Llama 2

Llama 2 is Meta's collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters. These models are designed for a variety of use cases including assistant-like chat applications and code-related tasks.

Hero Image Not Available

Key Features

  • Multiple Model Sizes: 7B, 13B, and 70B parameter versions
  • Open Source: Available for research and commercial use
  • High Performance: State-of-the-art results on many NLP benchmarks
  • Code Capabilities: Strong performance on code understanding and generation
  • Fine-tuned Variants: Chat-optimized versions available
  • Safety Features: Built-in safety measures and responsible AI practices

Technical Specifications

  • Architecture: Transformer-based autoregressive language model
  • Training Data: 2 trillion tokens of publicly available data
  • Context Length: 4,096 tokens
  • Vocabulary: 32,000 tokens using SentencePiece tokenizer

Development Applications

  • Code Completion: Intelligent code suggestions and completions
  • Code Review: Automated code analysis and improvement suggestions
  • Documentation: Generate code documentation and comments
  • Debugging: Assist in identifying and fixing code issues
  • API Integration: Help with API usage and integration patterns
  • Code Translation: Convert code between different programming languages

Model Variants

  • Llama 2: Base pretrained models
  • Llama 2-Chat: Fine-tuned for dialogue and instruction following
  • Code Llama: Specialized version for code generation (separate product)

Licensing

Llama 2 is released under a custom commercial license that allows for both research and commercial use, making it accessible for developers and organizations.