Vortex5/Scarlet-Shadow-31B
Full fine-tune vs ReadyArt/Dark-Scarlett-v2.0-31B. Fine-tune: 94% of the weights changed, across 60 of 60 layers.
Tensors changed: Embeddings 1/4, Norms 4/523, other 0/62, MLP 180/261, Attention 230/338
3 derivatives checked
- bartowski/Vortex5_Scarlet-Shadow-31B-GGUF: Same weights, converted (fewer layers)
- mradermacher/Scarlet-Shadow-31B-i1-GGUF: Quantization (fewer layers)
- mradermacher/Scarlet-Shadow-31B-GGUF: Quantization (fewer layers)