SriRamanaAtmic/AtmicEmbeddingv3

Partial fine-tune vs SriRamanaAtmic/AtmicEmbeddingv2. Partial change: 13% of the weights changed (attention, norms, other), in 6 of 24 layers.

Tensors changed: Embeddings 0/5, Norms 24/96, Attention 48/192, other 24/98

On the Hub · Downloads on Model Pulse