automerger/Experiment26Shadow-7B
Full fine-tune vs CorticalStack/shadow-clown-7B-slerp. Fine-tune: 95% of the weights changed, across 32 of 32 layers.
Tensors changed: MLP 91/96, Attention 121/128, Norms 12/65, Embeddings 1/1, lm_head 1/1
3 derivatives checked
- MaziyarPanahi/Experiment26Shadow-7B-GGUF: Quantization
- MaziyarPanahi/YamshadowInex12_Experiment26Shadow: Full fine-tune
- MaziyarPanahi/Experiment28M7_Experiment26Shadow: Full fine-tune