farbodtavakkoli/OTel-LLM-12B-IT

OTel-LLM-12B-IT: A Specialized Telecom Language Model

OTel-LLM-12B-IT is a 12 billion parameter language model developed by farbodtavakkoli, specifically fine-tuned for the telecommunications sector. It is built upon the google/gemma-3-12b-it base model and is part of the OTel Family of Models, an open-source initiative providing AI resources for global telecommunications. This model excels in context-grounded answer generation within Retrieval-Augmented Generation (RAG) pipelines, demonstrating a notable improvement of +4.5 percentage points in correctness over its base model on OTel-LLM held-out evaluation data.

Key Capabilities

• Context-Grounded Telecom QA: Optimized for generating accurate answers based on provided telecommunications context.
• Enhanced Correctness: Achieves 82.8% correctness in LLM-as-judge evaluations for telecom-specific tasks.
• Specialized Training Data: Fine-tuned on a curated dataset of 326,767 high-confidence telecom examples from sources like arXiv papers, 3GPP standards, GSMA documents, and O-RAN specifications.
• Full-Parameter Fine-Tuning: Utilizes full-parameter post-training for comprehensive adaptation to the telecom domain.

Good for

• Retrieval-Augmented Generation (RAG) in Telecom: Ideal for applications requiring precise, context-aware answers from telecom-specific documents.
• Domain-Specific AI Development: A valuable resource for building AI solutions tailored to the telecommunications industry.
• Research and Development: Provides a strong baseline for further research into specialized language models for technical domains.