NexesMess/Llama_3.1_70b_Luminhermoblivion_V1.21

Full fine-tune vs invisietch/L3.1-70Blivion-v0.1-rc1-70B. Fine-tune: 100% of the weights changed, across 80 of 80 layers.

Tensors changed: Norms 85/161, MLP 240/240, Attention 320/320

2 derivatives checked

On the Hub · Downloads on Model Pulse