Sorawiz/KunouSky-32B

Full fine-tune vs NovaSky-AI/Sky-T1-32B-Preview. Fine-tune: 94% of the weights changed, across 64 of 64 layers.

Tensors changed: Embeddings 0/1, Norms 128/129, MLP 190/192, Attention 258/448, lm_head 0/1

2 derivatives checked

• mradermacher/KunouSky-32B-i1-GGUF: Quantization
• mradermacher/KunouSky-32B-GGUF: Quantization

On the Hub · Downloads on Model Pulse