### Setup Beta9 SDK

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

Run this command to set up the Beta9 SDK. Ensure you have Python 3.8 or higher installed.

```bash
make setup-sdk
```

--------------------------------

### Setup On-Prem Agent

Source: https://github.com/beam-cloud/beta9/blob/main/docs/opa/README.md

Create a machine within a worker pool and obtain a setup command to install the agent. This command registers the machine with Beam's control plane and joins the Tailscale network.

```sh
$ beam machine create --pool my-h100-pool

=> Created machine with ID: 'c11d3030'. Use the following command to setup the node:
# -- Agent setup
sudo curl -L -o agent https://release.beam.cloud/agent/agent && \
sudo chmod +x agent && \
sudo ./agent --token "MY_TOKEN" \
    --machine-id "c11d3030" \
    --tailscale-url "" \
    --tailscale-auth "tskey-auth" \
    --pool-name "my-h100-pool" \
    --provider-name "generic"
```

--------------------------------

### Setup Beta9 Server Locally

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

Run this command to set up the Beta9 server for local development using k3d. Ensure Docker is installed.

```bash
make setup
```

--------------------------------

### Install Beam Client

Source: https://github.com/beam-cloud/beta9/blob/main/README.md

Install the Beam client library using pip. This is the first step to using Beam for your AI workloads.

```shell
pip install beam-client
```

--------------------------------

### Install Dependencies with Poetry

Source: https://github.com/beam-cloud/beta9/blob/main/sdk/bin/docgen.md

Install all project dependencies, including development dependencies, using Poetry.

```sh
poetry install
```

--------------------------------

### Monitor Setup Logs via SSH

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/crusoe/README.md

Connect to your deployed instance via SSH to monitor the user-data log in real-time. Replace `<instance-ip>` with the actual IP address of your instance.

```sh
ssh -i ~/.ssh/id_crusoecloud ubuntu@<instance-ip> 'tail -n+1 -f /var/log/user-data.log'
```

--------------------------------

### Fix Permissions for /usr/local/bin

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

If you encounter permission errors during setup, use this command to change the ownership of the /usr/local/bin directory to your current user.

```bash
sudo chown $(whoami) /usr/local/bin
```

--------------------------------

### Configure Local DNS with Dnsmasq

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

Install and configure dnsmasq for local DNS resolution, required for external file services and optional for subdomain middleware. This setup directs 'cluster.local' domain traffic to 127.0.0.1.

```shell
brew install dnsmasq
echo 'address=/cluster.local/127.0.0.1' >> /opt/homebrew/etc/dnsmasq.conf
sudo bash -c 'mkdir -p /etc/resolver'
sudo bash -c 'echo "nameserver 127.0.0.1" > /etc/resolver/cluster.local'
sudo brew services start dnsmasq
```

--------------------------------

### Deploy Serverless Inference Endpoint

Source: https://github.com/beam-cloud/beta9/blob/main/README.md

Create an autoscaling endpoint for your custom AI model. This example configures the image, GPU, CPU, memory, and autoscaling parameters.

```python
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 with Beam

Source: https://github.com/beam-cloud/beta9/blob/main/README.md

Schedule resilient background tasks using a simple decorator. This example defines input schema, task policy, and demonstrates invoking a background task.

```python
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
```

--------------------------------

### List Machines

Source: https://github.com/beam-cloud/beta9/blob/main/docs/opa/README.md

View all registered machines in your inventory, including their hardware specifications, status, and assigned worker pool. This helps in verifying the setup and availability of your on-prem resources.

```sh
$ beam machine list

| ID       | CPU     | Memory     | GPU     | Status     | Pool         |
|----------|---------|------------|---------|------------|--------------|
| edc9c2d2 | 30,000m | 222.16 GiB | H100    | registered | my-h100-pool |
```

--------------------------------

### Launch and Deploy fly App

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/fly-io/README.md

Initiate the deployment of the fly app, which includes copying configuration and prompting for confirmation. This command modifies the fly.toml file in-place.

```shell
fly launch --org my-org --name my-deployment --no-deploy --copy-config --yes
```

--------------------------------

### Create a Sandbox Environment

Source: https://github.com/beam-cloud/beta9/blob/main/README.md

Spin up isolated containers to run LLM-generated code. This snippet demonstrates creating a sandbox and running code within it.

```python
from beam import Image, Sandbox


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

print(response.result)
```

--------------------------------

### Initialize and Apply Terraform Configuration

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/crusoe/README.md

Run these commands to initialize your Terraform working directory and apply the configuration to deploy resources on Crusoe Cloud.

```sh
terraform init
```

```sh
terraform apply
```

--------------------------------

### Generate All Protocol Buffers and OpenAPI Schemas

Source: https://github.com/beam-cloud/beta9/blob/main/docs/openapi/README.md

Execute this command to regenerate all protocol buffers, including the OpenAPI schemas.

```bash
make protocol
```

--------------------------------

### Import Secrets and Deploy Application

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/fly-io/README.md

Import the previously prepared secrets into the fly.io environment, deploy the application, and scale it to a single instance in the 'iad' region. Ensure the max instances per region is set to 1.

```shell
cat .env | fly secrets import
```

```shell
fly deploy
```

```shell
fly scale count 1 --max-per-region=1 --region=iad
```

--------------------------------

### Generate SDK Reference Markdown

Source: https://github.com/beam-cloud/beta9/blob/main/sdk/bin/docgen.md

Run the docgen script to generate markdown for the SDK Reference documentation. On macOS, the output can be piped to pbcopy for immediate use.

```sh
# Generates markdown
./bin/docgen.py
```

```sh
# Generates markdown and copies to your clipboard (on macOS)
./bin/docgen.py | pbcopy
```

--------------------------------

### List Available Crusoe Compute Instance Types

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/crusoe/README.md

Use this CLI command to view the available instance types for deployment on Crusoe Cloud. This is useful for selecting the appropriate `instance_type` input.

```sh
crusoe compute vms types
```

--------------------------------

### Generate OpenAPI Schemas

Source: https://github.com/beam-cloud/beta9/blob/main/docs/openapi/README.md

Run this command to regenerate the OpenAPI schemas from the proto files.

```bash
make openapi
```

--------------------------------

### Define and Run a Function with Beta9 SDK

Source: https://github.com/beam-cloud/beta9/blob/main/sdk/README.md

Define a Python function decorated with `@function` to run on Beta9. Use `.map()` to invoke the function multiple times with different inputs. Ensure the `main` function is called when the script is executed directly.

```python
from beta9 import function

@function(
    cpu=0.25,
    gpu="L4",
)
def square(i: int):
    return i**2

def main():
    numbers = list(range(10))
    squared = []

    for result in square.map(numbers):
        squared.append(result)

    print("numbers:", numbers)
    print("squared:", sorted(squared))

if __name__ == "__main__":
    main()

```

--------------------------------

### Prepare Secrets for JuiceFS S3 Gateway

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/fly-io/README.md

Create a .env file to store secrets required by the JuiceFS S3 Gateway application. Ensure all sensitive information is correctly formatted.

```shell
cat <<'EOF' | tee > .env
TAILSCALE_AUTHKEY="my-tailscale-authkey"
JUICEFS_NAME="my-filesystem"
JUICEFS_REDIS_URI="rediss://username:password@hostname:6379/0"
JUICEFS_BUCKET="https://fly.storage.tigris.dev/my-bucket-name"
JUICEFS_ACCESS_KEY="my-access-key"
JUICEFS_SECRET_KEY="my-secret-key"
MINIO_ROOT_USER="my-username"
MINIO_ROOT_PASSWORD="my-password"
EOF
```

--------------------------------

### Configure Beta9 Gateway for Host-Based URLs

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

Add this configuration to your Beta9 settings to enable host-based URLs, which is necessary for using subdomain middleware.

```yaml
gateway:
  invokeURLType: host
```

--------------------------------

### Create a Worker Pool

Source: https://github.com/beam-cloud/beta9/blob/main/docs/opa/README.md

Use the CLI to create a new worker pool with specified attributes like GPU type and count. This defines a group of workers for specific tasks.

```sh
beam pool create --name my-h100-pool --gpu-type A100 --default-gpu-count 1
```

--------------------------------

### Beta9 Örnek Uygulama

Source: https://github.com/beam-cloud/beta9/blob/main/docs/tr/README.md

Bir tamsayıyı karekökleyen bir fonksiyon tanımlar ve bunu bir listedeki her öğe için uzak bir kapsayıcıda çalıştırır.

```python
from beta9 import function


@function(cpu=8)
def square(i: int):
    return i**2


def main():
    numbers = list(range(10))
    squared = []

    # Listedeki her öğe için uzak bir kapsayıcı çalıştır
    for result in square.map(numbers):
        squared.append(result)
```

--------------------------------

### List Available Crusoe Cloud Locations

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/crusoe/README.md

Execute this command to see the list of available deployment locations on Crusoe Cloud. This helps in setting the `location` input for your Terraform configuration.

```sh
crusoe locations list
```

--------------------------------

### Run Function on Worker Pool

Source: https://github.com/beam-cloud/beta9/blob/main/docs/opa/README.md

Define and run a Python function that will be executed on your on-prem worker pool. Specify hardware requirements like GPU type using the `@function` decorator.

```python
from beam import function

@function(gpu="H100")
def handler():
    return {"message": "This is running on your worker pool!"}

if __name__ == "__main__":
    handler.remote()
```

--------------------------------

### Terraform Variables for Crusoe Cloud Deployment

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/crusoe/README.md

Define the path to your SSH public key for instance access. Ensure this key is generated or available before applying Terraform.

```hcl
ssh_key_path  = "~/.ssh/id_crusoecloud.pub"
```

--------------------------------

### Remote Function Execution with Beta9

Source: https://github.com/beam-cloud/beta9/blob/main/docs/it/README.md

This Python code defines a remote function 'square' that computes the square of an integer and then maps this function over a list of numbers, executing each call in a remote container.

```python
from beta9 import function


@function(cpu=8)
def square(i: int):
    return i**2


def main():
    numbers = list(range(10))
    squared = []

    # Esegui un contenitore remoto per ogni elemento nell'elenco
    for result in square.map(numbers):
        squared.append(result)
```

--------------------------------

### Destroy fly.io Deployment

Source: https://github.com/beam-cloud/beta9/blob/main/deploy/fly-io/README.md

Remove the deployed application from fly.io. This command will prompt for confirmation before proceeding with the destruction.

```shell
fly app destroy --yes my-deployment
```

--------------------------------

### Check Beta9 Gateway Health

Source: https://github.com/beam-cloud/beta9/blob/main/CONTRIBUTING.md

Use curl to check the health of your local k3s instance via the Beta9 gateway using its internal service domain.

```shell
curl http://beta9-gateway.beta9.svc.cluster.local:1994/api/v1/health
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.