Open Models Almanac
Methodsoptimizationquantization-aware training

single mention · optimization

Quantize-Dequantize Training

Applying quantize–dequantize operations to experts after parameter updates.

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.

Also called quantize–dequantize (QDQ).

Filed under optimization › quantization-aware training