farbodtavakkoli/OTel-LLM-270M-IT

Hugging Face
TEXT GENERATIONConcurrent Unit Cost:1Model Size:0.27BQuant:BF16Context Size:32kPublished:Feb 11, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

The farbodtavakkoli/OTel-LLM-270M-IT is a 270 million parameter context-grounded telecom language model, full-parameter fine-tuned by farbodtavakkoli on OTel telecommunications data. Based on google/gemma-3-270m-it, this model is specifically optimized for generating accurate answers within Retrieval-Augmented Generation (RAG) pipelines using provided telecom context. It demonstrates improved context-grounded correctness, achieving a +9.0 percentage point increase over its base model on OTel evaluation splits. This model is intended for specialized telecom answer generation rather than general-purpose language tasks.

Loading preview...

OTel-LLM-270M-IT: Context-Grounded Telecom Language Model

OTel-LLM-270M-IT is a 270 million parameter language model developed by farbodtavakkoli, specifically fine-tuned on a curated dataset of telecommunications data. Part of the OTel Family of Models, it aims to provide open-source AI resources for the global telecom sector. This model is a full-parameter fine-tuned version of google/gemma-3-270m-it.

Key Capabilities & Differentiators

  • Domain-Specific Expertise: Optimized for telecommunications, trained on data curated by 100+ domain experts from sources like arXiv telecom papers, 3GPP standards, GSMA documents, and O-RAN specifications.
  • Context-Grounded Correctness: Achieves a significant improvement of +9.0 percentage points in LLM-as-judge correctness over its base model when generating answers grounded in retrieved context.
  • RAG Pipeline Integration: Designed for use in Retrieval-Augmented Generation (RAG) pipelines, where it generates answers based on provided telecom context.
  • Apache 2.0 License: Released under an Apache 2.0 license, promoting open use and development.

Intended Use Cases

  • Telecom Answer Generation: Ideal for generating precise, context-grounded answers to telecom-related queries within RAG systems.
  • Specialized QA: Suitable for applications requiring deep knowledge of telecommunications, such as technical support, documentation analysis, or network management assistance.

Limitations

  • Domain-Specific: Not intended as a general-purpose language model; performance outside telecommunications is not guaranteed.
  • English-Only: The current release is primarily text-centric and English-only.
  • Context-Dependent: Optimized for context-grounded generation; not designed for unrestricted, context-free question answering.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p