lynaNSFW/DaSiWa_MMH3_HybridTurbo_V3

Full fine-tune vs lynaNSFW/DaSiWa_MiniMax_H3. Fine-tune: 100% of the weights changed, across 50 of 50 layers.

Tensors changed: other 112/114, Norms 42/210, Attention 204/304, MLP 204/304

On the Hub · Downloads on Model Pulse