Rigour Deep Analysis Model (lite v5)
Fine-tuned Qwen3.5-0.8B for code quality analysis. Trained via RLAIF: strong teacher labels + structural verification + DPO.
| Property | Value |
|---|---|
| Base | Qwen3.5-0.8B |
| Quant | Q4_K_M |
| Size | 379MB |
| Tier | lite |
| Version | 5 |
| Date | 2026-03-07 |
Usage with Rigour
Update types.ts in rigour-core to point to this model,
or place the GGUF file in ~/.rigour/models/.
Built by: rigour-labs/driftbench RLAIF pipeline
- Downloads last month
- 4
Hardware compatibility
Log In to add your hardware
4-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support