True2456/gemma-4-12b-it-qat-4bit-frontierdistill

Full fine-tune vs mlx-community/gemma-4-12B-it-qat-4bit. Fine-tune: 94% of the weights changed, across 48 of 48 layers.

Tensors changed: Embeddings 0/20, Norms 0/289, other 0/48, MLP 431/432, Attention 550/552

On the Hub · Downloads on Model Pulse