nbeerbower/bruphin-kappa

Full fine-tune vs nbeerbower/bruphin-epsilon. Fine-tune: 94% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 65/65, MLP 90/96, Attention 120/128, lm_head 1/1

1 derivatives checked

• mradermacher/bruphin-kappa-i1-GGUF: Quantization

On the Hub · Downloads on Model Pulse