Surpem/Supertron2.1-0.6B
Supertron2.1-0.6B by Surpem is a compact, instruction-tuned causal language model built on Qwen3-0.6B, featuring 0.8 billion parameters and a 32768-token context length. Designed for efficiency on consumer hardware, it excels in reasoning, math, coding, and general knowledge tasks. This model serves as a lightweight generalist for assistant-style conversations and structured problem-solving.
Loading preview...
Supertron2.1-0.6B: Compact and Efficient Instruction-Tuned Model
Supertron2.1-0.6B, developed by Surpem, is an instruction-tuned language model based on Qwen3-0.6B. With 0.8 billion parameters and a 32768-token context length, it is engineered for efficiency, making it suitable for consumer hardware. The model retains the Qwen3 architecture, tokenizer, and chat format, ensuring compatibility with standard transformers workflows.
Key Capabilities
- Reasoning: Designed for structured reasoning, breaking down problems, comparing options, and explaining tradeoffs.
- Math: Assists with arithmetic, algebra, word problems, and step-by-step explanations.
- Coding: Capable of writing, debugging, and explaining code across multiple languages including Python, JavaScript, C++, and Java.
- General Knowledge: Explains concepts across STEM, history, business, and technology, providing summaries and technical explanations.
- Instruction Following: Adapts to natural language instructions and requested output formats like bullet lists, tables, and JSON.
Good For
- Lightweight assistant experiments and local coding assistance.
- Math practice, explanations, and general question answering.
- Summarization, rewriting, and prototype agent workflows.
- Educational and research applications requiring a compact generalist model.