> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usechamber.io/llms.txt
> Use this file to discover all available pages before exploring further.

# CLI Overview

> Submit GPU workloads in seconds — no Docker or Kubernetes expertise required

![Build](https://github.com/ChamberOrg/chamber-cli/actions/workflows/test.yml/badge.svg?branch=main)
![Coverage](https://codecov.io/gh/ChamberOrg/chamber-cli/branch/main/graph/badge.svg)

# Chamber CLI

**Go from Python code to running GPU workload in one command.** The Chamber CLI eliminates the complexity of containerization, registry management, and Kubernetes — so ML engineers and data scientists can focus on what matters: training models.

<Info>
  **No Docker expertise required.** Chamber auto-detects your project, generates optimized Dockerfiles, handles registry authentication, and submits your workload — all interactively guided.
</Info>

## AI Assistant — Ask Your Infrastructure Anything

Query your GPU infrastructure in natural language, directly from the terminal:

```bash theme={null}
chamber chat "how many GPUs are running right now?"
```

Start an interactive conversation for back-and-forth troubleshooting, capacity planning, or workload analysis:

```bash theme={null}
chamber chat
```

<CardGroup cols={2}>
  <Card title="AI Assistant Guide" icon="message-bot" href="/cli/chat">
    Interactive conversations, command execution, piped input, image analysis, and more
  </Card>

  <Card title="Auto-Containerize Guide" icon="wand-magic-sparkles" href="/cli/run">
    Deep dive into `chamber run`
  </Card>
</CardGroup>

## The Fastest Path to GPU Training

```bash theme={null}
chamber run ./my-training-project --gpus 4 --team my-team
```

That's it. One command. Chamber handles everything else:

<Steps>
  <Step title="Detects your project">
    Automatically identifies PyTorch, TensorFlow, or JAX. Finds your entrypoint and requirements.
  </Step>

  <Step title="Generates optimized Dockerfile">
    Creates a GPU-optimized container with CUDA, cuDNN, and your dependencies.
  </Step>

  <Step title="Guides you through setup">
    Missing Docker? No registry configured? Chamber walks you through each step interactively.
  </Step>

  <Step title="Builds, pushes, and submits">
    Handles authentication, builds your image, pushes to your registry, and submits to Chamber.
  </Step>
</Steps>

<Card title="Quick Start" icon="rocket" href="/cli/installation">
  Install and run your first workload
</Card>

## Interactive Setup — No Prior Configuration Needed

Chamber CLI guides you through everything. Don't have Docker? It'll help you install it. No registry configured? It'll walk you through setting one up:

```
$ chamber run ./my-training --gpus 4 --team abc123

No container registry configured

Chamber needs a container registry to store your Docker images.
You can use Google Artifact Registry, AWS ECR, or any Docker-compatible registry.

Select your registry type:

  [1] Google Artifact Registry (recommended for GCP users)
      Example: us-central1-docker.pkg.dev/my-project/ml-images

  [2] AWS ECR (recommended for AWS users)
      Example: 123456789012.dkr.ecr.us-east-1.amazonaws.com/ml-images

  [3] Other Docker registry

Select an option [1]: 1

Google Artifact Registry Setup

Enter your GAR registry URL: us-central1-docker.pkg.dev/my-project/ml-images

Save as default registry? (y/n) [y]: y
Default registry saved to ~/.chamber/config.json

[detect] Analyzing project structure...
[detect] Framework: PyTorch
[detect] Entrypoint: train.py
[setup] Checking prerequisites...
```

<Tip>
  **First-time users:** Just run `chamber run` on your project. The CLI will guide you through every step — from installing Docker to configuring your registry to submitting your workload.
</Tip>

## Automatic Prerequisite Detection

Missing a tool? Chamber detects it and offers to help:

```
$ chamber run ./my-project --gpus 4 --team abc123

⚠ Docker is not installed
  Docker is required to build and push container images.

Installation options:

  [1] Quick install (recommended)
      brew install --cask docker
  [2] Open installation guide in browser
  [3] Show manual installation instructions
  [4] Skip (continue anyway)

Select an option [1]:
```

The same guided experience applies to:

* **Docker** — Required for building images
* **AWS CLI** — For ECR authentication
* **gcloud CLI** — For Google Artifact Registry authentication
* **Azure CLI** — For ACR authentication

## Why ML Engineers Love Chamber CLI

<CardGroup cols={3}>
  <Card title="Zero Config Start" icon="play">
    Run `chamber run` and follow the prompts. No YAML files, no Docker knowledge needed.
  </Card>

  <Card title="Smart Defaults" icon="brain">
    Auto-detects frameworks, entrypoints, and optimal GPU configurations.
  </Card>

  <Card title="One-Time Setup" icon="check">
    Configure once, run forever. Settings are saved for future use.
  </Card>
</CardGroup>

<CardGroup cols={3}>
  <Card title="Preview First" icon="eye">
    Use `--dry-run` to see exactly what will be generated before building.
  </Card>

  <Card title="Works Everywhere" icon="server">
    Single binary with no dependencies. SSH-friendly for remote workstations.
  </Card>

  <Card title="Scriptable" icon="code">
    JSON output mode for CI/CD pipelines and automation.
  </Card>
</CardGroup>

## Quick Command Reference

| What you want to do            | Command                                         |
| ------------------------------ | ----------------------------------------------- |
| **Ask a question**             | `chamber chat "what GPUs are available?"`       |
| **Start a conversation**       | `chamber chat`                                  |
| **Submit a training workload** | `chamber run ./my-project --gpus 4 --team <id>` |
| Preview before submitting      | `chamber run ./my-project --gpus 4 --dry-run`   |
| List your workloads            | `chamber workloads list`                        |
| Check GPU capacity             | `chamber capacity`                              |
| View workload status           | `chamber workloads get <id>`                    |
| Cancel a workload              | `chamber workloads cancel <id>`                 |

## System Requirements

* **macOS** (Intel or Apple Silicon) or **Linux** (x86\_64 or ARM64)
* A Chamber account

**For auto-containerization (`chamber run`):**

* Docker — *Chamber will help you install it*
* Cloud CLI (gcloud/aws/az) — *Chamber will help you install it*

<Card title="Get Started" icon="arrow-right" href="/cli/installation">
  Install Chamber CLI and run your first workload
</Card>