Explore models

Blog posts

YouTube

VESSL AI

Examples

Reference

Get Started

Talk to Sales

Create a new pipeline

Launch and scale AI workloads without the hassle of managing infrastructure

Overview

Launch a barebone GPU-accelerated workload

Quickstart – Hello, world!

Launch a Jupyter Notebook server with an SSH connection

GPU-accelerated notebook

Launch an interactive web app for Stable Diffusion

Stable Diffusion playground

Fine-tune Llama 3.1 8B with instruction datasets

Llama 3.1 fine-tuning

Serve & deploy vLLM-accelerated Llama3-8B

Llama 3 deployment

Serve & deploy with Text Generation Inference (TGI)

Serverless deployment

Create a new run

Manage runs

Tracking runs

Create a new workspace

Explore workspaces

Create an SSH connection

Download & attach datasets

Run a server application

Create your own custom image in Workspace on VESSL AI

Build a custom image

Tips & Limitations

Create a new service

Manage API keys (Provisioned Mode)

Migrate from Service Beta

Pipeline revisions

Pipeline steps

Pipeline variables

Manage pipeline executions

Create & manage pipeline triggers

Create a new experiment

Manage experiments

Monitor experiment results

Run distributed training jobs

Run local experiments

Create a new sweep

Monitor sweeps

Monitor clusters

Create a new model

Deploy models

Create a dataset

Manage datasets

Create a new Organization

Project summary

Create a new project

Project-level repo & datasets

Changelog

See the full list of compute options and pricing

Pricing & Compute

See VESSL Run in action with common use cases and workflows

Use cases

Use VESSL Service to create a Slack Summarizer Bot with Open-Source Llama 3

Create a Slack Summarizer Bot with VESSL Service

Leverge the power of GPUs to efficiently train batch runs

Launch batch jobs on GPUs

Enable a real-time session of interacitve run on GPUs.

Spin-up a notebook server on GPUs

Backup & Restore data with Artifacts

Use VESSL AI Python SDK to log and track runs

Track runs with vessl.log()

Import, export, and mount your data in your cloud object storage like AWS S3 or GCP GCS.

Use object storage with VESSL Run

VESSL Run is configured through a single YAML file.

Run YAML

VESSL Serve is configured through a single YAML file.

Serve YAML

Cheat Sheet

vessl cluster

vessl configure

vessl dataset

vessl experiment

vessl image

vessl model

vessl organization

vessl project

vessl run

vessl service

vessl ssh-key

vessl sweep

vessl volume

vessl workspace

Keras

Tensorboard

Transformers

vessl.configure

vessl.init

vessl.log

vessl.Image

vessl.Audio

vessl.hp.update

vessl.progress

vessl.upload

vessl.finish

Cluster

Dataset

Experiment

Image

Model

Organization

Project

RunnerBase

Service

SSH-key

Sweep

Volume

Workspace

Send a request

Create an asynchronous request

Fetch output of an asynchronous request

Get Started

Compute

Resource

Admin

Resources

Create a new pipeline