adaptive-classifier/document-type

1---2language: multilingual3tags:4- adaptive-classifier5- text-classification6- continuous-learning7license: apache-2.08---9 10# Adaptive Classifier11 12This model is an instance of an [adaptive-classifier](https://github.com/codelion/adaptive-classifier) that allows for continuous learning and dynamic class addition.13 14You can install it with `pip install adaptive-classifier`.15 16## Model Details17 18- Base Model: answerdotai/ModernBERT-base19- Number of Classes: 520- Total Examples: 2521- Embedding Dimension: 76822 23## Class Distribution24 25```26contract: 5 examples (20.0%)27email: 5 examples (20.0%)28invoice: 5 examples (20.0%)29memo: 5 examples (20.0%)30report: 5 examples (20.0%)31```32 33## Usage34 35```python36from adaptive_classifier import AdaptiveClassifier37 38# Load the model39classifier = AdaptiveClassifier.from_pretrained("adaptive-classifier/model-name")40 41# Make predictions42text = "Your text here"43predictions = classifier.predict(text)44print(predictions)  # List of (label, confidence) tuples45 46# Add new examples47texts = ["Example 1", "Example 2"]48labels = ["class1", "class2"]49classifier.add_examples(texts, labels)50```51 52## Training Details53 54- Training Steps: 1355- Examples per Class: See distribution above56- Prototype Memory: Active57- Neural Adaptation: Active58 59## Limitations60 61This model:62- Requires at least 3 examples per class63- Has a maximum of 150 examples per class64- Updates prototypes every 10 examples65 66## Citation67 68```bibtex69@software{adaptive_classifier,70  title = {Adaptive Classifier: Dynamic Text Classification with Continuous Learning},71  author = {Sharma, Asankhaya},72  year = {2025},73  publisher = {GitHub},74  url = {https://github.com/codelion/adaptive-classifier}75}76```77