EssentialAI/eai-distill-0.5b

EssentialAI/eai-distill-0.5b is a 0.5 billion parameter language model, fine-tuned from Qwen2.5-0.5B-Instruct, specifically designed for high-throughput document classification across 12 taxonomic categories. It excels at generating structured metadata for large-scale web document curation and content quality assessment. The model achieves strong agreement with larger teacher models and human annotators, making it suitable for efficient, automated web content analysis.