farbodtavakkoli/OTel-LLM-32B-IT

Hugging Face
TEXT GENERATIONConcurrency Cost:2Model Size:32BQuant:FP8Ctx Length:32kPublished:Feb 11, 2026License:apache-2.0Architecture:Transformer Open Weights Warm

OTel-LLM-32B-IT by farbodtavakkoli is a 32 billion parameter language model fine-tuned from allenai/OLMo-3-32B, specifically specialized for the telecommunications domain. It was trained on extensive telecom-focused data curated by over 100 domain experts from various institutions. This model is designed to power Retrieval-Augmented Generation (RAG) pipelines for telecommunications, excelling at generating accurate, context-grounded responses by declining to answer when context is insufficient.

Loading preview...

OTel-LLM-32B-IT: A Specialized Telecom Language Model

OTel-LLM-32B-IT is a 32 billion parameter language model developed by farbodtavakkoli, fine-tuned from the allenai/OLMo-3-32B base model. It is a core component of the OTel Family of Models, an open-source initiative aimed at creating industry-standard AI for the global telecommunications sector. The model underwent full parameter fine-tuning using a comprehensive dataset curated by over 100 domain experts from institutions like Yale University, GSMA, NetoAI, Khalifa University, University of Leeds, and The University of Texas at Dallas.

Key Capabilities

  • Telecom Domain Specialization: Trained on a vast array of telecom-specific data, including arXiv papers, 3GPP standards, GSMA documents, IETF RFCs, industry whitepapers, and O-RAN specifications.
  • Retrieval-Augmented Generation (RAG) Optimized: Designed to work within RAG pipelines, complementing OTel Embedding and Reranker models to provide accurate, grounded responses.
  • Abstention Training: Features built-in abstention training, meaning it will decline to answer if it does not receive sufficient context, thereby minimizing hallucinations and ensuring context-grounded generation.
  • Open-Source License: Released under the Apache 2.0 license, promoting broad usability and integration.

Ideal Use Cases

  • Telecommunications RAG Systems: Powering end-to-end RAG pipelines for querying and generating responses based on telecom specifications, standards, and documentation.
  • Context-Grounded Information Retrieval: Generating accurate answers to questions within the telecom domain, relying heavily on provided context.
  • Specialized Telecom Applications: Developing AI solutions that require deep understanding and generation capabilities specific to the telecommunications industry.