Quantization-Aware Training

Per-Tensor Quantization

Method applying a single set of quantization parameters to an entire tensor, simplifying implementation.

← Wstecz