Run with local code

Start by creating an account at app.valohai.com

Install the tools

Install the valohai-cli tools by running in your local machine, or where you write your scripts in:

pipx install valohai-cli

# Then use your credentials to login
vh login

Create an project

Create a directory on your local machine

mkdir hello-valohai

Navigate to the folder to create a new Valohai project and link local commands to it. This ensures Valohai knows which project to use when running a job from your machine.

vh project create --name hello-valohai

Create a hello.py

First, create a file called train.py and paste a simple yolov8 example:

import shutil
from ultralytics import YOLO
import json

# Load a model
model = YOLO("yolov8n.pt")  # load a pretrained model (recommended for training)

# Use the model
model.train(data="coco128.yaml", epochs=1, verbose=False)  # train the model
path = model.export(format="onnx")  # export the model to ONNX format

Next, let’s edit the code to copy the generated model to Valohai outputs, and creating an alias, so we can easily reference the file later with the alias “latest-model”.

import shutil
from ultralytics import YOLO
import json

# Load a model
model = YOLO("yolov8n.pt")  # load a pretrained model (recommended for training)

# Use the model
model.train(data="coco128.yaml", epochs=1, verbose=False)  # train the model
path = model.export(format="onnx")  # export the model to ONNX format

# Copy the exported model to the Valohai outputs directory
shutil.copy(path, '/valohai/outputs/')

file_metadata = {
    "valohai.alias": "latest-model"
}

# Attach the metadata to our file
# So we can easily find this find with the alias defined
with open("/valohai/outputs/best.onnx.metadata.json", "w") as outfile:
    f.write(json.dumps(file_metadata))

Create a valohai.yaml

Finally, let’s create a valohai.yaml configuration file to describe our job for Valohai. This configuration will include a step name, a command to guide the execution of our script, environment and a base environment containing Python and YOLOv8 libraries, sourced from a Docker image on hub.docker.com.

Update the environment with the GPU machine you want to run the job on. You can see list of available GPU machines by typing vh environments --gpu. Use the slug name in the YAML.

- step:
    name: yolo
    image: docker.io/ultralytics/ultralytics:8.0.180-python
    command: python train.py
    environment: aws-eu-west-1-p3-2xlarge

Run from the command-line

Finally, execute our training on a remote cloud or on-premises machine by sending the job to the Valohai scheduler. Run the following command in your local command-line:

vh execution run yolo --adhoc --open-browser

A new browser tab will open where you’ll find:

• Valohai execution details, including a snapcode of the code used to run the job.
• All standard logs.
• Upon job completion, you’ll find the trained model in the outputs tab. It will be versioned and uploaded to the designated data store.