Valohai Help
Promoted articles
- Getting started with Valohai
- Getting started with pipelines
- Bring your existing projects to Valohai
- Introduction to the valohai-utils toolkit
- Introduction to parameters
- Attach a debugger (VSCode)
- Run a grid search
- Run Bayesian optimization
- Introduction to deployments
- Run notebooks on Valohai
- Add input files to your execution
- Data aliases
Categories
-
Getting Started
New to Valohai executions and pipelines? Start here. -
Executions
Working with parameters, metrics, and other execution details. -
Pipelines
Defining and scheduling pipelines for training, inference, and ETL. -
Tasks
Using Tasks for parameter sweeps and hyperparmeter optimization. -
Deployments
Manage deployments and collect custom metrics from real-time inference. -
Notebooks
Create executions directly from Notebooks using Jupyhai. -
Git repositories
Connect your Git repositories to Valohai projects. -
Data
Working with data files, versioning, aliases, and tracing. -
Docker
Building and using custom Docker images. -
Command-line interface
Using the Valohai CLI tools to manage projects, executions, and pipelines. -
API
Trigger new executions or fetch information about current executions and pipelines. -
Organization management
Invite users, configure shared data stores, and manage teams. -
Setup
Deploying Valohai to your own cloud or on-premises environment.
From the blog
-
Git for Data Science: What every data scientist should know about Git
Git is becoming a crucial tool for data scientists as teams work on larger, more production-focused ML projects. So Juha put together six rules of thumb to avoid the biggest pitfalls with Git -
What every data scientist should know about Python dependencies?
In this blog post Juha continues exploring engineering for data scientists with Dependency Management. So if you are curious about virtual environments and Python dependencies, this is an excellent place to start. -
Docker for Data Science: What every data scientist should know about Docker
Docker is the next layer of engineering practices in our series. Juha Kiili compares Docker to a spacesuit. It keeps your code insulated from the elements and the environment inside is always the same.
