⚡ SopuCoder Fast (v1.1.0)
SopuCoder Fast is an ultra-low-latency coding model created and maintained by Md. Saiful Islam (Saif), utilizing Alibaba's Qwen3.5 4B foundation model, optimized for real-time IDE inline completions, instant syntax diagnostics, and low-spec laptop execution.
🚀 Key Specifications
- Creator & Maintainer: Md. Saiful Islam (Saif)
- Base Architecture: Alibaba Qwen3.5 4B
- Context Length: 16,384 tokens
- Inference Speed: ~22.4 - 25+ tokens/second on Apple Silicon
- Fine-Tuned Standalone Candidate (1.5B): 68.7 tokens/second at 1.09 GB RAM
- Target Stacks: Real-time IDE inline completions, quick function generation, lint fixes, tool calling.
🦙 Quick Run with Ollama
# Run latest release
ollama run sopu175/sopu-coder-fast:v1.1.0
# Or run latest tag
ollama run sopu175/sopu-coder-fast:latest
🛠️ VS Code & Cursor IDE Autocomplete Configuration
Using the Continue extension (~/.continue/config.json):
{
"tabAutocompleteModel": {
"title": "SopuCoder Fast",
"provider": "ollama",
"model": "sopu175/sopu-coder-fast:v1.1.0"
}
}
📄 License & Attribution
- Creator Attribution: Created and maintained by Md. Saiful Islam (Saif).
- Foundation Model: Built upon Alibaba's Qwen3.5 4B (Apache License 2.0).
- SopuCoder Engine: Licensed under the Apache License 2.0.