sergeyzh/rubert-base-retriever
Full fine-tune vs sergeyzh/BERTA. Fine-tune: 100% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 5/5, Norms 48/48, Attention 96/96, other 50/50
Full fine-tune vs sergeyzh/BERTA. Fine-tune: 100% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 5/5, Norms 48/48, Attention 96/96, other 50/50