relation to MXFP4

#3
by wapxmas - opened

How does 8-bit quantization relate to MXFP4, on which OpenAI’s OSS model is built?

I was also wondering this, have you figured it out perhaps?

Sign up or log in to comment