zelk12/MT1-Gen4-BB-gemma-2-MTg2MTMM2MU-9B

Full fine-tune vs zelk12/MT-Gen2-gemma-2-9B. Fine-tune: 91% of the weights changed, across 42 of 42 layers.

Tensors changed: lm_head 0/1, Embeddings 1/1, Norms 107/169, MLP 126/126, Attention 168/168

1 derivatives checked