Llama

What is Llama?

Llama (Large Language Model Meta AI) is Meta’s family of foundation language models spanning multiple sizes and generations, widely used as open weights for research, fine-tuning, and application development.

Self-hosting Llama still requires safety and quality evaluation: jailbreaks, hallucinations, and tool misuse do not disappear with open weights. Test each fine-tune and deploy config.

Related Giskard articles

• Practical guide on AI security
• Giskard OSS v3
• Continuous LLM red teaming

Evaluate Llama deployments with Giskard

Run quality and red-team suites on each Llama base or fine-tune before production. Learn more.

Further reading

Authoritative reference: Meta Llama.