ertghiu256/Qwen3-4b-tcomanr-merge-v2.5
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Sep 18, 2025Architecture:Transformer0.0K Cold

ertghiu256/Qwen3-4b-tcomanr-merge-v2.5 is a 4 billion parameter language model based on the Qwen3-4B-Thinking-2507 architecture, specifically designed for enhanced reasoning, coding, and mathematical capabilities. This model integrates multiple Qwen3 4B fine-tuned models to create a robust and versatile solution. It excels in text generation, coding, math, and multi-turn conversations, offering specialized chat templates for controlled chain-of-thought reasoning.

Loading preview...

Overview

ertghiu256/Qwen3-4b-tcomanr-merge-v2.5 is an upgraded 4 billion parameter model built upon the Qwen3-4B-Thinking-2507 base. It is a merge of various Qwen3 4B fine-tuned models, created using the TIES method, to enhance its performance across multiple domains. The model is particularly strong in reasoning, code generation, and mathematical problem-solving, making it a versatile tool for diverse applications.

Key Capabilities

  • Enhanced Reasoning: Integrates advanced chain-of-thought reasoning modes (/think, /shortthink, /nothink) via a new chat template, allowing for controlled depth of reasoning.
  • Code Generation: Combines capabilities from several code-focused Qwen3 4B models.
  • Mathematical Tasks: Optimized for mathematical problem-solving through the inclusion of math-reasoner models.
  • Multi-turn Conversations: Designed to handle complex conversational flows effectively.
  • Versatile Applications: Suitable for general text generation, coding, and mathematical tasks.

Recommended Use Cases

This model is ideal for developers and researchers requiring a compact yet powerful model for:

  • Applications demanding explicit reasoning steps.
  • Code generation and analysis tasks.
  • Solving mathematical problems.
  • Building interactive chatbots that require robust multi-turn conversational abilities.