Sakalti/Saba2-14B-Preview

Full fine-tune vs arcee-ai/SuperNova-Medius. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 48/97, MLP 144/144, Attention 320/336

1 derivatives checked

• mradermacher/Saba2-14B-Preview-GGUF: Quantization

On the Hub · Downloads on Model Pulse