ArmBench-TextEmbed
Metric-AI-Lab/ArmBench-TextEmbed
This repo hosts the code for evaluating text embedding models on Armenian Benchmark (ArmBench).
Summary
ArmBench-TextEmbed is a specialized Python benchmark for evaluating the performance of text embedding models on the Armenian language. It provides standardized testing across semantic textual similarity (STS), passage retrieval, and reranking tasks, integrating with the MTEB framework. The tool supports any Hugging Face model, features configurable prompts, and outputs detailed performance metrics.