YOON1v/Apex-2

Full fine-tune vs YOON1v/Apex-2-Base. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 1/1, Norms 129/129, MLP 1568/1568, Attention 128/128

1 derivatives checked

• LiviiYS/Apex-2-apex: Quantization

On the Hub · Downloads on Model Pulse