### Install dstack Server with uv

Source: https://dstack.ai/docs/installation

Installs the dstack server and all its dependencies using uv. This command also starts the server.

```shell
$ uv tool install "dstack[all]" -U
$ dstack server

Applying ~/.dstack/server/config.yml...

The admin token is "bbae0f28-d3dd-4820-bf61-8f4bb40815da"
The server is running at http://127.0.0.1:3000/

```

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

### Install dstack Server with pip

Source: https://dstack.ai/docs/installation

Installs the dstack server and all its dependencies using pip. This command also starts the server.

```shell
$ pip install "dstack[all]" -U
$ dstack server

Applying ~/.dstack/server/config.yml...

The admin token is "bbae0f28-d3dd-4820-bf61-8f4bb40815da"
The server is running at http://127.0.0.1:3000/

```

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

### Initialize Dev Environment with Commands

Source: https://dstack.ai/docs/concepts/dev-environments

Pre-configure a dev environment by specifying a list of commands to run at startup using the `init` property. This example installs the `wandb` package.

```yaml
type: dev-environment
name: vscode

python: "3.11"
ide: vscode

init:
  - pip install wandb
```

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

### Install and Run dstack Server with uv

Source: https://dstack.ai/docs/guides/server-deployment

This snippet shows how to install the dstack server using uv, a fast Python package installer. Git and OpenSSH are prerequisites.

```shell
$ uv tool install 'dstack[all]' -U
$ dstack server

Applying ~/.dstack/server/config.yml...

The admin token is "bbae0f28-d3dd-4820-bf61-8f4bb40815da"
The server is running at http://127.0.0.1:3000/

```

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

### Install dstack CLI with uv

Source: https://dstack.ai/docs/installation

Installs the dstack CLI tool using uv.

```shell
$ uv tool install dstack -U

```

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

### Runpod Backend Selection Example

Source: https://dstack.ai/docs/concepts/backends

This example demonstrates how dstack lists available Runpod instances, differentiating between Secure Cloud and Community Cloud regions based on their identifiers.

```shell
$ dstack apply -f .dstack.yml -b runpod

 #  BACKEND  REGION    INSTANCE               SPOT  PRICE
 1  runpod   CA        NVIDIA A100 80GB PCIe  yes   $0.6
 2  runpod   CA-MTL-3  NVIDIA A100 80GB PCIe  yes   $0.82
```

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

### Install dstack with Fluent-bit extras

Source: https://dstack.ai/docs/guides/server-deployment

Install the necessary extras for Fluent-bit log storage functionality.

```shell
$ pip install "dstack[all]" -U
# or
$ pip install "dstack[fluentbit]" -U
```

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

### Install dstack CLI with pip

Source: https://dstack.ai/docs/installation

Installs the dstack CLI tool using pip.

```shell
$ pip install dstack -U

```

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

### Schedule Dev Environment with Cron

Source: https://dstack.ai/docs/dev-environments

Use the `schedule` property with cron syntax to start a dev environment periodically. This example schedules an environment to start at 8:00 UTC from Monday through Friday. It can be combined with `max_duration` or `utilization_policy` for automatic shutdown.

```yaml
type: dev-environment
ide: vscode
schedule:
  cron: "0 8 * * mon-fri" # at 8:00 UTC from Monday through Friday
.dstack.yml
```

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

### Install dstack Skills for AI Agents

Source: https://dstack.ai/docs/installation

Install the dstack skills package using npx to enable AI agents to use the CLI and manage configuration files.

```shell
$ npx skills add dstackai/dstack
```

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

### Define a Service with a Specific Python Version

Source: https://dstack.ai/docs/concepts/services

Configure a service to use a specific Python version. This example sets the Python version to 3.12 and starts a simple HTTP server.

```yaml
type: service
name: http-server-service

python: 3.12

commands:
  - python3 -m http.server
port: 8000
```

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

### Apply Run Plan

Source: https://dstack.ai/docs/reference/api/python

Apply a pre-generated run plan to start a run. This method is used when you have obtained a `RunPlan` object, for example, from `get_run_plan`.

```python
from dstack.api import Client, RunPlan

client = Client.from_config()

# Assume run_plan is a RunPlan object obtained elsewhere
run_plan: RunPlan = ... 

run = client.runs.apply_plan(run_plan=run_plan)

```

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

### Fluent-bit configuration for Elasticsearch/OpenSearch

Source: https://dstack.ai/docs/guides/server-deployment

Example Fluent-bit configuration to receive logs and forward them to Elasticsearch or OpenSearch.

```ini
[INPUT]
    Name        forward
    Listen      0.0.0.0
    Port        24224

[OUTPUT]
    Name            es
    Match           dstack.*
    Host            elasticsearch.example.com
    Port            9200
    Index           dstack-logs
    Suppress_Type_Name On
```

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

### Specifying Service Account Data with jq

Source: https://dstack.ai/docs/reference/server/config.yml

Example of how to specify service account file contents as a string using `jq` for processing.

```shell
cat my-service-account-file.json | jq -c | jq -R
```

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

### Display dstack init help

Source: https://dstack.ai/docs/reference/cli/dstack/init

View available options for initializing dstack, including how to set custom Git credentials.

```shell
$ dstack init --help
```

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

### Configure Vast.ai Backend

Source: https://dstack.ai/docs/concepts/backends

Configure the Vast.ai backend by providing your API key in the `config.yml` file. This is the standard setup for using Vast.ai with dstack.

```yaml
projects:
- name: main
  backends:
    - type: vastai
      creds:
        type: api_key
        api_key: d75789f22f1908e0527c78a283b523dd73051c8c7d05456516fc91e9d4efd8c5
```

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

### Init Repo

Source: https://dstack.ai/docs/reference/api/http

Initializes a new repository for a project.

```APIDOC
## POST /api/project/{project_name}/repos/init

### Description
Init Repo

### Method
POST

### Endpoint
/api/project/{project_name}/repos/init

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Enable NVCC for CUDA Development

Source: https://dstack.ai/docs/concepts/dev-environments

Set 'nvcc' to true to include nvcc in the base Docker image, required for building custom CUDA kernels. Includes an example of installing flash_attn.

```yaml
type: dev-environment
name: vscode

python: 3.12
nvcc: true

ide: vscode
init:
  - uv pip install flash_attn --no-build-isolation
```

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

### Run a Streamlit application as a dstack task

Source: https://dstack.ai/docs/guides/protips

Configure a dstack task to run a Streamlit application. This example installs Streamlit and then runs the 'streamlit hello' command. The task exposes port 8501.

```yaml
type: task
name: streamlit-task

python: 3.12

commands:
  - uv pip install streamlit
  - streamlit hello
ports: 
  - 8501

```

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

### Initialize a dstack Plugin Package

Source: https://dstack.ai/docs/reference/plugins/python

Use `uv init --library` to quickly set up the basic structure for a new Python library, which can then be developed into a dstack plugin.

```bash
uv init --library
```

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

### Get Next Redirect

Source: https://dstack.ai/docs/reference/api/http

Gets the next redirect URL for authentication.

```APIDOC
## POST /api/auth/get_next_redirect

### Description
Get Next Redirect

### Method
POST

### Endpoint
/api/auth/get_next_redirect
```

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

### Configure Lambda Backend

Source: https://dstack.ai/docs/concepts/backends

Set up the Lambda backend by providing your API key. Ensure the API key is generated from your Lambda Cloud account.

```yaml
projects:
- name: main
  backends:
    - type: lambda
      creds:
        type: api_key
        api_key: eersct_yrpiey-naaeedst-tk-_cb6ba38e1128464aea9bcc619e4ba2a5.iijPMi07obgt6TZ87v5qAEj61RVxhd0p
```

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

### Dev Environment with Initialization Commands

Source: https://dstack.ai/docs/dev-environments

Pre-configure a dev environment by specifying initialization commands to run at startup using the `init` property.

```yaml
type: dev-environment
name: vscode

python: "3.11"
ide: vscode

init:
  - pip install wandb
examples/.dstack.yml
```

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

### Schedule Dev Environment Startup with Cron

Source: https://dstack.ai/docs/concepts/dev-environments

Use the `schedule` property with cron syntax to start a dev environment periodically at specific UTC times. This allows for automated, time-based environment provisioning.

```yaml
type: dev-environment
ide: vscode
schedule:
  cron: "0 8 * * mon-fri" # at 8:00 UTC from Monday through Friday
```

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

### Display help for dstack offer

Source: https://dstack.ai/docs/reference/cli/dstack/offer

Use the --help flag to view all available options and arguments for the `dstack offer` command.

```shell
dstack offer --help
```

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

### Get Instance Details

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves detailed information about a specific instance. Use this endpoint to get configuration and status of an instance.

```APIDOC
## GET /instances/{id}

### Description
Retrieves detailed information about a specific instance.

### Method
GET

### Endpoint
/instances/{id}

### Parameters
#### Path Parameters
- **id** (string) - Required - The unique identifier of the instance.

### Response
#### Success Response (200)
- **id** (string) - The unique identifier of the instance.
- **name** (string) - The name of the instance.
- **status** (string) - The current status of the instance (e.g., 'running', 'stopped').
- **ip_address** (string) - The IP address assigned to the instance.

#### Response Example
```json
{
  "id": "a1b2c3d4-e5f6-7890-1234-567890abcdef",
  "name": "my-instance",
  "status": "running",
  "ip_address": "192.168.1.100"
}
```
```

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

### dstack secret get --help

Source: https://dstack.ai/docs/reference/cli/dstack/secret

View help for the dstack secret get command. Use this to retrieve the value of a specific secret.

```shell
dstack secret get --help
```

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

### Initialize Repository

Source: https://dstack.ai/docs/reference/api/python

Initializes a repository and configures its credentials. Must be called before mounting the repo to a run. Supports overriding default Git credentials.

```python
repo = RemoteRepo.from_url(
    repo_url="https://github.com/dstackai/dstack-examples",
    repo_branch="main",
)
client.repos.init(repo)
```

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

### Start dstack Server with PostgreSQL

Source: https://dstack.ai/docs/guides/server-deployment

Set the DSTACK_DATABASE_URL environment variable to connect to a PostgreSQL database. Ensure the database instance meets minimum requirements or adjust pool sizes.

```shell
$ DSTACK_DATABASE_URL=postgresql+asyncpg://user:password@db-host:5432/dstack dstack server
```

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

### Get a specific secret value

Source: https://dstack.ai/docs/concepts/secrets

Use `dstack secret get` to retrieve the actual value of a specific secret. This is useful for inspecting secret contents.

```shell
$ dstack secret get my_secret
 NAME       VALUE             
 my_secret  some_secret_value 

```

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

### Initialize Repository

Source: https://dstack.ai/docs/reference/api/python

Initializes a repository and configures its credentials. This must be called before mounting the repo to a run.

```APIDOC
## dstack.api.RepoCollection.init(repo, git_identity_file=None, oauth_token=None, creds=None)

### Description
Initializes the repo and configures its credentials in the project. Must be invoked before mounting the repo to a run.

### Parameters
#### Path Parameters
- **repo** (Repo) - Required - The repo to initialize.
- **git_identity_file** (Optional[PathLike]) - Optional - The private SSH key path for accessing the remote repo. Defaults to `None`.
- **oauth_token** (Optional[str]) - Optional - The GitHub OAuth token to access the remote repo. Defaults to `None`.
- **creds** (Optional[RemoteRepoCreds]) - Optional - Optional prepared repo credentials. If specified, both `git_identity_file` and `oauth_token` are ignored. Defaults to `None`.

### Example
```python
repo = RemoteRepo.from_url(
    repo_url="https://github.com/dstackai/dstack-examples",
    repo_branch="main",
)
client.repos.init(repo)
```

By default, it uses the default Git credentials configured on the machine. You can override these credentials via the `git_identity_file` or `oauth_token` arguments of the `init` method.
Once the repo is initialized, you can pass the repo object to the run:
```python
run = client.runs.apply_configuration(
    configuration=...,
    repo=repo,
)
```
```

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

### Configure a Service Run

Source: https://dstack.ai/docs/quickstart

Deploy a model or web app as a service endpoint. Specify the Docker image, commands to run the service, exposed ports, model name (if applicable), and required resources like GPUs.

```yaml
type: service
name: qwen36-service

image: lmsysorg/sglang:v0.5.10.post1

commands:
  - |
    sglang serve \
      --model-path Qwen/Qwen3.6-27B \
      --host 0.0.0.0 \
      --port 30000 \
      --reasoning-parser qwen3
# Expose the SGLang server port
port: 30000

# Specify a name if it's an OpenAI-compatible model
model: Qwen/Qwen3.6-27B

# Required resources
resources:
  gpu: H100
```

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

### Display dstack server help

Source: https://dstack.ai/docs/reference/cli/dstack/server

Use the --help flag to view all available options for the dstack server command.

```shell
$ dstack server --help
```

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

### Use Instance Volume for Pip Caching

Source: https://dstack.ai/docs/concepts/volumes

Mount a directory on the instance to a specific path within the container to cache pip installations. This speeds up subsequent runs by reusing previously installed packages.

```yaml
type: task

volumes:
  - /dstack-cache/pip:/root/.cache/pip
```

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

### Display help for dstack project list

Source: https://dstack.ai/docs/reference/cli/dstack/project

View the help message for the `dstack project list` command to understand how to list all configured projects on the client.

```shell
$ dstack project list --help
```

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

### Get Instance

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Returns an instance given its ID.

```APIDOC
## POST /api/project/{project_name}/instances/get

### Description
Returns an instance given its ID.

### Method
POST

### Endpoint
/api/project/{project_name}/instances/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - Project Name

### Request Body
- **request** (GetInstanceRequestRequest) - Required

### Response
#### Success Response (200)
- **response** (InstanceRequest) - Successful Response
```

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

### Get Server Info

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves information about the server.

```APIDOC
## POST /api/server/get_info

### Description
Retrieves information about the server.

### Method
POST

### Endpoint
/api/server/get_info

### Response
#### Success Response (200)
- **schema**: ServerInfoRequest
```

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

### Apply Run Configuration

Source: https://dstack.ai/docs/reference/api/python

Apply a run configuration (Task, Service, or DevEnvironment) to start a run. The `repo` argument can be used to mount additional files or code.

```python
from dstack.api import Task, Client

client = Client.from_config()

task = Task(
    name="my-task",
    image="my-image:latest",
    commands=["echo hello"]
)

run = client.runs.apply_configuration(configuration=task)

```

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

### OCI Backend with Client Credentials

Source: https://dstack.ai/docs/concepts/backends

Configure an OCI backend using client credentials. This requires user, tenancy, region, fingerprint, and a private key file.

```yaml
projects:
- name: main
  backends:
  - type: oci
    creds:
      type: client
      user: ocid1.user.oc1..g5vlaeqfu47akmaafq665xsgmyaqjktyfxtacfxc4ftjxuca7aohnd2ev66m
      tenancy: ocid1.tenancy.oc1..ajqsftvk4qarcfaak3ha4ycdsaahxmaita5frdwg3tqo2bcokpd3n7oizwai
      region: eu-frankfurt-1
      fingerprint: 77:32:77:00:49:7c:cb:56:84:75:8e:77:96:7d:53:17
      key_file: "~/.oci/private_key.pem"
```

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

### Get Run

Source: https://dstack.ai/docs/reference/api/http

Retrieves details of a specific run.

```APIDOC
## POST /api/project/{project_name}/runs/get

### Description
Get Run

### Method
POST

### Endpoint
/api/project/{project_name}/runs/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Configure a Dev Environment Run

Source: https://dstack.ai/docs/quickstart

Use this configuration to provision a development instance accessible with your desktop IDE. Specify the Python version and optionally request resources like GPUs.

```yaml
type: dev-environment
name: vscode

# If `image` is not specified, dstack uses its default image
python: "3.11"
#image: dstackai/base:py3.13-0.7-cuda-12.1

ide: vscode

# Uncomment to request resources
#resources:
#  gpu: 24GB
```

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

### Get Repo

Source: https://dstack.ai/docs/reference/api/http

Retrieves details of a specific repository.

```APIDOC
## POST /api/project/{project_name}/repos/get

### Description
Get Repo

### Method
POST

### Endpoint
/api/project/{project_name}/repos/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Create dstack Client

Source: https://dstack.ai/docs/reference/api/python

Creates a Client instance using the default configuration from `~/.dstack/config.yml`. Optionally specify project name, server URL, and user token.

```python
from dstack.api import Client

client = Client.from_config()

```

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

### Get Instance

Source: https://dstack.ai/docs/reference/api/http

Retrieves details of a specific instance.

```APIDOC
## POST /api/project/{project_name}/instances/get

### Description
Get Instance

### Method
POST

### Endpoint
/api/project/{project_name}/instances/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Get Fleet

Source: https://dstack.ai/docs/reference/api/http

Retrieves details of a specific fleet.

```APIDOC
## POST /api/project/{project_name}/fleets/get

### Description
Get Fleet

### Method
POST

### Endpoint
/api/project/{project_name}/fleets/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### OCI Backend with Default Credentials

Source: https://dstack.ai/docs/concepts/backends

Configure an OCI backend using default credentials. Ensure you have `~/.oci/config` set up.

```yaml
projects:
- name: main
  backends:
  - type: oci
    creds:
      type: default
```

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

### Install dstack Server with Docker

Source: https://dstack.ai/docs/installation

Runs the dstack server using a Docker container, mapping the host's .dstack/server directory to the container and exposing port 3000.

```shell
$ docker run -p 3000:3000 \
    -v $HOME/.dstack/server/:/root/.dstack/server \
    dstackai/dstack

Applying ~/.dstack/server/config.yml...

The admin token is "bbae0f28-d3dd-4820-bf61-8f4bb40815da"
The server is running at http://127.0.0.1:3000/

```

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

### Get Project

Source: https://dstack.ai/docs/reference/api/http

Retrieves details of a specific project.

```APIDOC
## POST /api/projects/{project_name}/get

### Description
Get Project

### Method
POST

### Endpoint
/api/projects/{project_name}/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Configure Nebius Backend with Private Key File

Source: https://dstack.ai/docs/concepts/backends

Configure the Nebius backend using a service account and a private key file. The private key file path should be correctly specified.

```yaml
projects:
- name: main
  backends:
  - type: nebius
    creds:
      type: service_account
      service_account_id: serviceaccount-e00dhnv9ftgb3cqmej
      public_key_id: publickey-e00ngaex668htswqy4
      private_key_file: ~/path/to/key.pem
```

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

### Get User

Source: https://dstack.ai/docs/reference/api/http

Retrieves information about a specific user.

```APIDOC
## POST /api/users/get_user

### Description
Get User

### Method
POST

### Endpoint
/api/users/get_user
```

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

### Get Run Plan

Source: https://dstack.ai/docs/reference/api/http

Retrieves the plan for a specific run.

```APIDOC
## POST /api/project/{project_name}/runs/get_plan

### Description
Get Plan

### Method
POST

### Endpoint
/api/project/{project_name}/runs/get_plan

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Create Project

Source: https://dstack.ai/docs/reference/api/http

Creates a new project.

```APIDOC
## POST /api/projects/create

### Description
Create Project

### Method
POST

### Endpoint
/api/projects/create
```

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

### Get Fleet Plan

Source: https://dstack.ai/docs/reference/api/http

Retrieves the plan for a specific fleet.

```APIDOC
## POST /api/project/{project_name}/fleets/get_plan

### Description
Get Plan

### Method
POST

### Endpoint
/api/project/{project_name}/fleets/get_plan

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
```

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

### Display dstack apply help information

Source: https://dstack.ai/docs/reference/cli/dstack/apply

Use this command to view the available options and usage instructions for `dstack apply`.

```shell
dstack apply --help
```

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

### Get Backend Yaml

Source: https://dstack.ai/docs/reference/api/http

Retrieves the YAML configuration of a backend.

```APIDOC
## POST /api/project/{project_name}/backends/{backend_name}/get_yaml

### Description
Get Backend Yaml

### Method
POST

### Endpoint
/api/project/{project_name}/backends/{backend_name}/get_yaml

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
- **backend_name** (string) - Required - The name of the backend
```

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

### MPI with Slurm

Source: https://dstack.ai/docs/guides/migration/slurm

Set up Slurm for MPI workloads, including GPU allocation and memory. This example configures the environment variables for master address and port, creates a hostfile, and runs an NCCL performance test.

```bash
#!/bin/bash
#SBATCH --nodes=2
#SBATCH --ntasks=16
#SBATCH --gres=gpu:8
#SBATCH --mem=200G
#SBATCH --time=24:00:00

export MASTER_ADDR=$(scontrol show hostnames $SLURM_NODELIST | head -n1)
export MASTER_PORT=12345

# Convert SLURM_JOB_NODELIST to hostfile format
HOSTFILE=$(mktemp)
scontrol show hostnames $SLURM_JOB_NODELIST | awk -v slots=$SLURM_NTASKS_PER_NODE '{print $0" slots="slots}' > $HOSTFILE

# MPI with NCCL tests or custom MPI application
mpirun \
--allow-run-as-root \
--hostfile $HOSTFILE \
-n $SLURM_NTASKS \
--bind-to none \
/opt/nccl-tests/build/all_reduce_perf -b 8 -e 8G -f 2 -g 1

rm -f $HOSTFILE

```

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

### Get My User

Source: https://dstack.ai/docs/reference/api/http

Retrieves information about the currently authenticated user.

```APIDOC
## POST /api/users/get_my_user

### Description
Get My User

### Method
POST

### Endpoint
/api/users/get_my_user
```

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

### dstack secret get

Source: https://dstack.ai/docs/reference/cli/dstack/secret

Shows the value of a specified secret.

```APIDOC
## dstack secret get

### Description
Shows the value of a specified secret.

### Usage
```shell
dstack secret get [name]
```

### Parameters
#### Path Parameters
- **name** (string) - Required - The name of the secret to retrieve.

### Example
```shell
dstack secret get MY_API_KEY
```
```

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

### Get Fleet Plan

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves the plan for a fleet based on its specification.

```APIDOC
## POST /fleet/plan

### Description
Retrieves the plan for a fleet based on its specification.

### Method
POST

### Endpoint
/fleet/plan

### Request Body
- **spec** (FleetSpecRequest) - Required - The specification of the fleet for which to retrieve the plan.
```

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

### Apply Task Configuration

Source: https://dstack.ai/docs/quickstart

Apply the task configuration using `dstack apply`. This will provision an instance and run the specified commands. Forwarded ports will be accessible via `localhost`.

```shell
$ dstack apply -f task.dstack.yml

 #  BACKEND  REGION           RESOURCES                 SPOT  PRICE
 1  gcp      us-west4         2xCPU, 8GB, 100GB (disk)  yes   $0.010052
 2  azure    westeurope       2xCPU, 8GB, 100GB (disk)  yes   $0.0132
 3  gcp      europe-central2  2xCPU, 8GB, 100GB (disk)  yes   $0.013248
 
Submit the run streamlit? [y/n]: y

Provisioning `streamlit`...
---> 100%

  Welcome to Streamlit. Check out our demo in your browser.

  Local URL: http://localhost:8501
```

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

### Initiate Rolling Deployment

Source: https://dstack.ai/docs/concepts/services

Use `dstack apply` to update a running service. `dstack` detects changes and prompts for a rolling update.

```shell
$ dstack apply -f my-service.dstack.yml

Active run my-service already exists. Detected changes that can be updated in-place:
- Repo state (branch, commit, or other)
- File archives
- Configuration properties:
  - env
  - files

Update the run? [y/n]:
```

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

### Get Models

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves a list of models associated with a specific project.

```APIDOC
## GET /proxy/models/{project_name}/models

### Description
Retrieves a list of models for a given project.

### Method
GET

### Endpoint
/proxy/models/{project_name}/models

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project.

### Response
#### Success Response (200)
- Description: Successful Response
- **ModelsResponseRequest**: A response object containing models.

#### Error Response (422)
- Description: Validation Error
```

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

### Inspect Available Offers

Source: https://dstack.ai/docs/guides/troubleshooting

Use the `dstack offer` command to inspect available instance offers. You can filter by GPU type or by a specific fleet.

```shell
# All matching offers, ignoring fleet configurations
$ dstack offer --gpu H100

# Offers available through a specific fleet
$ dstack offer --gpu H100 --fleet my-fleet
```

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

### Get Gateway

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves details of a specific gateway within a project.

```APIDOC
## POST /api/project/{project_name}/gateways/get

### Description
Retrieves details of a specific gateway within a project.

### Method
POST

### Endpoint
/api/project/{project_name}/gateways/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project.

#### Request Body
- **name** (string) - Required - The name of the gateway to retrieve.

### Request Example
```json
{
  "name": "example_gateway_name"
}
```

### Response
#### Success Response (200)
- **name** (string) - The name of the gateway.
- **url** (string) - The URL of the gateway.

#### Response Example
```json
{
  "name": "example_gateway_name",
  "url": "http://example.com"
}
```
```

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

### dstack.api.Client.from_config

Source: https://dstack.ai/docs/reference/api/python

Creates a Client using the default configuration from ~/.dstack/config.yml if it exists.

```APIDOC
## `dstack.api.Client.from_config`

### Description
Creates a Client using the default configuration from `~/.dstack/config.yml` if it exists.

### Method
`staticmethod`

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Request Example
```python
client = Client.from_config()
```

### Response
#### Success Response (200)
- **client** (`Client`) - A client instance.

#### Response Example
```json
{
  "client": "<dstack.api.Client object>"
}
```

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Parameters
#### Path Parameters
None

#### Query Parameters
- **project_name** (Optional[str]) - The name of the project. required if `server_url` and `user_token` are specified. Default: `None`
- **server_url** (Optional[str]) - The dstack server URL (e.g. `http://localhost:3000/` or `https://sky.dstack.ai`). Default: `None`
- **user_token** (Optional[str]) - The dstack user token. Default: `None`

### Request Example
```python
client = Client.from_config(project_name='my-project', server_url='http://localhost:3000/', user_token='my-token')
```

### Response
#### Success Response (200)
- **client** (`Client`) - A client instance.

#### Response Example
```json
{
  "client": "<dstack.api.Client object>"
}
```
```

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

### Get Secret

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves a specific secret by its name within a project.

```APIDOC
## POST /api/project/{project_name}/secrets/get

### Description
Retrieves a specific secret by its name within a project.

### Method
POST

### Endpoint
/api/project/{project_name}/secrets/get

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project.

#### Request Body
- **name** (string) - Required - The name of the secret to retrieve.

### Request Example
```json
{
  "name": "example_secret_name"
}
```

### Response
#### Success Response (200)
- **name** (string) - The name of the secret.
- **value** (string) - The value of the secret.

#### Response Example
```json
{
  "name": "example_secret_name",
  "value": "example_secret_value"
}
```
```

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

### Slurm Container Data Access and Mounting

Source: https://dstack.ai/docs/guides/migration/slurm

This example is for Slurm jobs using containers. It shows how to explicitly mount shared filesystems into the container and utilize local scratch space for training.

```bash
#!/bin/bash
#SBATCH --nodes=4
#SBATCH --gres=gpu:8
#SBATCH --time=24:00:00

# Shared filesystem mounted at /datasets and /checkpoints
DATASET_PATH=/datasets/imagenet

# Local scratch accessible via $SLURM_TMPDIR (host storage mounted into container)
# Copy dataset to local scratch, then train
srun --container-image=/shared/images/pytorch-2.0-cuda11.8.sif \
  --container-mounts=/shared/datasets:/datasets,/shared/checkpoints:/checkpoints \
  cp -r $DATASET_PATH $SLURM_TMPDIR/dataset

srun --container-image=/shared/images/pytorch-2.0-cuda11.8.sif \
  --container-mounts=/shared/datasets:/datasets,/shared/checkpoints:/checkpoints \
  python train.py \
    --data=$SLURM_TMPDIR/dataset \
    --checkpoint-dir=/checkpoints \
    --epochs=100

# \$SLURM_TMPDIR automatically cleaned when job ends
# Checkpoints saved to mounted shared filesystem persist

```

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

### Get Job Metrics

Source: https://dstack.ai/docs/reference/api/http

Retrieves metrics for a specific job run.

```APIDOC
## GET /api/project/{project_name}/metrics/job/{run_name}

### Description
Get Job Metrics

### Method
GET

### Endpoint
/api/project/{project_name}/metrics/job/{run_name}

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project
- **run_name** (string) - Required - The name of the run
```

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

### Initialize and Mount RemoteRepo

Source: https://dstack.ai/docs/reference/api/python

Initialize a RemoteRepo instance with the client and then pass it to a run configuration. This mounts the repository into the run environment. Git credentials can be overridden using `git_identity_file` or `oauth_token`.

```python
client.repos.init(repo)
```

```python
run = client.runs.apply_configuration(
    configuration=...,
    repo=repo,
)
```

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

### Get File Archive by Hash

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves a file archive using its hash.

```APIDOC
## GET /file/archive

### Description
Retrieves a file archive using its hash.

### Method
GET

### Endpoint
/file/archive

### Parameters
#### Query Parameters
- **hash** (string) - Required - The hash of the file archive to retrieve.
```

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

### Get Upstream

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Retrieves upstream information, likely for SSH proxy configuration.

```APIDOC
## POST /api/sshproxy/get_upstream

### Description
Fetches upstream details for SSH proxy.

### Method
POST

### Endpoint
/api/sshproxy/get_upstream

### Request Body
- **GetUpstreamRequestRequest** (object) - Required - Request body for getting upstream information.

### Responses
#### Success Response (200)
- **GetUpstreamResponseRequest** (object) - Successful Response

#### Error Responses
- **400** - Bad request
- **403** - Access denied
- **422** - Validation Error
```

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

### Configure Hot Aisle Backend

Source: https://dstack.ai/docs/concepts/backends

Set up the Hot Aisle backend with your team handle and API key. Ensure the API key has the necessary Owner and Operator roles.

```yaml
projects:
- name: main
  backends:
    - type: hotaisle
      team_handle: hotaisle-team-handle
      creds:
        type: api_key
        api_key: 9c27a4bb7a8e472fae12ab34.3f2e3c1db75b9a0187fd2196c6b3e56d2b912e1c439ba08d89e7b6fcd4ef1d3f
```

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

### Get Plan

Source: https://dstack.ai/docs/reference/api/http/openapi.json

Generates a fleet plan based on the provided fleet configuration.

```APIDOC
## POST /api/project/{project_name}/fleets/get_plan

### Description
Returns a fleet plan for the given fleet configuration.

### Method
POST

### Endpoint
/api/project/{project_name}/fleets/get_plan

### Parameters
#### Path Parameters
- **project_name** (string) - Required - The name of the project.

#### Request Body
- **GetFleetPlanRequestRequest** (object) - Required - Request body schema for getting a fleet plan.

### Response
#### Success Response (200)
- **FleetPlanRequest** (object) - The fleet plan object.

#### Error Responses
- **400** - Bad request
- **403** - Access denied
- **422** - Validation Error
```

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

### Get Repository

Source: https://dstack.ai/docs/reference/api/python

Retrieves repository information by its ID. Optionally includes credentials.

```APIDOC
## dstack.api.RepoCollection.get(repo_id, *, with_creds=False)

### Description
Returns the repo by `repo_id`

### Parameters
#### Path Parameters
- **repo_id** (str) - Required - The repo ID.
- **with_creds** (bool) - Optional - include repo credentials in the response. Defaults to `False`.

### Returns
- **Optional[Union[RepoHead, RepoHeadWithCreds]]** - The repo or `None` if the repo is not found.
```