Sakalti/Neptuno-3B

Partial fine-tune vs aayanmishra-ml/Athena-1-3B. Partial change: 90% of the weights changed (attention, mlp), in 36 of 36 layers.

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

3 derivatives checked

• mradermacher/Neptuno-3B-i1-GGUF: Quantization
• mradermacher/Neptuno-3B-GGUF: Same weights, converted
• Sakalti/Neptuno-Alpha: Partial fine-tune

On the Hub · Downloads on Model Pulse