cs-552-2026-baseline/math_model

Qwen3-1.7B: A Dual-Mode Reasoning Model

Qwen3-1.7B is a 1.7 billion parameter causal language model from the Qwen series, designed with a unique dual-mode operation. It features a 'thinking mode' for intricate logical reasoning, mathematics, and coding, and a 'non-thinking mode' for general, efficient dialogue. This allows the model to adapt its processing for optimal performance across diverse tasks.

Key Capabilities

• Enhanced Reasoning: Significantly improves performance in mathematics, code generation, and commonsense logical reasoning compared to previous Qwen models.
• Flexible Operation: Seamlessly switches between thinking and non-thinking modes, which can be controlled via enable_thinking parameter or dynamic /think and /no_think tags in user prompts.
• Agentic Expertise: Excels in tool-calling capabilities, integrating precisely with external tools in both modes, and achieving leading performance in complex agent-based tasks among open-source models.
• Multilingual Support: Supports over 100 languages and dialects with strong capabilities for instruction following and translation.
• Human Preference Alignment: Delivers natural, engaging, and immersive conversational experiences, excelling in creative writing, role-playing, and multi-turn dialogues.

Good For

• Complex Problem Solving: Ideal for applications requiring deep logical reasoning, such as advanced mathematical problems or intricate coding challenges.
• Dynamic Conversational Agents: Suitable for building agents that need to switch between analytical processing and general conversation based on user input.
• Multilingual Applications: Effective for tasks involving instruction following or translation across a wide range of languages.
• Creative and Role-Playing Scenarios: Excels in generating engaging and contextually appropriate responses for creative writing and interactive role-play.