DevQuasar/nvidia.Llama-3_1-Nemotron-Ultra-253B-CPT-v1-GGUF
Quantization vs nvidia/Llama-3_1-Nemotron-Ultra-253B-CPT-v1. DevQuasar/nvidia.Llama-3_1-Nemotron-Ultra-253B-CPT-v1-GGUF is a GGUF quantization (Q8_0) of the same weights: 146 tensors match exactly and 501 within quantization error.
Tensors changed: Norms 0/146, MLP 0/243, Attention 0/256, Embeddings 0/1, lm_head 0/1