@if (showTabs) { } @if (step === 1) {
See the pre-loaded examples to quickly get familiar with ClearML’s various capabilities.
Browse "Pro Tips" in the Help menu to jump start your work flow.
Check out the ClearML docs for advanced information and in depth how-to’s.
} @else {
@if (queue) {
{{entityName}} have been enqueued on the {{queue?.name}} queue, which is currently not serviced by any worker. They will remain in the 'pending' state until a ClearML worker services this queue.
} @if (!currentLink || currentLink === links[0]) { @for (step of steps; track step.id) {
{{step.header}}
{{step.title}}
@if (step.title==='2. Configure') { }
{{isJupyter? 'Set the ClearML environment for your notebook':'Run the ClearML setup script'}}
{{step.code}}
@if (step.subNote || (isJupyter && ((queue && $index === 2) || (!queue && $index === 1)))) {
@if ((isJupyter && ((queue && $index === 2) || (!queue && $index === 1)))) { @if (accessKey) { Manage your app credentials in the workspace settings page } @else { Fill in your API credentials (if you don’t have any, create new credentials) } } @else { {{step.subNote}} } {{step.subNote}}
}
}
@if (!isJupyter) {
Complete the clearml configuration information as prompted.
} @if (!credentialsCreated) {
}
api { @if (credentialsComment) {
# {{credentialsComment}}
}
web_server : {{WEB_SERVER_URL}}
api_server: {{configService.apiServerUrl()}}
@if (configService.fileServerUrl()) {
files_server: {{configService.fileServerUrl()}}
}
credentials {
"access_key"="{{accessKey}}"
"secret_key"="{{secretKey}}"
}
}
@if (!showTabs) {
Manage your app credentials in the workspace settings page
}
} @if (!currentLink || currentLink === links[1]) { @if (!queue) {
@if (!showTabs) {
3. Integrate
}
Add ClearML to your code. For example:
@if (!showTabs) {
from {{gettingStartedContext?.packageName || 'clearml'}} import Task
task = Task.init(project_name="my project", task_name="my task")
} @else {
import numpy as np
import matplotlib.pyplot as plt
# Add the following two lines to your code, to have ClearML automatically log your experiment
from clearml import Task
task = Task.init(project_name='My Project', task_name='My Experiment')
# Create a plot using matplotlib, or you can also use plotly
plt.scatter(np.random.rand(50), np.random.rand(50), c=np.random.rand(50), alpha=0.5)
# Plot will be reported automatically to clearml
plt.show()
# Report some scalars
for i in range(100):
task.get_logger().report_scalar(title="graph title", series="linear", value=i*2, iteration=i)
}
} } @if (!queue && (currentLink === links[2])) {
1. Clone a previously run experiment
Use the “Clone” action to create a new draft copy.
You can change any of the new experiments’ configurations.
Experiments table
2. Run your experiment
Use the “Enqueue” action to enqueue the new experiment on an available queue.
To have an agent service a queue, on your target machine run:
clearml-agent daemon —queue <queue name>
} @if (!queue && (currentLink === 'videoTab')) {
} @if (queue) {
}
}