Lambent/Iris-12B-v1.4

Full fine-tune vs Lambent/Iris-12B-v1.3.3. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: Embeddings 12/12, Norms 193/289, other 0/48, MLP 144/144, Attention 184/184

3 derivatives checked