Vortex5/Glimmering-Citrus-31B

Full fine-tune vs ConicCat/Gemma4-Garnet-31B. Fine-tune: 94% of the weights changed, across 60 of 60 layers.

Tensors changed: Embeddings 1/4, Norms 0/523, other 0/62, MLP 180/261, Attention 230/338

2 derivatives checked

• mradermacher/Glimmering-Citrus-31B-i1-GGUF: Quantization (fewer layers)
• mradermacher/Glimmering-Citrus-31B-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse