Dataset and models for "ESNLIR: Expanding Spanish NLI Benchmarks with Multi-genre and Causal Annotation".