Model techniques map
Techniquesinference & servinginference quantization

specific method · filed under inference & serving

FP8 W8A8 quantization with dynamic activation scaling

Uses FP8 for weights and activations with amax-based dynamic activation scaling that does not require calibration.

source
1
model
1
lab adopt it
1
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 1

Documented in

Evidence

1 span quoted from the sources, strongest treatment first.

For FP8 quantization (W8A9), we used a amax-based quantization scheme with dynamic activation scaling that does not require calibration.

usedpost trainingin Laguna XS.2Poolside

Filed alongside

Other methods under inference & serving :: inference quantization.