farbodtavakkoli/OTel-LLM-20B-IT
OTel-LLM-20B-IT is a 20 billion parameter language model developed by Farbod Tavakkoli, fine-tuned from openai/gpt-oss-20b. It is specifically specialized for the telecommunications domain, trained on extensive telecom-focused data curated by industry experts and institutions. This model is designed to power Retrieval-Augmented Generation (RAG) pipelines, providing accurate, context-grounded responses within the telecom sector.
Loading preview...
OTel-LLM-20B-IT: A Specialized Telecom Language Model
OTel-LLM-20B-IT is a 20 billion parameter language model, fine-tuned from the openai/gpt-oss-20b base model. It is a core component of the OTel Family of Models, an open-source initiative by Farbod Tavakkoli aimed at developing industry-standard AI for the global telecommunications sector. The model was trained using full parameter fine-tuning on a comprehensive dataset curated by over 100 domain experts, including data from arXiv telecom papers, 3GPP standards, GSMA documents, IETF RFCs, and O-RAN specifications.
Key Capabilities
- Telecom Domain Specialization: Highly proficient in understanding and generating content related to telecommunications, leveraging a unique dataset from various institutional partners.
- RAG Pipeline Integration: Designed to function as the generative component within a Retrieval-Augmented Generation (RAG) pipeline, working alongside OTel Embedding and Reranker models.
- Context-Grounded Generation: Optimized for generating accurate responses based on provided context, with abstention training to decline answers when insufficient context is available, preventing hallucination.
Intended Use Cases
- Answering technical queries within the telecommunications domain.
- Summarizing and extracting information from telecom specifications, standards, and documentation.
- Supporting knowledge retrieval systems for telecom professionals and researchers.
- Developing specialized AI applications for the global telecommunications industry.