cs-552-2026-barn/group_model

TEXT GENERATIONConcurrency Cost:1Model Size:2BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:May 21, 2026Architecture:Transformer Cold

The cs-552-2026-barn/group_model is a merged language model based on Qwen/Qwen3-1.7B-Base, created using the TIES merge method. This model integrates capabilities from specialized models focusing on general knowledge, mathematics, multilingual understanding, and safety. It is designed to offer a balanced performance across these diverse domains, leveraging a 1.7 billion parameter architecture.

Loading preview...

Model Overview

The cs-552-2026-barn/group_model is a composite language model built upon the Qwen/Qwen3-1.7B-Base architecture. It was developed using the TIES merge method via mergekit, combining the strengths of several specialized models.

Key Capabilities

This model integrates functionalities from four distinct components, aiming for a versatile performance profile:

  • General Knowledge: Incorporates a model focused on broad factual understanding.
  • Mathematics: Includes a component specialized in mathematical reasoning and problem-solving.
  • Multilingual Understanding: Features a model designed for processing and generating text in multiple languages.
  • Safety: Integrates a safety-focused model to enhance responsible AI interactions.

Merge Details

The merge process assigned equal weights (0.25) and a density of 0.5 to each contributing model: cs-552-2026-barn/general_knowledge_model, cs-552-2026-barn/math_model, cs-552-2026-barn/multilingual_model, and cs-552-2026-barn/safety_model. The configuration also specified bfloat16 data type and int8_mask for parameters, with a union tokenizer source.