Outcome
diff --git a/docs/how-to-guides/support/index.md b/docs/how-to-guides/support/index.md
new file mode 100644
index 0000000..7cef255
--- /dev/null
+++ b/docs/how-to-guides/support/index.md
@@ -0,0 +1,9 @@
+---
+title: Support
+description: Submit support tickets and get help.
+---
+
+# Support
+
+- [Submit a ticket from the CLI](/how-to-guides/support/submit-ticket-cli)
+- [Submit a ticket from the UI](/how-to-guides/support/submit-ticket-ui)
diff --git a/docs/how-to-guides/teams/tutorials.mdx b/docs/how-to-guides/teams/tutorials.mdx
index 8b6cddd..a86fc5f 100644
--- a/docs/how-to-guides/teams/tutorials.mdx
+++ b/docs/how-to-guides/teams/tutorials.mdx
@@ -29,7 +29,7 @@ sidebar_label: Team setup walkthrough
## Create a team, add members, and grant cluster access
-This tutorial walks through setting up a team with controlled access to a cluster.
+This walkthrough covers setting up a team with controlled access to a cluster.
**Prerequisites:** Admin or Team Admin permissions in your organization, a connected cluster.
diff --git a/docs/how-to-guides/workbench/observability.mdx b/docs/how-to-guides/workbench/observability.mdx
index ce23048..d29c069 100644
--- a/docs/how-to-guides/workbench/observability.mdx
+++ b/docs/how-to-guides/workbench/observability.mdx
@@ -45,7 +45,7 @@ The top of the page shows six tiles:
| Spend MTD | Month-to-date spend with an end-of-month forecast |
| Idle GPU hours | GPU hours where utilization was below 5%; the metric to drive down |
-If **Idle GPU hours** exceeds 20% of **Spend today**, sessions are likely being left open without active use.
+If idle GPU-hour cost exceeds 20% of today's spend, sessions are likely being left open without active use.
## Charts
diff --git a/docs/how-to-guides/workbench/serve/canary-rollouts.mdx b/docs/how-to-guides/workbench/serve/canary-rollouts.mdx
index 2377cf9..1969ec9 100644
--- a/docs/how-to-guides/workbench/serve/canary-rollouts.mdx
+++ b/docs/how-to-guides/workbench/serve/canary-rollouts.mdx
@@ -56,7 +56,7 @@ Success looks like this: the endpoint detail page shows the current canary perce
-The current Workbench UI displays canary traffic state but does not expose promote or rollback buttons on the endpoint detail page. Use your team's release process or API workflow to change rollout state.
+The current Workbench UI displays canary traffic state but does not expose promote or rollback buttons on the endpoint detail page. Use your team's release process or the [API reference](/reference/developer-tools/api) to change rollout state.
diff --git a/docs/how-to-guides/workbench/train/pipelines/creating.mdx b/docs/how-to-guides/workbench/train/pipelines/creating.mdx
index f8a093e..7b886c3 100644
--- a/docs/how-to-guides/workbench/train/pipelines/creating.mdx
+++ b/docs/how-to-guides/workbench/train/pipelines/creating.mdx
@@ -6,7 +6,7 @@ description: Build a multi-step pipeline from the Vantage SDK and view it in the
# Create a pipeline
:::warning
-Pipelines are in early access. UI authoring is not yet available. Use the SDK to create pipelines.
+Pipelines are in preview. UI authoring is not yet available. Use the SDK to create pipelines.
:::
Pipelines are multi-step DAGs (directed acyclic graphs) that turn one-off scripts into reproducible, schedulable workflows. Each step runs in its own container with explicit inputs, outputs, and dependencies.
diff --git a/docs/how-to-guides/workbench/troubleshooting.mdx b/docs/how-to-guides/workbench/troubleshooting.mdx
index f5cb38a..392b329 100644
--- a/docs/how-to-guides/workbench/troubleshooting.mdx
+++ b/docs/how-to-guides/workbench/troubleshooting.mdx
@@ -27,7 +27,7 @@ By default the Logs tab streams the trainer master pod. If your master finished
## Pipeline and sweep limitations
-Both tabs are in **Preview**. Today:
+Both tabs are in **preview**. Today:
- Pipeline authoring is API-first. The UI shows experiments, pipelines, runs, recurring runs, artifacts, and executions.
- Sweep creation is available from the UI when algorithms, runtimes, and compute pools are configured. The list and detail views show trials and sweep progress.
diff --git a/docs/index.md b/docs/index.md
index 1376c97..6cfb4ac 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -6,7 +6,7 @@ slug: /
# Vantage Compute Documentation
-**Virtually Limitless™ Compute:** Execute AI, HPC, and Quantum workloads across any environment with a unified platform for Slurm+Kubernetes orchestration.
+**Virtually Limitless™ Compute:** Execute AI, HPC, and Quantum workloads across any environment with a unified platform for Slurm and Kubernetes orchestration.
[Tutorial](/tutorial) · [How-to Guides](/how-to-guides/cloud-accounts/get-started) · [Explanation](/explanation) · [Book a Demo](https://vantagecompute.ai/contact)
diff --git a/docs/reference/clusters/kubernetes/manage.mdx b/docs/reference/clusters/kubernetes/manage.mdx
index 99ebbf4..728f516 100644
--- a/docs/reference/clusters/kubernetes/manage.mdx
+++ b/docs/reference/clusters/kubernetes/manage.mdx
@@ -13,8 +13,8 @@ A Kubernetes cluster has two status indicators:
| Status | Meaning |
|---|---|
-| **CONNECTED** | The cluster agent is online and communicating normally |
-| **NOT CONNECTED** | The agent is unreachable. Check network connectivity or agent health |
+| **Connected** | The cluster agent is online and communicating normally |
+| **Not Connected** | The agent is unreachable. Check network connectivity or agent health |
**Deploy status** shows the current lifecycle state of the cluster:
@@ -108,6 +108,10 @@ Grafana is automatically deployed when the cluster is created.
## Adding compute pools post-creation
+:::tip
+For step-by-step instructions, see [Manage a Kubernetes cluster](/how-to-guides/clusters/manage-cluster).
+:::
+
1. Open the cluster detail page and go to **Workloads**, then **Compute Pools** in the cluster sidebar.
2. Click **Add Compute Pool**.
3. Configure the pool: name, role, instance type (or profile for non-AWS), min/max size, and optional labels/taints.
@@ -123,6 +127,10 @@ The Applications page is a read-only view of what is deployed on the cluster. It
## Deleting a cluster
+:::tip
+For step-by-step instructions, see [Manage a Kubernetes cluster](/how-to-guides/clusters/manage-cluster).
+:::
+
1. Open the cluster detail page.
2. Click the overflow menu (three dots) in the top-right corner.
3. Select **Delete Cluster** and confirm.
diff --git a/docs/reference/clusters/kubernetes/node-groups.mdx b/docs/reference/clusters/kubernetes/node-groups.mdx
index e356dd1..fe6d2a8 100644
--- a/docs/reference/clusters/kubernetes/node-groups.mdx
+++ b/docs/reference/clusters/kubernetes/node-groups.mdx
@@ -81,6 +81,10 @@ For AWS clusters, the autoscaler manages EC2 Fleet instances tagged with `vantag
## Best practices
+:::tip
+These are general recommendations. See the [Kubernetes how-to guides](/how-to-guides/clusters) for step-by-step procedures.
+:::
+
- **Separate workload types into different compute pools:** GPU training jobs and CPU preprocessing should use different pools so they don't compete for resources.
- **Set min sizes conservatively:** Idle nodes cost money. Start with min=0 and adjust once you understand your workload patterns.
- **Use labels and taints for scheduling control:** Mark GPU compute pools with `nvidia.com/gpu: "true"` so only GPU workloads land on them.
diff --git a/docs/reference/clusters/slurm/manage.mdx b/docs/reference/clusters/slurm/manage.mdx
index df4e981..a2fab7e 100644
--- a/docs/reference/clusters/slurm/manage.mdx
+++ b/docs/reference/clusters/slurm/manage.mdx
@@ -11,16 +11,16 @@ A Slurm cluster moves through these statuses:
| Status | Meaning |
|---|---|
-| `preparing` | Provisioning in progress. Vantage is creating cloud resources (if applicable) and waiting for nodes to register. |
-| `ready` | Cluster is connected, Slurm configuration is uploaded, and the cluster is accepting jobs. |
-| `failed` | Provisioning or runtime error. Check `creation_status_details` on the detail page for the specific error. |
-| `deleting` | Cluster teardown in progress. Vantage is deprovisioning cloud resources and removing the database record. |
+| **Preparing** | Provisioning in progress. Vantage is creating cloud resources (if applicable) and waiting for nodes to register. |
+| **Ready** | Cluster is connected, Slurm configuration is uploaded, and the cluster is accepting jobs. |
+| **Failed** | Provisioning or runtime error. Check the detail page for the specific error. |
+| **Deleting** | Cluster teardown in progress. Vantage is deprovisioning cloud resources and removing the database record. |
Transitions:
-- `preparing` to `ready`: All nodes registered and Slurm config uploaded.
-- `preparing` to `failed`: CloudFormation error, provisioning timeout, or node registration failure.
-- `ready` to `deleting`: User initiated deletion.
-- `failed` to `deleting`: User initiated deletion.
+- **Preparing** to **Ready**: All nodes registered and Slurm config uploaded.
+- **Preparing** to **Failed**: CloudFormation error, provisioning timeout, or node registration failure.
+- **Ready** to **Deleting**: User initiated deletion.
+- **Failed** to **Deleting**: User initiated deletion.
## Detail page sidebar
diff --git a/docs/reference/clusters/slurm/partitions.mdx b/docs/reference/clusters/slurm/partitions.mdx
index e503660..663f7a0 100644
--- a/docs/reference/clusters/slurm/partitions.mdx
+++ b/docs/reference/clusters/slurm/partitions.mdx
@@ -63,6 +63,10 @@ Cloud Slurm partitions use Vantage's autoscaler to manage node counts:
## Best practices
+:::tip
+These are general recommendations. See the [Slurm how-to guides](/how-to-guides/clusters) for step-by-step procedures.
+:::
+
- **Separate workloads by resource profile:** Create CPU and GPU partitions so batch preprocessing doesn't block GPU nodes for training.
- **Set max node counts conservatively:** Idle nodes cost money. Start low and raise when you've validated your workload patterns.
- **Use the default partition:** Mark one partition as default so jobs that don't specify a partition still get scheduled.
diff --git a/docs/reference/developer-tools/cli/commands.mdx b/docs/reference/developer-tools/cli/commands.mdx
index 3901b76..c31f20a 100644
--- a/docs/reference/developer-tools/cli/commands.mdx
+++ b/docs/reference/developer-tools/cli/commands.mdx
@@ -43,6 +43,10 @@ This document provides a comprehensive reference for all available CLI commands
```
+:::note
+`vdeployer-web` (Vantage's internal deployment service) handles background deployment operations for clusters. It is referenced in several commands below, including `cluster extend`, `cluster kubeflow create`, and `cluster slurm create`.
+:::
+
## Authentication Commands
### Login
diff --git a/docs/reference/developer-tools/cli/index.md b/docs/reference/developer-tools/cli/index.md
index 63fbfd7..f508626 100644
--- a/docs/reference/developer-tools/cli/index.md
+++ b/docs/reference/developer-tools/cli/index.md
@@ -22,9 +22,11 @@ Or from source:
git clone https://github.com/vantagecompute/v8x
cd v8x
uv sync
-uvx v8x --help
+uv run v8x --help
```
+> **`uvx v8x` vs `uv run v8x`:** Use `uvx v8x` (shown above) to run the published package without a permanent install. Use `uv run v8x` only when running from a local source checkout. For a persistent install, run `uv tool install v8x`, which makes `v8x` available as a standalone command.
+
#### Authenticate
Authenticate against the Vantage platform using the `login` command.
@@ -33,7 +35,7 @@ Authenticate against the Vantage platform using the `login` command.
uvx v8x login
```
-#### Create a Multipass Singlenode Cluster
+#### Create a Multipass Single-node Cluster
Create a cloud account first, then create the cluster:
@@ -67,8 +69,8 @@ uvx v8x cluster create my-slurm-microk8s-cluster \
### Next Steps
-- [CLI Quickstart](/how-to-guides/developer-tools/cli/quickstart) – Install & Configure
-- [Commands Reference](/reference/developer-tools/cli/commands) – Complete Command Reference
-- [Private Installation Configuration](/how-to-guides/developer-tools/cli/private-vantage-installation) – Partner Vantage Deployment CLI Profile Configuration
-- [Usage Examples](/how-to-guides/developer-tools/cli/usage) – Practical Command Patterns
-- [Troubleshooting](/how-to-guides/developer-tools/cli/troubleshooting) – Common Issues and Solutions
+- [CLI Quickstart](/how-to-guides/developer-tools/cli/quickstart) - Install and Configure
+- [Commands Reference](/reference/developer-tools/cli/commands) - Complete Command Reference
+- [Private Installation Configuration](/how-to-guides/developer-tools/cli/private-vantage-installation) - Partner Vantage Deployment CLI Profile Configuration
+- [Usage Examples](/how-to-guides/developer-tools/cli/usage) - Practical Command Patterns
+- [Troubleshooting](/how-to-guides/developer-tools/cli/troubleshooting) - Common Issues and Solutions
diff --git a/docs/reference/licenses/reference.md b/docs/reference/licenses/reference.md
index 56421ec..67eda2e 100644
--- a/docs/reference/licenses/reference.md
+++ b/docs/reference/licenses/reference.md
@@ -16,6 +16,8 @@ description: License server ports, protocols, and configuration fields.
| DSLS | Varies by vendor |
| OLicense | 7374 |
+
+
## Registration fields
When adding a license server in Vantage, these fields are available:
diff --git a/docs/reference/settings/iam/reference.md b/docs/reference/settings/iam/reference.md
index 09a1d14..302c34a 100644
--- a/docs/reference/settings/iam/reference.md
+++ b/docs/reference/settings/iam/reference.md
@@ -5,17 +5,19 @@ description: Permission groups, authentication options, and API key settings.
# IAM, Reference
+
+
## Built-in permission groups
| Group | Permissions |
|---|---|
| Regular User | Standard platform access, create and manage own resources |
-| Full Admin | Full platform administration, users, billing, all settings |
+| Full Admin | Full platform administration, users, billing, all settings. Platform-wide access across all organizations. |
| Cluster Admin | Manage clusters, create, update, delete, monitor |
| Jobs Admin | Manage jobs and scripts, create, submit, cancel |
| Licenses Admin | Manage license servers, add, configure, monitor |
| Notebook Admin | Manage Workbench notebook sessions |
-| Organization Admin | Manage organization settings, billing, and members |
+| Organization Admin | Manage organization settings, billing, and members. Scoped to a single organization. |
| Team Admin | Manage teams, create, add/remove members, configure access |
## Authentication methods
diff --git a/docs/reference/workbench/compute-profiles/reference.mdx b/docs/reference/workbench/compute-profiles/reference.mdx
index 9b58301..4f40325 100644
--- a/docs/reference/workbench/compute-profiles/reference.mdx
+++ b/docs/reference/workbench/compute-profiles/reference.mdx
@@ -7,6 +7,8 @@ description: Every field on a compute pool.
Compute pools are managed by workspace admins. These are the fields you'll see on each pool.
+{/* TODO(vantage): verify -- g5.48xlarge has A10G GPUs, not A100. Should this be p4d.24xlarge? */}
+
| Field | Description |
|---|---|
| Display name | What people see, e.g. `A100 x 4`. |
diff --git a/docs/reference/workbench/endpoints/presets.mdx b/docs/reference/workbench/endpoints/presets.mdx
index ecd413a..596c51f 100644
--- a/docs/reference/workbench/endpoints/presets.mdx
+++ b/docs/reference/workbench/endpoints/presets.mdx
@@ -5,6 +5,10 @@ description: Pre-configured endpoint defaults for quick deployment.
# Presets
+:::info
+Not to be confused with [Session presets](/reference/workbench/presets), which define IDE, image, and compute defaults for interactive sessions.
+:::
+
Endpoint presets are reusable configurations that bundle compute, replica counts, and runtime selections. Instead of filling out the full endpoint form every time, pick a preset and the deployment form pre-fills around it.
## The presets list
diff --git a/docs/reference/workbench/endpoints/reference.mdx b/docs/reference/workbench/endpoints/reference.mdx
index 99e1589..0ec30b8 100644
--- a/docs/reference/workbench/endpoints/reference.mdx
+++ b/docs/reference/workbench/endpoints/reference.mdx
@@ -11,8 +11,10 @@ Endpoint URL pattern, auth, and key fields.
The detail page header shows the URL. Auth is handled by your workspace's ingress, pass your Vantage API key as a bearer token. The full URL pattern is:
+{/* TODO(vantage): confirm the actual inference API hostname */}
+
```
-POST https://{workspace}.vantage.ai/v1/{endpoint-name}/infer
+POST https://{workspace}.vantagecompute.ai/v1/{endpoint-name}/infer
Authorization: Bearer {your-api-key}
```
diff --git a/docs/reference/workbench/index.mdx b/docs/reference/workbench/index.mdx
index 0c69a1f..759b26e 100644
--- a/docs/reference/workbench/index.mdx
+++ b/docs/reference/workbench/index.mdx
@@ -15,6 +15,7 @@ Workbench is where you build, train, and serve machine-learning models on Vantag
- **[Remote Desktop](/reference/workbench/remote-desktop):** VNC-based remote desktop sessions for full GUI access to GPU nodes.
- **[PVC Viewer](/reference/workbench/pvc-viewer):** Browser-based file browser for PersistentVolumeClaims.
- **[Compute Profiles](/reference/workbench/compute-profiles):** The reusable shape of your compute: GPU type, count, autoscaling bounds, instance class.
+{/* TODO(vantage): confirm Apple-silicon nodes are supported for cloud training */}
- **[Training Jobs](/reference/workbench/training-jobs):** Distributed training on PyTorch, DeepSpeed, or MLX runtimes. Retry, suspend, resume.
- **[Pipelines](/reference/workbench/pipelines):** Multi-step DAGs that orchestrate ingestion, training, evaluation, and deployment.
- **[Sweeps](/reference/workbench/sweeps):** Hyperparameter search with Bayesian, grid, or random algorithms, tracked end-to-end.
diff --git a/docs/reference/workbench/models/reference.mdx b/docs/reference/workbench/models/reference.mdx
index 6efd19f..f161ab0 100644
--- a/docs/reference/workbench/models/reference.mdx
+++ b/docs/reference/workbench/models/reference.mdx
@@ -5,6 +5,8 @@ description: "Model fields and metadata."
# Reference
+{/* TODO(vantage): expand -- add versioning/format/size fields, explain Labels vs Custom properties */}
+
Every model carries these fields beyond name and version.
| Field | Description |
diff --git a/docs/reference/workbench/pipelines/runs.mdx b/docs/reference/workbench/pipelines/runs.mdx
index 4473bdb..3fbf9c0 100644
--- a/docs/reference/workbench/pipelines/runs.mdx
+++ b/docs/reference/workbench/pipelines/runs.mdx
@@ -17,10 +17,20 @@ The list view shows every run across all experiments. Each row displays:
- **Name:** the run's display name.
- **Pipeline:** which pipeline version this run executes.
-- **Status:** current phase (Running, Succeeded, Failed, Skipped).
+- **Status:** current phase (Pending, Running, Succeeded, Failed, Skipped).
- **Started:** when the run started.
- **Duration:** elapsed time.
+## Run statuses
+
+| Status | Meaning |
+|---|---|
+| **Pending** | The run is queued and waiting for resources to become available. |
+| **Running** | The run is actively executing pipeline steps. |
+| **Succeeded** | All steps completed successfully. |
+| **Failed** | One or more steps errored. Check step-level logs for details. |
+| **Skipped** | The run was bypassed due to a condition evaluation or manual cancellation. |
+
## Run detail
Click a run to see:
diff --git a/docs/reference/workbench/presets/index.mdx b/docs/reference/workbench/presets/index.mdx
index 6ae2202..494a064 100644
--- a/docs/reference/workbench/presets/index.mdx
+++ b/docs/reference/workbench/presets/index.mdx
@@ -5,6 +5,10 @@ description: Reusable templates that define which images, compute sizes, and sto
# Presets
+:::info
+Not to be confused with [Endpoint presets](/reference/workbench/endpoints/presets), which define compute and replica defaults for inference endpoints.
+:::
+
A **preset** (WorkspaceKind) is a reusable template that bundles the choices you make when creating a session, which IDE, which container images, which compute sizes, and which storage volumes are available. Instead of picking images and pod sizes from scratch every time, you select a preset and the wizard pre-fills around it.
Presets appear across Workbench in three contexts:
diff --git a/docs/reference/workbench/sessions/lifecycle.mdx b/docs/reference/workbench/sessions/lifecycle.mdx
index 67fe14d..00295fc 100644
--- a/docs/reference/workbench/sessions/lifecycle.mdx
+++ b/docs/reference/workbench/sessions/lifecycle.mdx
@@ -36,6 +36,8 @@ Click any row name to open it. The detail page is split into:
- **Activity:** chronological log of phase transitions, conditions, and reconcile errors.
- **Observability:** embedded Grafana panel scoped to this session.
+{/* TODO(vantage): verify this cost figure is current; consider moving to a pricing page */}
+
:::warning
-**Idle sessions cost money.** A 4×A100 session left running overnight is roughly $400. Workbench will surface an "idle" warning, but it's on you to pause sessions you're not actively using.
+**Idle sessions cost money.** A 4xA100 session left running overnight is roughly $400. Workbench will surface an "idle" warning, but it is on you to pause sessions you are not actively using.
:::
\ No newline at end of file
diff --git a/docs/reference/workbench/sessions/notebooks.md b/docs/reference/workbench/sessions/notebooks.md
index a5d2715..7811112 100644
--- a/docs/reference/workbench/sessions/notebooks.md
+++ b/docs/reference/workbench/sessions/notebooks.md
@@ -4,6 +4,8 @@ title: Notebooks
description: Jupyter, VS Code, and RStudio notebook environments.
---
+
+
Vantage Workbench sessions provide JupyterLab, VS Code, and RStudio environments running on managed compute infrastructure. Launch a session, pick your compute pool, and start coding in minutes.
## Quickstart
diff --git a/docs/reference/workbench/sessions/remote-desktops.md b/docs/reference/workbench/sessions/remote-desktops.md
index af46c3f..88b0e24 100644
--- a/docs/reference/workbench/sessions/remote-desktops.md
+++ b/docs/reference/workbench/sessions/remote-desktops.md
@@ -3,6 +3,8 @@ title: Remote Desktops
description: Remote desktop access to Vantage compute resources.
---
+
+
Vantage provides remote desktop sessions (X2Go/Linux) for interactive GUI access to your compute resources. Launch a desktop session from the Workbench and connect from anywhere.
## Getting started
diff --git a/docs/reference/workbench/sweeps/index.mdx b/docs/reference/workbench/sweeps/index.mdx
index 9ac46d1..72b5926 100644
--- a/docs/reference/workbench/sweeps/index.mdx
+++ b/docs/reference/workbench/sweeps/index.mdx
@@ -17,6 +17,10 @@ A sweep is a hyperparameter-search experiment that runs many *trials*, each tria
- **Presets:** pre-defined configurations for common search patterns.
- **Trials:** individual parameter combinations within a sweep, each running as a training job.
+## See also
+
+- [Training Jobs](/reference/workbench/training-jobs): each trial runs as a training job
+
## Next steps
- [Algorithms](/explanation/workbench/sweep-algorithms): the supported search strategies
diff --git a/docs/reference/workbench/sweeps/presets.mdx b/docs/reference/workbench/sweeps/presets.mdx
index 0dc7a90..de39300 100644
--- a/docs/reference/workbench/sweeps/presets.mdx
+++ b/docs/reference/workbench/sweeps/presets.mdx
@@ -8,7 +8,7 @@ description: Hyperparameter search presets for sweep configuration.
Sweep presets are pre-defined configurations for common hyperparameter search patterns. They combine an algorithm with a set of parameter ranges so you can start a sweep without manually configuring every parameter.
:::caution[Preview]
-The Sweeps tab is in early access. Preset browsing is available in the UI, but sweep creation is API-only via the Vantage SDK.
+Sweeps are in preview. Preset browsing and preset-based sweep creation are available in the UI. Sweeps can also be created programmatically via the Vantage SDK.
:::
## The presets list
diff --git a/docs/reference/workbench/sweeps/trials.mdx b/docs/reference/workbench/sweeps/trials.mdx
index a0071be..bc3b538 100644
--- a/docs/reference/workbench/sweeps/trials.mdx
+++ b/docs/reference/workbench/sweeps/trials.mdx
@@ -26,7 +26,11 @@ Click a trial to see its full parameter set, metric history, and logs.
## Best trial
-The sweep detail page highlights the **best trial:** the one with the highest (or lowest, depending on the objective) metric value. This is the configuration you'd use for your production model.
+The sweep detail page highlights the **best trial:** the one with the highest (or lowest, depending on the [objective direction](/explanation/workbench/sweep-algorithms)) metric value. This is the configuration you would use for your production model.
+
+## See also
+
+- [Training Jobs](/reference/workbench/training-jobs): each trial runs as a training job
## Next steps
diff --git a/docs/reference/workbench/training-jobs/lifecycle.mdx b/docs/reference/workbench/training-jobs/lifecycle.mdx
index 7fc0578..c50e5e8 100644
--- a/docs/reference/workbench/training-jobs/lifecycle.mdx
+++ b/docs/reference/workbench/training-jobs/lifecycle.mdx
@@ -18,5 +18,9 @@ Pending to Running to Succeeded/Failed and the actions at each phase.
Lifecycle actions:
Suspend,
Resume,
Retry (clones the job and re-runs from scratch),
Delete.
:::warning
-If `active=N, ready
-## Create your first cluster.
+## Create your first cluster
Provision a new Slurm cluster:
This step uses the `slurm-multipass` app, which provisions a Slurm cluster inside a Multipass VM. Make sure [Multipass is installed](https://multipass.run/install) before running this command.
diff --git a/sidebars-main.js b/sidebars-main.js
index f1cb9da..5a4d9eb 100644
--- a/sidebars-main.js
+++ b/sidebars-main.js
@@ -139,7 +139,7 @@ module.exports = {
doc('how-to-guides/licenses/dsls/high-availability', 'High availability'),
doc('how-to-guides/licenses/dsls/troubleshooting', 'Troubleshooting'),
]),
- category('LM-X', 'how-to-guides/licenses/lmx/introduction', [
+ category('LMX', 'how-to-guides/licenses/lmx/introduction', [
doc('how-to-guides/licenses/lmx/server-setup', 'Server setup'),
doc('how-to-guides/licenses/lmx/vantage-managed-server-setup', 'Vantage-managed setup'),
doc('how-to-guides/licenses/lmx/user-managed-server-setup', 'User-managed setup'),
@@ -148,13 +148,14 @@ module.exports = {
doc('how-to-guides/licenses/lmx/troubleshooting', 'Troubleshooting'),
]),
category('LS-DYNA', 'how-to-guides/licenses/ls-dyna/introduction', [
+ doc('how-to-guides/licenses/ls-dyna/server-setup', 'Server setup'),
doc('how-to-guides/licenses/ls-dyna/vantage-managed-server-setup', 'Vantage-managed setup'),
doc('how-to-guides/licenses/ls-dyna/user-managed-server-setup', 'User-managed setup'),
doc('how-to-guides/licenses/ls-dyna/monitoring', 'Monitoring'),
doc('how-to-guides/licenses/ls-dyna/high-availability', 'High availability'),
doc('how-to-guides/licenses/ls-dyna/troubleshooting', 'Troubleshooting'),
]),
- category('Olicense', 'how-to-guides/licenses/olicense/introduction', [
+ category('OLicense', 'how-to-guides/licenses/olicense/introduction', [
doc('how-to-guides/licenses/olicense/server-setup', 'Server setup'),
doc('how-to-guides/licenses/olicense/vantage-managed-server-setup', 'Vantage-managed setup'),
doc('how-to-guides/licenses/olicense/user-managed-server-setup', 'User-managed setup'),