| --- |
| 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} |
| } |
| ``` |
| |
