mradermacher/Velvet-Eclipse-v0.1-4x12B-MoE-EVISCERATED-GGUF
Quantization vs SuperbEmphasis/Velvet-Eclipse-v0.1-4x12B-MoE-EVISCERATED. mradermacher/Velvet-Eclipse-v0.1-4x12B-MoE-EVISCERATED-GGUF is a GGUF quantization (Q8_0) of the same weights: 109 tensors match exactly and 146 within quantization error.
Tensors changed: Embeddings 0/1, Norms 0/73, Attention 0/144, lm_head 0/1, other 0/36