Sakalti/Neptuno-Alpha

Partial fine-tune vs Sakalti/Neptuno-3B. Partial change: 82% of the weights changed (attention, mlp), in 36 of 36 layers.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/73, MLP 108/108, Attention 145/252

2 derivatives checked

• mradermacher/Neptuno-Alpha-i1-GGUF: Quantization
• mradermacher/Neptuno-Alpha-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse