transformers

Add "selecting a quantization method" doc

#37159

Merged