blascotobasco/L3.2-Ascendant-8E-A6B
MoE built from dense models vs DavidAU/Llama-3.2-3B-Instruct-heretic-ablitered-uncensored. blascotobasco/L3.2-Ascendant-8E-A6B is a mixture of 8 experts assembled from dense models (mergekit-moe style): expert 6 is an exact copy of DavidAU/Llama-3.2-3B-Instruct-heretic-ablitered-uncensored's MLP; the other experts come from other models. Shared tensors (attention, norms, embeddings): 52 of 170 identical.
2 derivatives checked
- mradermacher/L3.2-Ascendant-8E-A6B-i1-GGUF: Quantization
- mradermacher/L3.2-Ascendant-8E-A6B-GGUF: Quantization