tsmatz/xlm-roberta-ner-japanese
Partial fine-tune vs FacebookAI/xlm-roberta-base. Partial change: 31% of the weights changed (attention, embeddings, norms, other), in 12 of 12 layers.
Tensors changed: Embeddings 3/5, Norms 48/48, Attention 96/96, other 48/48