Metrics
- Bert_score
- Perplexity
- Perplexity_a
- Perplexity_chat
- Perplexity_q
- Rag
- Reward
- Sentence_bert
- accuracy
- bleu
- char_edit_dist_accuracy
- f1_macro
- f1_macro_multi_label
- f1_micro
- f1_micro_multi_label
- f1_weighted
- kpa
- map
- matthews_correlation
- mrr
- ndcg
- ner
- normalized_sacrebleu
- precision_macro_multi_label
- precision_micro_multi_label
- recall_macro_multi_label
- recall_micro_multi_label
- regard
- retrieval_at_k
- rouge
- rouge_with_confidence_intervals
- sacrebleu
- safety
- spearman
- squad
- string_containment
- token_overlap
- token_overlap_with_context
- wer