Llama 2
Meta·
LLMsopen-weightWorkstation
Overview
First commercially licensed open LLM. Made open-weight models viable for business use and kickstarted the open-source LLM ecosystem.
Capabilities and innovations
7B / 13B / 70B parametersCommercial license4K context windowChat-tuned variants (Llama 2-Chat)Grouped-Query Attention (GQA) in 70BGhost Attention (GAtt) for multi-turnRLHF with rejection sampling
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| MMLU | 68.9% | unsourced |
Architecture and hardware
- Parameters
- 70B total (Dense)
- Estimated VRAM at Q4
- ~41 GB, Workstation class
- Quantization formats
- GGUF, GPTQ, AWQ
- Recommended runtime
- Ollama
- License
- Llama 2 Community License
Links
More from Meta
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.