prefill/decode separation in practice

[thetaplane]

I think the bottleneck is somewhere else. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model.

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at quant drift. Once you know the plain path works, then add the faster path back in piece by piece.