wiki#quantization#quantization4pages tagged with "quantization"modelsmxfp4 quantization and gpu compute requirements— understanding gpu compute capability requirements for running mxfp4 quantized models like gpt-ossprogrammingpytorch 2.8 release— pytorch 2.8 release notes with focus on training and inference performancequantization— 18+ quantization methods as first-class citizens in transformers v5.0UE8M0 data format— 8-bit unsigned floating-point format for ai microscaling applicationsback to wiki