Overview

ertghiu256/Qwen3-4b-tcomanr-merge-v2.2 is a 4 billion parameter language model built upon the Qwen3 architecture, created by ertghiu256. It is a TIES (Task-Independent Embedding Space) merge of several Qwen3 finetunes, with a base of Qwen/Qwen3-4B-Thinking-2507. This model is an iteration of the previous v2 merge, focusing on combining and improving capabilities across code, mathematics, and general reasoning.

Key Capabilities

• Enhanced Reasoning: Designed to improve logical and analytical thinking, particularly in complex scenarios.
• Code Proficiency: Integrates finetunes specifically aimed at strengthening code generation and understanding.
• Mathematical Acumen: Incorporates models optimized for mathematical problem-solving.
• Extended Context: Supports a context length of up to 32768 tokens, allowing for processing longer inputs and maintaining coherence over extended dialogues or documents.

Good For

• Complex Problem Solving: Ideal for applications requiring deep analytical capabilities in technical domains.
• Code Generation and Analysis: Suitable for developers needing assistance with programming tasks.
• Mathematical Computations: Can be utilized for tasks involving numerical reasoning and calculations.
• Research and Development: Provides a strong base for further fine-tuning or integration into specialized AI systems.