Fortytwo-Network/Strand-Rust-Coder-14B-v1

Fortytwo-Network/Strand-Rust-Coder-14B-v1 is a 14 billion parameter, Rust-specialized language model fine-tuned from Qwen2.5-Coder-14B. Developed by Fortytwo-Network using Swarm Inference, it excels at Rust-specific programming tasks, achieving 43-48% accuracy on Rust benchmarks, outperforming larger proprietary models. This model is optimized for Rust code generation, completion, and documentation, leveraging a 191K-example peer-validated synthetic dataset.