copiglet/medgemma-nuslerp-27b

Full fine-tune vs google/gemma-3-27b-it. Fine-tune: 100% of the weights changed, across 62 of 62 layers.

Tensors changed: Norms 484/484, MLP 294/294, Attention 464/464, other 1/1, Embeddings 3/3

2 derivatives checked

• mradermacher/medgemma-nuslerp-27b-i1-GGUF: Quantization (fewer layers)
• mradermacher/medgemma-nuslerp-27b-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse