VALE: A Multimodal Visual and Language Explanation Framework for Image Classifiers using eXplainable AI and Language Models