Model Overview

mylesgoose/Llama-3.2-1B-Instruct-abliterated is a 1.23 billion parameter instruction-tuned model from Meta's Llama 3.2 family, designed for multilingual text-in/text-out generative tasks. It utilizes an optimized transformer architecture with Grouped-Query Attention (GQA) and was fine-tuned using Supervised Fine-Tuning (SFT) and Reinforcement Learning with Human Feedback (RLHF) for helpfulness and safety. The model supports a substantial 128k context length and was trained on up to 9 trillion tokens, with a knowledge cutoff of December 2023.

Key Capabilities

• Multilingual Dialogue: Optimized for assistant-like chat and agentic applications in multiple languages, including English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.
• Agentic Use Cases: Excels in knowledge retrieval, summarization, mobile AI-powered writing assistants, and query/prompt rewriting.
• Performance: Outperforms many open-source and closed chat models on common industry benchmarks, particularly in multilingual contexts.
• Efficiency: The 1B and 3B models are designed for deployment in constrained environments, such as mobile devices.

Good For

• Developing multilingual chatbots and virtual assistants.
• Implementing agentic systems for information retrieval and text summarization.
• Applications requiring efficient, instruction-tuned language generation in resource-constrained settings.
• Research into safety fine-tuning and robust model deployment.