AhmedZaky1/DIMI-embedding-matryoshka-arabic

Full fine-tune vs aubmindlab/bert-base-arabertv02. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: Embeddings 5/5, Norms 48/48, Attention 96/96, other 49/50

1 derivatives checked

• s1nju/darija-embedding-model: Full fine-tune

On the Hub · Downloads on Model Pulse