Exscientia/IgBert

Full fine-tune vs Exscientia/IgBert_unpaired. Fine-tune: 100% of the weights changed, across 30 of 30 layers.

Tensors changed: Embeddings 5/5, Norms 122/122, Attention 240/240, other 123/123

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse