Quick Reference
One-page cheat sheet for DS01 commands. This covers the most common workflows; for the complete list of all available commands and options, see the Command Reference.
Daily Workflow
Project-Oriented (Recommended)
Familiar if you're used to local Python/Jupyter development.
# First time
user setup # Complete setup wizard
# New project
project init my-thesis --type=llm # Create project
project launch my-thesis --open # Start working
# Resume work
project launch my-thesis --open
# Done
exit
Container-Oriented (More Control)
Cloud-native style, closer to Docker/Kubernetes.
# Deploy container directly
container-deploy my-project --background
# Attach terminal
container-attach my-project
# Done
container-retire my-project
Help Commands
# List available aliases
commands # lists all commands
# Return to workspace
home # = cd /home/<user-id>/
Project Commands
# Create new project
project init [name] # Interactive if no name
project init my-project # Named project
project init my-project --type=cv # With template (ml/cv/nlp/rl/audio/ts/llm/custom)
project init --guided # With explanations
# Launch project (builds image if needed)
project launch [name] # Interactive if no name
project launch my-project # Named project
project launch my-project --open # Launch and enter terminal
project launch my-project --background # Start in background
project launch my-project --rebuild # Force image rebuild
Container Commands
# Create + start (orchestrator)
container-deploy [name] [image] # Interactive if no args
container-deploy my-project # Default base image
container-deploy my-project pytorch # Specify base image
container-deploy my-project --open # Create and enter
container-deploy my-project --background # Start in background
container-deploy my-project --cpu-only # No GPU
container-deploy my-project -w /path/to/dir # Custom workspace
container-deploy my-project --dry-run # Show what would happen
# Stop + remove + free GPU
container-retire [name] # Interactive if no name
container-retire my-project # Named container (prompts to save new packages)
container-retire my-project --force # Skip confirmation
# Status
container-list # Your containers
container-list --all # Include stopped
container-stats # Resource usage
# Individual steps (atomic - for advanced users)
container-create my-project # Create container (& allocate resources)
container-start my-project # Start in background
container-run my-project # Start + enter
container-attach my-project # Enter running container
container-pause my-project # Freeze processes (GPU stays allocated)
container-unpause my-project # Resume frozen container
container-stop my-project # Stop only
container-remove my-project # Remove only
Image Commands
image-create [name] # Interactive wizard
image-create my-project # Named image
image-create my-project -f pytorch # Specify framework (pytorch, tensorflow, jax)
image-list # Your images
image-update # Interactive GUI to add/remove packages
image-update my-project --rebuild # Rebuild after manual Dockerfile edit
image-delete my-project # Remove image
System Status
dashboard # System overview
dashboard gpu # GPU utilisation
dashboard cpu # CPU by user
dashboard --watch # Live monitoring (2s refresh)
dashboard --full # All sections expanded
check-limits # Your quotas and usage
Inside Container
# Check GPU
nvidia-smi
# Python with GPU
python
>>> import torch
>>> torch.cuda.is_available()
True
# Files location
/workspace/ # Your persistent files
File Locations
Host Container
---- ---------
~/workspace/my-project/ -> /workspace/
~/workspace/my-project/Dockerfile (image build source)
Getting Help
Every command supports 4 help modes:
| Flag | Type | Purpose |
|---|---|---|
--help, -h | Reference | Quick reference (usage, main options) |
--info | Reference | Full reference (all options, examples) |
--concepts | Education | Pre-run learning (what is X?) |
--guided | Education | Interactive learning (explanations during) |
Examples:
project init --concepts # Learn about projects before creating one
container-deploy --info # See all deploy options
image-create --guided # Step-by-step with explanations
Common Patterns
# Multiple experiments
container-deploy exp-1 --background
container-deploy exp-2 --background
# View logs
docker logs <project-name>._.$(whoami)
# Enter running container
container-attach my-project
# Recreate with fresh image
container-retire my-project
project launch my-project --rebuild
Troubleshooting
# Check status
container-list
dashboard
# View logs
docker logs <project-name>._.$(whoami)
# Recreate (fixes most issues)
container-retire my-project
project launch my-project
Alternative Syntax
Commands also work with spaces instead of hyphens:
container deploy my-project # Same as container-deploy
image create my-project # Same as image-create
project init my-project # Same as project-init
See Dispatcher Commands for details.
Docker Commands (Advanced Users)
If you prefer working directly with Docker, these commands are still subject to DS01 resource limits and GPU allocation:
# List containers
docker ps # Running containers
docker ps -a # All containers (including stopped)
# Inspect containers
docker logs <container-id> # View container output
docker logs -f <container-id> # Follow logs (live)
docker inspect <container-id> # Full container details (JSON)
# Execute commands
docker exec <container-id> ls /workspace # Run command in container
docker exec -it <container-id> bash # Interactive shell
# Manage images
docker images # List images
docker build -t my-image . # Build image from Dockerfile
docker rmi <image-id> # Remove image
# Resource info
docker stats # CPU, memory, GPU usage
docker stats --no-stream # Single snapshot
Note: All Docker commands run through DS01's enforcement system:
- Resource limits (CPU, memory, GPU) are enforced via cgroups
- GPU allocations are tracked and validated
- Containers are organized in DS01 systemd slices
Detailed docs: See Command Reference | Troubleshooting