fix: device consistence (#1891)

Commit

1 year ago

fix: device consistence (#1891) * fix: device consistence * style: make style on ./optimum/gptq/quantizer.py