hon9kon9ize/bert-base-cantonese

Full fine-tune vs indiejoseph/bert-base-cantonese. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: Embeddings 5/5, Norms 50/50, Attention 96/96, other 51/51

On the Hub · Downloads on Model Pulse