accelerate

23ac55fc - [`core`] Add Quantization support for `dispatch_model` (#1237)

[`core`] Add Quantization support for `dispatch_model` (#1237) * add quantization support for `dispatch_model` * fix multi-gpu * more chaecks * fix bias issue * Update src/accelerate/utils/modeling.py Co-authored-by: Andrei Panferov <andrei@BlackSamorez.ru> * make style * add tests * left some todos --------- Co-authored-by: Andrei Panferov <andrei@BlackSamorez.ru>