nomic-ai/nomic-embed-text-v2-moe-unsupervised

Full fine-tune vs nomic-ai/nomic-xlm-2048. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: other 2/2, Embeddings 2/2, Attention 48/48, MLP 24/24, Norms 48/48

1 derivatives checked

• nomic-ai/nomic-embed-text-v2-moe: Full fine-tune

On the Hub · Downloads on Model Pulse