Cyclone-Labs/Lucid-Shadow-31B
Full fine-tune vs Jackrong/Gemopus-4-31B-it. Fine-tune: 98% of the weights changed, across 60 of 60 layers.
Tensors changed: Embeddings 2/4, Norms 0/523, other 0/62, MLP 181/261, Attention 231/338
3 derivatives checked
- mradermacher/Lucid-Shadow-31B-i1-GGUF: Quantization (fewer layers)
- mradermacher/Lucid-Shadow-31B-GGUF: Quantization (fewer layers)
- Vortex5/Helios-Flare-31B: Full fine-tune