nebuly
nebuly copied to clipboard
Add support for Intel GPUs
Follows this discussion. We should analyse the impact of the AUTO flag on quantization performances. If no issue is found we can safely switch from "CPU"
to "AUTO"
.