def handler():
    return {"label": "cat", "confidence": 0.97}

Run background tasks

Schedule resilient background tasks (or replace your Celery queue) by adding a simple decorator:

from beam import Image, TaskPolicy, schema, task_queue


class Input(schema.Schema):
    image_url = schema.String()


@task_queue(
    name="image-processor",
    image=Image(python_version="python3.11"),
    cpu=1,
    memory=1024,
    inputs=Input,
    task_policy=TaskPolicy(max_retries=3),
)
def my_background_task(input: Input, *, context):
    image_url = input.image_url
    print(f"Processing image: {image_url}")
    return {"image_url": image_url}


if __name__ == "__main__":
    # Invoke a background task from your app (without deploying it)
    my_background_task.put(image_url="https://example.com/image.jpg")

    # You can also deploy this behind a versioned endpoint with:
    # beam deploy app.py:my_background_task --name image-processor

Self-Hosting vs Cloud

Beta9 is the open-source engine powering Beam, our fully-managed cloud platform. You can self-host Beta9 for free or choose managed cloud hosting through Beam.

👋 Contributing

We welcome contributions big or small. These are the most helpful things for us:

❤️ Thanks to Our Contributors