Hands-on lesson

Run Your First Local AI Model

Run Ollama in Docker, verify its local API, and generate your first response with llama3.2:3b.

14 minFree Core9 Hands-On AI Lessonsbeginner

What you'll build

Official Ollama running in Docker

llama3.2:3b available from the retained model volume

A real local model response

Docker

compose up -d

→

Ollama container

enjyra-v6-local-ai-02

→

API :11434

/api/version

→

Model

llama3.2:3b

→

Response

local answer

docker stop

container exits

→

curl --fail

connection failed

→

docker ps -a

Exited

→

docker start

API responds

Check

Before You Build

2 quick questions. This is diagnostic only — it does not block the lesson.

1 / 4Core

Where are model weights stored in this lesson?

The official runtime uses a persistent Docker volume.

2 / 4Core

What does /api/version verify?

A JSON response proves the local service is listening.

understand

build

break

debug

improve

deploy

Mission

Run Ollama in Docker, verify its local API, and generate your first response with llama3.2:3b.

Understand

Ollama is the model runtime. It stays in its own official container and stores model data in a persistent volume; ENJYRA applications call its HTTP API without embedding model weights.

Build

What you're doing

Create compose.yaml, paste the service definition, save it, and verify the saved file.

Why it matters

Compose records the exact image, container name, local port, and persistent model volume.

Create, open, paste, and save~/enjyra-free-ai/compose.yaml

Where to work VS Code FileLanguage yaml

Create and open the fileRun this from ~/enjyra-free-ai.

touch compose.yaml
code compose.yaml

touch compose.yaml
code compose.yaml

if (-not (Test-Path "compose.yaml")) { New-Item -ItemType File -Path "compose.yaml" | Out-Null }
code "compose.yaml"

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Visual Studio Code is installed; the optional code shell command is available, or you can open the file manually.

Why You're Running This

compose.yaml is the project’s yaml file at ~/enjyra-free-ai/compose.yaml. It must exist before you can paste the supplied content. The commands create its folder when needed, create the file, and open it in VS Code; the next stage tells you what to paste and reminds you to save it.

Command Breakdown

touch / New-Item File: Creates an empty file at the requested path without requiring code to exist first.
code: Opens the file or folder in Visual Studio Code from the terminal.

Expected Result

compose.yaml

compose.yaml exists and is ready for the supplied content.

What Changed

Beforecompose.yaml may not exist yet, or may still contain its previous content.

Aftercompose.yaml now exists in the lesson workspace and is ready for the next instruction.

Verify It

macOSls -l compose.yaml

Linuxls -l compose.yaml

Windows PowerShellGet-Item "compose.yaml"

Confirm the file exists at the expected path.

If It Fails

Check that the terminal is in the lesson project, the path spelling is exact, and VS Code is installed. If the code command is unavailable, use File → Open Folder in VS Code.

If code is unavailable, open VS Code, choose File → Open Folder, select enjyra-free-ai, then create compose.yaml in the Explorer.

Paste this contentCopy the complete block into compose.yaml.

services:
  ollama:
    image: ollama/ollama@sha256:57d60e686821ea81a7748a3ec8141308c8b8f95b27105713954abf7a6529e700
    container_name: enjyra-v6-local-ai-02
    ports:
      - "127.0.0.1:11434:11434"
    volumes:
      - docker_ollama_data:/root/.ollama
volumes:
  docker_ollama_data:
    external: true

Save before continuing: press Cmd + S on macOS or Ctrl + S on Linux and Windows.

Verify the saved fileThe command should print the content you just saved.

cat compose.yaml

cat compose.yaml

Get-Content "compose.yaml"

Before You Run

Your terminal is open in ~/enjyra-free-ai.

Why You're Running This

Reading compose.yaml back from disk proves that the yaml content was saved at the correct path, rather than remaining only in an unsaved editor tab.

Command Breakdown

cat / Get-Content: Reads the saved file and prints its contents so you can verify it.

Expected Result

The terminal prints the complete content you pasted into compose.yaml.

What Changed

BeforeThe current state has not yet been checked.

AfterNo persistent state changed; the command printed evidence you can compare with the expected result.

Verify It

macOScat compose.yaml

Linuxcat compose.yaml

Windows PowerShellGet-Content "compose.yaml"

Run the command and compare its output with the Expected Result block.

If It Fails

Save the file, confirm the filename and capitalisation, and check the current folder with pwd on macOS/Linux or Get-Location on Windows.

If it fails

Check that your terminal is in ~/enjyra-free-ai, the filename and capitalisation match exactly, and the file was saved. Use pwd on macOS/Linux or Get-Location on Windows to check your current folder.

What changed

compose.yaml now exists in the project root.

What you're doing

Start Ollama from the saved Compose file.

Why it matters

The remaining checks need a running model API.

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory ~/enjyra-free-ai

What to run

cd ~/enjyra-free-ai && docker compose up -d

cd ~/enjyra-free-ai && docker compose up -d

Set-Location $HOME\enjyra-free-ai
docker compose up -d

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Docker Desktop or Docker Engine is running.

Why You're Running This

The service must be running before the lesson can connect to or test it.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
docker compose up: Creates and starts the services defined in the Compose file; -d keeps them running in the background.
&&: Runs the next command only when the previous command succeeds.

Expected Result

Container enjyra-v6-local-ai-02 started

Docker Desktop shows the friendly container name.

What Changed

BeforeThe requested local service or lab containers are not yet confirmed running.

AfterThe local service or lab containers are running in the background and can be verified.

Verify It

macOSdocker ps

Linuxdocker ps

Windows PowerShelldocker ps

Confirm the expected container is listed as running or healthy.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

Container enjyra-v6-local-ai-02 started

Success looks like: Docker Desktop shows the friendly container name.

If it fails

command not found → verify the required tool is installed and reopen the terminal.
No such file or directory → check your current folder with pwd or Get-Location.
Permission denied → confirm the lesson workspace is writable.

What changed

The enjyra-v6-local-ai-02 container is running.

What you're doing

Check the HTTP endpoint and confirm the retained model is available.

Why it matters

Both the runtime and the model must be healthy before inference.

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

curl -s http://localhost:11434/api/version

curl -s http://localhost:11434/api/version

curl.exe -s http://localhost:11434/api/version

Before You Run

Your terminal is open in Current lesson workspace.
The local service named in the URL is already running.

Why You're Running This

This command completes the “this step” step and produces the state needed by the next instruction.

Command Breakdown

curl: Makes an HTTP request to the local service and prints its response.

Expected Result

{"version":"<installed-version>"}

The local Ollama HTTP endpoint returns JSON.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.

What you should see

{"version":"<installed-version>"}

Success looks like: The local Ollama HTTP endpoint returns JSON.

If it fails

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

docker exec enjyra-v6-local-ai-02 ollama list

docker exec enjyra-v6-local-ai-02 ollama list

docker exec enjyra-v6-local-ai-02 ollama list

Before You Run

Your terminal is open in Current lesson workspace.
Docker Desktop or Docker Engine is running.

Why You're Running This

This checks the current state without changing it, so you can verify the previous action worked.

Command Breakdown

docker exec: Inspects or interacts with local Docker state without using a cloud service.

Expected Result

llama3.2:3b

The retained model appears without a new download.

What Changed

BeforeThe current state has not yet been checked.

AfterNo persistent state changed; the command printed evidence you can compare with the expected result.

Verify It

macOSdocker exec enjyra-v6-local-ai-02 ollama list

Linuxdocker exec enjyra-v6-local-ai-02 ollama list

Windows PowerShelldocker exec enjyra-v6-local-ai-02 ollama list

Run the command and compare its output with the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

llama3.2:3b

Success looks like: The retained model appears without a new download.

If it fails

What changed

You have evidence for the API and llama3.2:3b.

What you're doing

Ask the retained model to explain Docker in one sentence.

Why it matters

A real response proves the full local inference path works.

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

docker exec enjyra-v6-local-ai-02 ollama run llama3.2:3b "Explain Docker in one sentence."

docker exec enjyra-v6-local-ai-02 ollama run llama3.2:3b "Explain Docker in one sentence."

docker exec enjyra-v6-local-ai-02 ollama run llama3.2:3b "Explain Docker in one sentence."

Before You Run

Your terminal is open in Current lesson workspace.
Docker Desktop or Docker Engine is running.

Why You're Running This

This command completes the “this step” step and produces the state needed by the next instruction.

Command Breakdown

docker exec: Inspects or interacts with local Docker state without using a cloud service.

Expected Result

<one-sentence response>

The model returns a meaningful local response.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

<one-sentence response>

Success looks like: The model returns a meaningful local response.

If it fails

What changed

The model has generated a response on your machine.

Break

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

docker stop enjyra-v6-local-ai-02 && curl --fail http://localhost:11434/api/version

docker stop enjyra-v6-local-ai-02 && curl --fail http://localhost:11434/api/version

docker stop enjyra-v6-local-ai-02
curl.exe --fail http://localhost:11434/api/version

Before You Run

Your terminal is open in Current lesson workspace.
Docker Desktop or Docker Engine is running.
The local service named in the URL is already running.

Why You're Running This

This command completes the “this step” step and produces the state needed by the next instruction.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.
curl: Makes an HTTP request to the local service and prints its response.
&&: Runs the next command only when the previous command succeeds.

Expected Result

curl: connection failed

The request fails because the runtime is stopped.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

curl: connection failed

Success looks like: The request fails because the runtime is stopped.

If it fails

Debug

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

docker ps -a --filter name=enjyra-v6-local-ai-02

docker ps -a --filter name=enjyra-v6-local-ai-02

docker ps -a --filter name=enjyra-v6-local-ai-02

Before You Run

Your terminal is open in Current lesson workspace.
Docker Desktop or Docker Engine is running.

Why You're Running This

This checks the current state without changing it, so you can verify the previous action worked.

Command Breakdown

docker ps: Inspects or interacts with local Docker state without using a cloud service.

Expected Result

Exited

Docker shows the container exists but is stopped.

What Changed

BeforeThe current state has not yet been checked.

AfterNo persistent state changed; the command printed evidence you can compare with the expected result.

Verify It

macOSdocker ps -a --filter name=enjyra-v6-local-ai-02

Linuxdocker ps -a --filter name=enjyra-v6-local-ai-02

Windows PowerShelldocker ps -a --filter name=enjyra-v6-local-ai-02

Run the command and compare its output with the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

Exited

Success looks like: Docker shows the container exists but is stopped.

If it fails

Improve

Keep the model volume. Stop the container when it is not needed; do not delete the shared volume unless you intend to remove downloaded models.

Verify

TerminalChoose macOS, Linux, Windows, or Explain

Where to run Terminal or VS Code TerminalWorking directory Current lesson workspace

What to run

docker start enjyra-v6-local-ai-02 && curl -s http://localhost:11434/api/version

docker start enjyra-v6-local-ai-02 && curl -s http://localhost:11434/api/version

docker start enjyra-v6-local-ai-02
curl.exe -s http://localhost:11434/api/version

Before You Run

Your terminal is open in Current lesson workspace.
Docker Desktop or Docker Engine is running.
The local service named in the URL is already running.

Why You're Running This

This command completes the “this step” step and produces the state needed by the next instruction.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.
curl: Makes an HTTP request to the local service and prints its response.
&&: Runs the next command only when the previous command succeeds.

Expected Result

{"version":"<installed-version>"}

The API responds again after restart.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

No such file or directory → verify the current folder and the path spelling.
Permission denied → work inside your own lesson folder and confirm it is writable.
Docker daemon not running → start Docker Desktop or Docker Engine, then retry.

What you should see

{"version":"<installed-version>"}

Success looks like: The API responds again after restart.

If it fails

Challenge

Repeat the verification from a fresh terminal and explain which observable signal proves each dependency is healthy.

Ship

A local AI runtime that responds through both the Ollama CLI and HTTP API.

Next Step

Continue to Lesson 03: Call AI from Python.

03 · Prerequisites

Before You Start the Lab

Confirm your machine has everything required for this lesson. Continue only when the verification commands succeed.

This lesson needs

DockerVS Code — requiredOllamaInternet: only for first-time pullsGPU: optional

ENJYRA Lab Setup completed

Prior lesson 01 completed

Docker Desktop or Docker Engine running

Terminal access

Visual Studio Code — required for this guided lab

Disk10 GB free if the runtime or model is not cached

Memory8 GB recommended

GPUOptional

InternetRequired only for uncached dependencies or images

Docker setup

Docker CLI

Docker Engine

Test container

Ollama API

Docker not installed? Show setup instructions ▾

Check your Mac's chip architecture (Apple Silicon vs Intel) before downloading.
Where to runTerminalPlatformmacOS
COMMAND
```
uname -m
```
Expected output"arm64" (Apple Silicon) or "x86_64" (Intel)
If error
If nothing prints, open Terminal from Applications → Utilities and try again.
Download Docker Desktop for your chip from docs.docker.com/desktop/setup/install/mac-install and open the downloaded .dmg, then drag Docker to Applications.
Launch Docker Desktop from Applications and wait for the whale icon in the menu bar to show "Docker Desktop is running".
Verify the CLI is installed.
Where to runTerminalPlatformmacOS
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “command not found”, reopen Terminal after installing so your PATH refreshes, or reinstall Docker Desktop.
Verify the engine is actually running (not just installed).
Where to runTerminalPlatformmacOS
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
If it says “Cannot connect to the Docker daemon”, open Docker Desktop and wait for the whale icon to stop animating.
Run a real test container end-to-end.
Where to runTerminalPlatformmacOS
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If the image can't be pulled, check your internet connection; this is the only step that needs it.

Download Docker Desktop for Windows from docs.docker.com/desktop/setup/install/windows-install and run the installer. Docker Desktop will guide you through any compatible Windows backend requirement; this lesson does not require a WSL terminal.
If the installer asks you to enable a Windows feature or restart, follow that prompt and finish the Docker Desktop setup before continuing.
Start Docker Desktop from the Start menu and wait for the whale icon in the system tray to report "Docker Desktop is running".
Verify the CLI, using PowerShell or the built-in VS Code terminal (not a Bash-syntax terminal).
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “not recognized”, close and reopen your terminal so PATH changes take effect.
Verify the engine is running.
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
If it can't connect to the daemon, open Docker Desktop, wait until it reports Running, and review Docker Desktop's own Troubleshoot panel.
Run a real test container.
Where to runPowerShell or VS Code terminalPlatformWindows
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If the pull hangs, check your internet connection and that Docker Desktop shows as running, not starting.

Prefer Docker Engine (not Docker Desktop) on Linux. Follow the official instructions for your distribution at docs.docker.com/engine/install — for Ubuntu specifically, use docs.docker.com/engine/install/ubuntu.
Start and enable the Docker service.
Where to runTerminalPlatformLinux
COMMAND
```
sudo systemctl enable --now docker
```
Expected outputNo output on success
If error
If systemctl isn't available, your distro may use a different init system — check its Docker Engine install page.
Confirm the service is active.
Where to runTerminalPlatformLinux
COMMAND
```
systemctl status docker
```
Expected outputActive: active (running)
If error
If it shows “inactive” or “failed”, re-run the enable command above and check the install steps were completed.
Verify the CLI.
Where to runTerminalPlatformLinux
COMMAND
```
docker --version
```
Expected outputDocker version 2x.x.x, build xxxxxxx
If error
If “command not found”, the Engine install likely didn't finish — re-check the distro-specific install page.
Verify the engine responds.
Where to runTerminalPlatformLinux
COMMAND
```
docker info
```
Expected outputA block of server information, including Server Version and Containers
If error
A permission-denied error here usually means your user isn't in the docker group yet — see the note below rather than running commands as root.
Run a real test container.
Where to runTerminalPlatformLinux
COMMAND
```
docker run --rm hello-world
```
Expected output"Hello from Docker!" followed by an explanation of what happened
If error
If you see a permission error, add your user to the docker group yourself (sudo usermod -aG docker $USER, then log out and back in) — we won't do this for you automatically.

Note: running Docker commands as root or with a global chmod on the socket is not recommended. Add your user to the docker group instead, as shown above.

Before You Run

Install Docker Desktop on macOS or Windows, or Docker Engine/Desktop using the official instructions for your Linux distribution. Start Docker and open Terminal or native Windows PowerShell.

Why Do We Need Docker?

The ENJYRA AWS, Azure, and GCP labs run locally inside Docker containers. No real AWS, Azure, or GCP account is required.

Command Breakdown

docker --version: Confirms the Docker command-line client is installed.
docker info: Asks the Docker Engine for server information, proving it is running and reachable.
docker run --rm hello-world: Runs a small official test container end to end and removes it when finished.

Expected Result

Docker version 2x.x.x
docker info: server details
Hello from Docker!

What Changed

BeforeDocker is not yet confirmed.

AfterDocker CLI and Docker Engine are available to launch the local lab.

Verify It

macOSdocker --version docker info

Linuxdocker --version docker info

Windows PowerShelldocker --version docker info

If It Fails

Open Docker Desktop or start Docker Engine, wait until it reports running, and retry.
Open a new terminal after installation so PATH changes are loaded.
On Linux, follow the official instructions for your distribution rather than assuming Ubuntu commands apply.

Lesson-specific check

Where to runTerminalPlatformmacOS / Windows / Linux

COMMAND

curl http://localhost:11434/api/version

Expected output{"version":"0.x.x"}

If error

If connection is refused, the Ollama container isn't running yet — start it with this lesson's start command, then retry.

Verify your setup

Run each command in your own terminal. This page cannot run commands on your machine, so tick a box only after you see the expected result.

Docker CLIConfirms the Docker command-line client is installed.

COMMAND

docker --version

docker --version

docker --version

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.

Why You're Running This

Confirms the Docker command-line client is installed.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

A Docker version is printed

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Docker CLI is installed or running, then retry the command in a new terminal.

Expected resultA Docker version is printed

I got the expected result

Docker EngineConfirms Docker is actually running, not only installed.

COMMAND

docker info

docker info

docker info

Before You Run

Your terminal is open in Any terminal folder.
Docker Desktop or Docker Engine is running.

Why You're Running This

Confirms Docker is actually running, not only installed.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

Docker server information is returned

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm Docker Engine is installed or running, then retry the command in a new terminal.

Expected resultDocker server information is returned

I got the expected result

Visual Studio CodeOpen Visual Studio Code from your applications. The optional code --version shell command is not required; VS Code works without it.

Expected resultVisual Studio Code opens successfully

Visual Studio Code ready

Ready to start?

Docker CLI verified
Docker Engine running
Visual Studio Code ready

Need help with a prerequisite?

Docker not running? Open Docker Desktop or start Docker Engine, then run docker info.

Permission denied? Confirm your account can access the Docker daemon; do not use a global prune.

Port already in use? Use docker ps to identify the container that owns the port.

Ollama API not responding? Run docker ps, then restart the Ollama service from the lesson command.

Model not found? Run docker exec enjyra-ollama ollama list.

04 · ENJYRA Lab

Launch the ENJYRA Lab

Start only after the prerequisite checks above are complete.

This button confirms readiness; commands still run in your own terminal.

ENJYRA Lab · pinned digest

AvailabilityREMOTE VERIFIED

RuntimeDocker + Ollama

Imageollama/ollama:pinned digest

PlatformsDocker Desktop · Docker Engine

Architecturelinux/arm64

Start lab

Where to runTerminalWorking directory~/enjyra-free-aiPlatformmacOS · Linux · Windows

COMMAND

cd ~/enjyra-free-ai && docker compose up -d

cd ~/enjyra-free-ai && docker compose up -d

Set-Location "$HOME\enjyra-free-ai"
docker compose up -d

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Docker Desktop or Docker Engine is running.

Why You're Running This

Creates and starts the verified ENJYRA lesson container on your machine.

Command Breakdown

cd / Set-Location: Changes the terminal’s current working directory before the next command runs.
docker compose up: Creates and starts the services defined in the Compose file; -d keeps them running in the background.
&&: Runs the next command only when the previous command succeeds.

Expected Result

The lesson workspace is ready.

What Changed

BeforeThe requested local service or lab containers are not yet confirmed running.

AfterThe local service or lab containers are running in the background and can be verified.

Verify It

macOSdocker ps

Linuxdocker ps

Windows PowerShelldocker ps

Confirm the expected container is listed as running or healthy.

If It Fails

Check the current folder, confirm Docker is running when required, and verify the image or script name before retrying.

What this command does Creates and starts the verified ENJYRA lesson container on your machine.

Expected result

The lesson workspace is ready.

If it fails

Check the current folder, confirm Docker is running when required, and verify the image or script name before retrying.

Verify running state

Where to runTerminalWorking directory~/enjyra-free-aiPlatformmacOS · Linux · Windows

COMMAND

curl -s http://localhost:11434/api/version

curl -s http://localhost:11434/api/version

curl.exe -s http://localhost:11434/api/version

Before You Run

Your terminal is open in ~/enjyra-free-ai.
The local service named in the URL is already running.

Why You're Running This

Checks the local runtime state required by this lesson without changing it.

Command Breakdown

curl: Makes an HTTP request to the local service and prints its response.

Expected Result

The expected version or runtime information is printed.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Confirm the start command completed first, then compare the image, container, port, or script name exactly with this lesson.

What this command does Checks the local runtime state required by this lesson without changing it.

Expected result

The expected version or runtime information is printed.

If it fails

Confirm the start command completed first, then compare the image, container, port, or script name exactly with this lesson.

Reset My Lab

Reset lab

Where to runTerminalWorking directory~/enjyra-free-aiPlatformmacOS · Linux · Windows

COMMAND

docker restart enjyra-v6-local-ai-02

docker restart enjyra-v6-local-ai-02

docker restart enjyra-v6-local-ai-02

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Docker Desktop or Docker Engine is running.

Why You're Running This

Returns this lesson’s local runtime or resources to a known starting state without touching unrelated projects.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

This lesson’s runtime state is reset or the named lesson container is removed.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Check Docker is running and make sure you are using this lesson’s exact container or launcher name.

What this command does Returns this lesson’s local runtime or resources to a known starting state without touching unrelated projects.

Expected result

This lesson’s runtime state is reset or the named lesson container is removed.

If it fails

Check Docker is running and make sure you are using this lesson’s exact container or launcher name.

Clean up the lab

Clean up lab

Where to runTerminalWorking directory~/enjyra-free-aiPlatformmacOS · Linux · Windows

COMMAND

docker stop enjyra-v6-local-ai-02

docker stop enjyra-v6-local-ai-02

docker stop enjyra-v6-local-ai-02

Before You Run

Your terminal is open in ~/enjyra-free-ai.
Docker Desktop or Docker Engine is running.

Why You're Running This

Stops or removes only this lesson’s local runtime so it does not keep using machine resources.

Command Breakdown

docker: Runs the docker tool with the shown arguments to complete this lesson step.

Expected Result

The lesson-scoped service or container is no longer running.

What Changed

BeforeThe action shown by this card has not yet been completed or verified.

AfterThe command has completed and the next lesson step has the state it needs.

Verify It

macOSInspect the command output shown above.

LinuxInspect the command output shown above.

Windows PowerShellInspect the command output shown above.

Continue only when the observed result matches the Expected Result block.

If It Fails

Inspect this lesson’s status first, then retry with the exact launcher or container name shown here.

What this command does Stops or removes only this lesson’s local runtime so it does not keep using machine resources.

Expected result

The lesson-scoped service or container is no longer running.

If it fails

Inspect this lesson’s status first, then retry with the exact launcher or container name shown here.

Final Check

Prove What You Learned

This lesson contains 4 focused questions. You answered 2 before the build; complete the remaining 2 now.

3 / 4Interview

Why keep Ollama separate from learner images?

One separate runtime avoids duplicating model weights.

4 / 4Exam

What should you inspect after a connection failure?

Container state is the first observable signal for a stopped runtime.

Skills you practised

Run Your First Local AI Model

What you'll build

Before You Build

Where are model weights stored in this lesson?

What does /api/version verify?

Mission

Understand

Build

▸Step 1Create the Ollama Compose file

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

▸Step 2Start the local AI runtime

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

▸Step 3Verify the API and model

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

▸Step 4Generate the first local response

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Break

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Debug

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Improve

Verify

Before You Run

Why You're Running This

Command Breakdown

Expected Result

What Changed

Verify It

If It Fails

Challenge