malteos/most-embed-de
Partial fine-tune vs nvidia/Nemotron-3-Embed-1B-BF16. Partial change: 76% of the weights changed (attention, mlp, norms), in 16 of 16 layers.
Tensors changed: Embeddings 0/1, Norms 1/33, MLP 48/48, Attention 64/64
Partial fine-tune vs nvidia/Nemotron-3-Embed-1B-BF16. Partial change: 76% of the weights changed (attention, mlp, norms), in 16 of 16 layers.
Tensors changed: Embeddings 0/1, Norms 1/33, MLP 48/48, Attention 64/64