---
language:
  - en
tags:
  - deckergui
  - token-classification
  - entity-recognition
  - ecosystem
  - gate-prediction
license: mit
pipeline_tag: token-classification
---

# DGUI-GatePredictor

Token classification model for predicting DGM gate outcomes in the DeckerGUI ecosystem.

## Model Details

- **Model Type:** Token Classification (Sequence Labeling)
- **Training Data:** deckergui-seed-status-events, deckergui-agent-coordination
- **Architecture:** DistilBERT-base with custom classification head
- **Labels:** gate_pass, gate_block, quota_exceeded, key_revoked, scope_violation
- **DeckerGUI Version:** v2.0.0

## Usage

```python
from transformers import pipeline

classifier = pipeline("token-classification", model="ctaxnagomi/DGUI-GatePredictor")

result = classifier("Agent dgui-emitter requested context.bundle with key 0xb553de32")
print(result)
```

## Training

Trained on synthetic DeckerGUI ecosystem data:
- 120 seed status events
- 180 agent coordination records
- Gate decision patterns from CaaS federation

## Citation

```bibtex
@software{deckergui_gate_predictor_2026,
  title={DGUI-GatePredictor: Token Classification for DGM Gate Outcomes},
  author={Wan Mohd Azizi bin Wan Hosen},
  year={2026},
  url={https://huggingface.co/ctaxnagomi/DGUI-GatePredictor}
}
```