Tensor Parallelism

---
name: tensor-parallelism
description: Distributing weights across multiple GPUs in a node.
category: model-serving
level: standard
---

# Tensor Parallelism

## Overview

Distributing weights across multiple GPUs in a node.

## When to Use

- Standard workflow tasks related to tensor parallelism.

## When NOT to Use

- Out-of-scope scenarios.

## Prompt Template

```markdown
Execute the tensor-parallelism process targeting the parameters below.
Parameters: {{parameters}}
```

## Evaluation Checklist

- [ ] Task matches structural schemas.
- [ ] Output is clean and formatted.