File size: 1,768 Bytes
cef688b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 | ---
language: en
library_name: transformers
pipeline_tag: text-classification
tags:
- bert
- privacy
- text-classification
- gradio
model-index:
- name: JayNightmare/PrERT-CNM-v4-privacybert
results: []
---
# JayNightmare/PrERT-CNM-v4-privacybert
PrERT-CNM v4 PrivacyBERT is a Transformers sequence-classification model prepared from the local checkpoint at `artifacts\phase-3-privacybert\classifier_checkpoint\privacybert`.
## Intended Use
Use this model for text classification in the privacy/CNM workflow it was trained for. It is intended for research and application prototyping unless your own validation shows it is suitable for production use.
## Labels
- `user`
- `system`
- `organization`
## Usage
```python
from transformers import pipeline
classifier = pipeline("text-classification", model="JayNightmare/PrERT-CNM-v4-privacybert", top_k=None)
scores = classifier("Paste text to classify.")
print(scores)
```
## Training Details
- Base architecture: BERT-compatible sequence classifier
- Source checkpoint: `artifacts\phase-3-privacybert\classifier_checkpoint\privacybert`
- Training metadata: included when available in the checkpoint folder
## Evaluation
Add the final held-out metrics before publishing if they are available. Include dataset split details, label distribution, and any thresholding used by downstream consumers.
## Limitations
The model can be sensitive to domain shift, ambiguous language, long inputs, and label definitions that differ from the training data. Review outputs before using them in automated decisions.
## Gradio Demo
The companion Space can be prepared from `huggingface/space` and pointed at this model with the `MODEL_ID` environment variable.
|