penghui wei
penghui wei
@Njuapp thanks for your quickly response ^^, as you said, the quantization params (scale, zp) of Q, K, V are obtained based the calibration, so is it means there is...
I think n_position is used to generate sin(x) and cos(x) value list, then slice [:SL], so only need n_position >= SL (seq_len)
遇到相同的转换错误
Have you ever integrated the "FP8-Emulation-Toolkit" into TransformerEngine, and run a simple network ?