Open Models Almanac
Methodsoptimizationquantization-aware training

single mention · optimization

Random Hadamard Transform

Applying random Hadamard transforms to inputs to the weight-gradient computation in NVFP4 quantized pretraining.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See quantization-aware training for the methods filed beside it.

Filed under optimization › quantization-aware training