Nlp

A comprehensive reference guide for selecting and optimizing embedding models for vector search and RAG applications. Covers model comparisons (OpenAI, Voyage, BGE, E5), chunking strategies (token-based, sentence-based, semantic), domain-specific pipelines, and retrieval quality evaluation metrics.

This skill helps you implement comprehensive evaluation strategies for LLM applications. It covers automated metrics like BLEU, ROUGE, and BERTScore for measuring text quality, LLM-as-judge patterns for using stronger models to evaluate outputs, human evaluation frameworks with inter-rater agreement calculations, and statistical A/B testing for comparing model variants with proper significance testing.

This skill should be used when working with pre-trained transformer models for natural language processing, computer vision, audio, or multimodal tasks. Use for text generation, classification, question answering, translation, summarization, image classification, object detection, speech recognition, and fine-tuning models on custom datasets.