powermove72/Qwen2-NextGen-8b

Full fine-tune vs ATH-MaaS/Marco-o1. Fine-tune: 100% of the weights changed, across 28 of 28 layers.

Tensors changed: Norms 55/57, MLP 84/84, Attention 196/196

1 derivatives checked

• mradermacher/Qwen2-NextGen-8b-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse