Description

Falcon 180B is a causal decoder-only text generation model from the Technology Innovation Institute, available on Hugging Face as a 180B-parameter model trained on 3,500B tokens of RefinedWeb plus curated corpora.

The model page shows support for multiple languages, transformer-based usage, and deployment options such as vLLM, SGLang, Docker, and Docker Model Runner. Access requires accepting the Falcon-180B TII license and acceptable use policy.

How we innovate

Falcon 180B is distinctive for pairing a very large 180B-parameter decoder-only architecture with multilingual language support and an inference-oriented setup, including guidance for Text Generation Inference and other deployment frameworks.

Use Case / Scenario

1. Text generation

The page lists Falcon 180B as a text generation model, making it suitable for generating and completing natural language text.

2. Research and fine-tuning

The model card describes direct use for research on large language models and as a foundation for further specialization and finetuning.

3. Deployment and local inference

Visible instructions cover Transformers, vLLM, SGLang, Docker, and Docker Model Runner, indicating use in notebooks, local apps, and server-based inference setups.

Visit Website