mmBERT Fact-Check Classifier (Merged for Rust)

This is a merged mmBERT model for fact-check classification, optimized for Rust inference using the candle framework.

Model Details

• Base Model: jhu-clsp/mmBERT-base
• Task: Binary classification (fact-check needed vs not needed)
• Languages: Multilingual (1800+ languages)
• Training: LoRA fine-tuned then merged with base model
• Inference: Optimized for Rust candle-binding

Usage with Rust (candle-binding)

use candle_semantic_router::model_architectures::traditional::TraditionalModernBertClassifier;

let classifier = TraditionalModernBertClassifier::load_from_directory("path/to/model", true)?;
let (class, confidence) = classifier.classify_text("The moon is made of cheese")?;
let needs_fact_check = class == 1;

Classes

Training Configuration

• LoRA Rank: 32
• LoRA Alpha: 64
• Epochs: 10
• Batch Size: 64
• Learning Rate: 2e-5

License

Apache 2.0