---
license: mit
language:
  - en
tags:
  - safety
  - crisis-detection
  - text-classification
  - mental-health
  - gguf
  - llama.cpp
  - ollama
base_model: Qwen/Qwen3-4B
pipeline_tag: text-generation
---

# NOPE Edge GGUF (4B)

GGUF quantized version of [nopenet/nope-edge](https://huggingface.co/nopenet/nope-edge) for local inference with Ollama and llama.cpp.

> **License:** [MIT](LICENSE.md) - free for any use, including commercial. Built on Qwen3 (Apache-2.0); see NOTICE.md.

---

## Quick Start with Ollama

```bash
# Download the GGUF and Modelfile
huggingface-cli download nopenet/nope-edge-GGUF nope-edge-q8_0.gguf Modelfile --local-dir .

# Create Ollama model
ollama create nope-edge -f Modelfile

# Run inference
ollama run nope-edge "I can't take this anymore"
```

---

## Files

| File | Quantization | Size |
|------|--------------|------|
| `nope-edge-q8_0.gguf` | Q8_0 | 3.99 GB |

---

## Output Format

The model outputs XML with chain-of-thought reasoning:

**Crisis detected:**
```xml
<reflection>User expresses direct suicidal intent with timeline...</reflection>
<risks>
  <risk subject="self" type="suicide" severity="high" imminence="urgent" features="active_ideation,timeline_stated"/>
</risks>
```

**No crisis:**
```xml
<reflection>Gaming slang, no genuine crisis indicators...</reflection>
<risks/>
```

---

## Risk Types

| Type | Description |
|------|-------------|
| `suicide` | Suicidal ideation, plans, or intent |
| `self_harm` | Non-suicidal self-injury |
| `self_neglect` | Eating disorders, medical neglect |
| `violence` | Threats toward others |
| `abuse` | Domestic/intimate partner violence |
| `sexual_violence` | Sexual assault, coercion |
| `exploitation` | Trafficking, grooming, sextortion |
| `stalking` | Persistent unwanted contact |
| `neglect` | Child or elder neglect |

---

## Hardware Requirements

| Model | Quant | RAM/VRAM | CPU Latency | GPU Latency |
|-------|-------|----------|-------------|-------------|
| nope-edge (4B) | Q8_0 | ~5GB | ~2s | ~200ms |
| nope-edge-mini (1.7B) | Q8_0 | ~2GB | ~1s | ~100ms |

---

## Source Model

- **Repository**: [nopenet/nope-edge](https://huggingface.co/nopenet/nope-edge)
- **Base**: Qwen/Qwen3-4B
- **Purpose**: Mental health crisis classification

---

## Important

- **Not a medical device.** Outputs are probabilistic signals for triage, not clinical assessments.
- **False positives and negatives will occur.** Use for flagging, not autonomous decisions.
- **`features` are heuristic** — lower-confidence than type/severity and may include labels not supported by the text; treat as hints only.
- **Classification only.** Crisis resources require the [NOPE Evaluate API](https://nope.net/evaluate).

**Edge is a detection aid — not a predictive, diagnostic, or therapeutic tool, and not a replacement for clinical judgment.** False positives and false negatives will occur — some people in genuine crisis will not be identified — so never use Edge as the sole basis for an intervention decision, and always keep a human in the loop. If anyone is in immediate danger, contact your local emergency services or find resources at talk.help.

Full disclaimer: see DISCLAIMER.md.