farbodtavakkoli/OTel-LLM-14B-IT

OTel-LLM-14B-IT: A Specialized Telecom Language Model

OTel-LLM-14B-IT is a 14 billion parameter language model developed by farbodtavakkoli, specifically fine-tuned for the telecommunications sector. Built upon the Qwen/Qwen3-14B base model, it leverages the OTel-LLM dataset, a comprehensive collection of telecom data curated by over 100 domain experts. This model is a key component of the OTel Family of Models, an open-source effort to provide AI resources for global telecommunications.

Key Capabilities

• Context-Grounded Telecom QA: Optimized for generating accurate answers to telecom questions when provided with relevant context, making it ideal for RAG applications.
• Improved Correctness: Demonstrates a +5.0 percentage point increase in LLM-as-judge correctness over its base model on held-out OTel evaluation partitions.
• Domain-Specific Training: Fine-tuned on a high-quality dataset derived from arXiv telecom papers, 3GPP standards, GSMA documents, IETF RFCs, O-RAN specifications, and industry whitepapers.
• Full-Parameter Fine-Tuning: Utilizes full-parameter post-training for enhanced domain adaptation.

Good for

• Retrieval-Augmented Generation (RAG) Systems: Excellent for building RAG pipelines that require precise, context-aware answers to telecommunications queries.
• Telecom-Specific Applications: Ideal for use cases involving technical documentation, standards interpretation, and operational support within the telecom industry.
• Research and Development: Provides a strong foundation for further research into AI applications in telecommunications.