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?
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?