virtuous7373/Mira-Demeter

Full fine-tune vs Lambent/Mira-v1.21-27B-rlvr. Fine-tune: 99% of the weights changed, across 62 of 62 layers.

Tensors changed: Embeddings 2/4, Norms 223/484, MLP 197/294, Attention 277/464, other 1/1

2 derivatives checked

• mradermacher/Mira-Demeter-i1-GGUF: Quantization (fewer layers)
• mradermacher/Mira-Demeter-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse