I intersted your project. specillay model compression using both quantization and pruning solution. it seems to be very nice approach toward on-deivce. your quantization technique only focus on mixed precision. but I hope your team to provide other technique. speciall y combined pruning and quantization soltuion
I intersted your project. specillay model compression using both quantization and pruning solution. it seems to be very nice approach toward on-deivce. your quantization technique only focus on mixed precision. but I hope your team to provide other technique. speciall y combined pruning and quantization soltuion