cl-nagoya/ruri-v3-pt-130m
Full fine-tune vs sbintuitions/modernbert-ja-130m. Fine-tune: 100% of the weights changed, across 19 of 19 layers.
Tensors changed: Embeddings 2/2, Norms 38/38, Attention 38/38, MLP 38/38
2 derivatives checked
- cl-nagoya/ruri-v3-130m: Full fine-tune
- hotchpotch/japanese-reranker-base-v2: Full fine-tune