---
license: apache-2.0
base_model: sriram1983007/SRA-RiskGate-4B
base_model_relation: quantized
library_name: gguf
pipeline_tag: text-generation
tags:
- qwen3
- gguf
- stablecoin
- risk-gate
- payments
- llama-cpp
---

# SRA-RiskGate-4B — GGUF

GGUF quantizations of `sriram1983007/SRA-RiskGate-4B`, a Qwen3-4B fine-tune for stablecoin payment risk gating, produced with llama.cpp.

## Available files

| File | Quantization | Description |
|---|---|---|
| `model-F16.gguf` | F16 | Full precision (~8 GB) |
| `model-Q8_0.gguf` | Q8_0 | 8-bit |
| `model-Q6_K.gguf` | Q6_K | 6-bit |
| `model-Q5_K_M.gguf` | Q5_K_M | 5-bit medium |
| `model-Q5_K_S.gguf` | Q5_K_S | 5-bit small |
| `model-Q5_0.gguf` | Q5_0 | 5-bit |
| `model-Q4_K_M.gguf` | Q4_K_M | 4-bit medium (recommended) |
| `model-Q4_K_S.gguf` | Q4_K_S | 4-bit small |
| `model-Q4_0.gguf` | Q4_0 | 4-bit |
| `model-Q3_K_L.gguf` | Q3_K_L | 3-bit large |
| `model-Q3_K_M.gguf` | Q3_K_M | 3-bit medium |
| `model-Q3_K_S.gguf` | Q3_K_S | 3-bit small |
| `model-Q2_K.gguf` | Q2_K | 2-bit |

## Usage
   
With llama.cpp:

```bash
./llama-cli -m model-Q4_K_M.gguf -p "Your prompt" -n 256
```

**Prompt format:** the model only performs as benchmarked with its training prompt (a system prompt plus `<context>`, `<payload>` and `<tool_results>` blocks) at temperature 0. See the [main model card's Quickstart](https://huggingface.co/sriram1983007/SRA-RiskGate-4B). Q2_K and Q3 builds have not been evaluated.

Or any GGUF-compatible runtime (Ollama, LM Studio, koboldcpp, etc.).

## License

Apache License 2.0. Built on `Qwen/Qwen3-4B-Instruct-2507`.

## Disclaimer

Payment risk-gate model. Evaluate before use in production payments.