DevQuasar/deepcogito.cogito-v2-preview-deepseek-671B-MoE-GGUF
Quantization vs deepcogito/cogito-v2-preview-deepseek-671B-MoE. DevQuasar/deepcogito.cogito-v2-preview-deepseek-671B-MoE-GGUF is a GGUF quantization (Q8_0) of the same weights: 303 tensors match exactly and 429 within quantization error.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/245, MLP 0/241, Attention 0/244