RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2
Partial fine-tune vs TomGrc/FusionNet_7Bx2_MoE_14B. Partial change: 88% of the weights changed (mlp, other), in 32 of 32 layers.
Tensors changed: lm_head 0/1, Embeddings 0/1, MLP 192/192, Norms 0/65, other 32/32, Attention 0/128
2 derivatives checked
- mradermacher/LogoS-7Bx2-MoE-13B-v0.2-i1-GGUF: Quantization
- mradermacher/LogoS-7Bx2-MoE-13B-v0.2-GGUF: Quantization