datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
tiny-aya-global-finance-evalgoldfish-crosslingual-cka
Cross-lingual representational alignment (CKA) of the Goldfish bilingual merge/train models
Reproduction of the §3 Representational Alignment protocol of the Goldfish cross-lingual paper, applied to the 111 bilingual merge/train checkpoints under suchirsalhan/* created 2025-12.
Generated 2026-08-26 19:17 UTC. Code: scripts/ in this repo.
TL;DR — read this before the numbers
The uploaded checkpoints do not contain the merges. All 111 repos ship a LoRA adapter over… See the full description on the dataset page: https://huggingface.co/datasets/suchirsalhan/goldfish-crosslingual-cka.nayana-beir-eval-multilang_v7_crosslingual_only