Hugging Face Overview
Model Overview
alpindale/Llama-3.2-1B-Instruct is a 1.23 billion parameter instruction-tuned model from Meta's Llama 3.2 family, built on an optimized transformer architecture. It is fine-tuned using supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. The model supports a 32768 token context length and was trained on a new mix of publicly available online data, incorporating knowledge distillation from larger Llama 3.1 models.
Key Capabilities
- Multilingual Support: Officially supports English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai, with training on a broader set of languages.
- Dialogue Optimization: Specifically optimized for multilingual dialogue, agentic retrieval, and summarization tasks.
- Instruction Following: Demonstrates strong instruction following capabilities, achieving 59.5 on IFEval.
- Long Context: Features a 32768 token context window, with a 75.0 recall on Needle in Haystack (NIH/Multi-needle).
Good For
- Commercial and Research Use: Intended for a wide range of applications in both commercial products and academic research.
- Assistant-like Applications: Ideal for chat assistants, knowledge retrieval, summarization, and mobile AI-powered writing assistants.
- Constrained Environments: Designed for deployment in environments with limited resources, such as mobile devices, due to its smaller size.