Download
“Llama” is the repository from Meta (formerly Facebook/Meta Research) containing the inference code for LLaMA (Large Language Model Meta AI) models. It provides utilities to load pre-trained LLaMA model weights, run inference (text generation, chat, completions), and work with tokenizers. Tokenizer utilities, download scripts, shell helpers to fetch model weights with correct licensing/permissions. Includes example scripts for chat completions and text completions to show how to call the models in code. This repo is a core piece of the Llama model infrastructure, used by researchers and developers to run LLaMA models locally or in their infrastructure. It is meant for inference (not training from scratch) and connects with aspects like model cards, responsible use, licensing, etc.
Features
- Provides reference code to load various LLaMA pre-trained weights (7B, 13B, 70B, etc.) and perform inference (chat or completion)
- Tokenizer utilities, download scripts, shell helpers to fetch model weights with correct licensing / permissions
- Support for multi-parameter setups (batch size, context length, number of GPUs / parallelism) to scale to larger models / machines
- License / Responsible Use guidance; a model card and documentation for how the model may be used or restricted
- Includes example scripts for chat completions and text completions to show how to call the models in code
- Compatibility with standard deep learning frameworks (PyTorch etc.) for inference usage, including ensuring the required dependencies and setup scripts are included
Project Samples
