MuXodious/gemma-3n-E4B-it-PaperWitch-heresy

Partial fine-tune vs google/gemma-3n-E4B-it. Partial change: 15% of the weights changed (attention, mlp, other), in 35 of 35 layers.

Tensors changed: Attention 9/363, Embeddings 0/25, Norms 0/489, other 3/687, MLP 35/112

2 derivatives checked

• mradermacher/gemma-3n-E4B-it-PaperWitch-heresy-GGUF: Same weights, converted (fewer layers)
• mradermacher/gemma-3n-E4B-it-PaperWitch-heresy-i1-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse