gss1147/Qwen3-0.6B-rStar-Coder-IF-Expert

Overview

gss1147/Qwen3-0.6B-rStar-Coder-IF-Expert is an 0.8 billion parameter model built on the Qwen3 architecture, developed by gss1147. It is a merged model, created using SLERP (Spherical Linear Interpolation) via mergekit, combining the strengths of two specialized base models: prithivmlmods/rStar-Coder-Qwen3-0.6B for coding expertise and suayptalha/Qwen3-0.6B-IF-Expert for instruction fidelity. This strategic merge aims to deliver superior performance in generating code that precisely follows complex instructions.

Key Capabilities

• Expert Code Generation: Inherits production-grade Python/TypeScript code generation with algorithmic precision from rStar-Coder DNA.
• Precise Instruction Following: Excels at strict adherence to complex instructions and format constraints, derived from IF-Expert DNA.
• Instruction-Compliant Code: Demonstrates emergent strength in producing code that not only functions but also reliably satisfies multiple, detailed constraints within a single generation.

Good for

• Developers needing to generate code that must strictly adhere to multi-part, complex instructions.
• Automating tasks requiring code output with specific formatting or functional constraints.
• Scenarios where both high-quality code and precise instruction adherence are critical, such as generating FastAPI endpoints with multiple validation and output requirements.