jaspionjader/Kosmos-Elusive-8b
Full fine-tune vs Khetterman/Kosmos-8B-v1. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 65/65, MLP 93/96, Attention 124/128
3 derivatives checked
- mradermacher/Kosmos-Elusive-8b-i1-GGUF: Quantization
- mradermacher/Kosmos-Elusive-8b-GGUF: Same weights, converted
- jaspionjader/Kosmos-Elusive-VENN-8B: Full fine-tune
1 other repos hold byte-for-byte identical weight files.