zelk12/MT1-Gen5-BMM-gemma-2-9B
Partial fine-tune vs zelk12/MT1-Gen5-BB-gemma-2-S4Av4d-9B. Partial change: 78% of the weights changed (attention, mlp, norms), in 40 of 42 layers.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 117/169, MLP 120/126, Attention 160/168
2 derivatives checked
- mradermacher/MT1-Gen5-BMM-gemma-2-9B-GGUF: Same weights, converted
- zelk12/MT1-Gen5-BMMIMU-gemma-2-9B: Full fine-tune