[Question] model compression for NLP

[nugabom]

your project focus on CNN model compression. but I'm intersted in NLP and other transformer. I have a question that your model compession technique also apply on NLP and other transformer?. if you send message that these of your projects also apply on NLP and other transformer, it is very nice

[asdfrv100]

we focus model compression for CNN on on-device. but our ADMM , quantization technique also apply NLP and transformer. maybe you have knowledge of NLP and transformer, you can apply our technique