Mahesh111000/qwen3-8b-hanabi-rl-base-1to1-24k-step_185
Mahesh111000/qwen3-8b-hanabi-rl-base-1to1-24k-step_185 is an 8.2 billion parameter causal language model from the Qwen3 series, developed by Qwen. This model uniquely supports seamless switching between a 'thinking mode' for complex logical reasoning, math, and coding, and a 'non-thinking mode' for efficient general-purpose dialogue. It excels in reasoning capabilities, human preference alignment, and agentic tasks, supporting over 100 languages with a native context length of 32,768 tokens.
Loading preview...
Qwen3-8B: A Versatile Language Model with Adaptive Reasoning
Mahesh111000/qwen3-8b-hanabi-rl-base-1to1-24k-step_185 is an 8.2 billion parameter causal language model from the Qwen3 series, designed to offer advanced capabilities across various tasks. A key differentiator is its unique ability to seamlessly switch between a 'thinking mode' for complex logical reasoning, mathematics, and code generation, and a 'non-thinking mode' for efficient, general-purpose dialogue. This adaptive approach ensures optimal performance tailored to specific scenarios.
Key Capabilities
- Adaptive Reasoning: Dynamically switches between thinking and non-thinking modes, enhancing performance in complex problem-solving and general conversation.
- Enhanced Reasoning: Demonstrates significant improvements in mathematical, coding, and commonsense logical reasoning compared to previous Qwen models.
- Human Preference Alignment: Excels in creative writing, role-playing, multi-turn dialogues, and instruction following, providing a more natural conversational experience.
- Agentic Expertise: Achieves leading performance among open-source models in complex agent-based tasks, with precise integration with external tools.
- Multilingual Support: Supports over 100 languages and dialects, offering strong multilingual instruction following and translation capabilities.
- Extended Context: Natively handles a context length of 32,768 tokens, extendable up to 131,072 tokens using the YaRN method.
Best Practices for Usage
- Sampling Parameters: Recommended settings vary for thinking (
Temperature=0.6,TopP=0.95,TopK=20,MinP=0) and non-thinking modes (Temperature=0.7,TopP=0.8,TopK=20,MinP=0). Greedy decoding is not recommended for thinking mode. - Output Length: Suggests 32,768 tokens for most queries, and up to 38,912 for highly complex problems.
- Standardized Output: Use specific prompts for math problems (e.g., "Please reason step by step, and put your final answer within \boxed{}") and multiple-choice questions (e.g., JSON structure for answers) to standardize model outputs during benchmarking.