jetbrains-cadence 1.2.0

JetBrains Cadence CLI

Cadence CLI

Install

pip install jetbrains-cadence

This will create a cadence script in your current environment.

Getting Started

cadence login

If you want to use Cadence CLI in the non-interactive environment, you can create the token manually and pass it via CADENCE_TOKEN environment variable

Start the execution from YAML config

cadence execution start --preset path/to/config.yaml

This will print the ID of the started execution.

Basic Example

working_dir: .
cmd:
  - python3 main.py
description: basic preset

provisioning:
  gpu_type: H200
  gpu_count: 1

env:
  python:
    version: 3.11
    pip:
      requirements_path: requirements.txt

project_sync:
  local:
    root: .
    storage_name: Cadence Storage
    uri: ""
    storage_type: DEFAULT

outputs:
  - type: OUTPUT
    storage_name: Cadence Storage
    uri: ""
    path: outputs # `./outputs` will be saved to S3
    storage_type: DEFAULT

More Examples

Please see the cadence-examples repository for more in-depth examples.