Run Llama in the browser
Before You Start
This will run a language model directly in your browser using WebGPU. Here's what to expect:
Downloads ~2GB once
Uses 2-4GB RAM
Requires WebGPU
The model will be cached locally for future use. This only works in Chrome/Edge on desktop.
...
Anti-Pattern: Ai Washing
Overview
Brief description of this antipattern and why it matters.
The Problem
- What specific issue does this address?
- Why do people fall into this trap / miss this opportunity?
- What are the symptoms you see?
The Solution / What Not To Do
Key Points
- Main principle or warning
- Specific techniques or pitfalls
- When to apply / avoid this approach
Examples
Good Example
Concrete example showing the right way
Bad Example
Concrete example showing what not to do
Why This Works / Why This Fails
- Underlying reasons
- Connection to how LLMs actually work
- Broader implications
Related Patterns
- Links to other relevant patterns/anti-patterns
- See also: Other Pattern
Summary
- Key takeaways
- When to remember this pattern
- Red flags to watch for