YanbingJiang comments

Results 76 comments of


                                            YanbingJiang

[Quant] Vectorize scalar remainder in quantized kernel for normalization

Hi @jerryzh168 , could you please help review this PR? Thank you!

[Quant] Vectorize scalar remainder in quantized kernel for normalization

Hi @jerryzh168 , could you please help review this PR? Thank you!

add qscheme check for quantization observer

Hi @jerryzh168 , could you please help review this PR? Thank you!

add qscheme check for quantization observer

Hi @jerryzh168 , could you please help review this PR? Thank you!

Update ideep for NNC post-op

Hi @frank-wei , performance measurement is in progress. Thank you for your inform.

Update ideep for NNC post-op

Hi @eellison @frank-wei , perf number has been updated in description. Thank you!

Update ideep for NNC post-op

Ideep tag has created. https://github.com/intel/ideep/releases/tag/pytorch-rls-v2.6.0-2

Update ideep for NNC post-op

@pytorchbot merge

Time consumption in primitive creation between conv1d (nwc) and conv2d (block format)

Hi, I also change conv2d's src and dst desc to `tag::acdb`. Its primitive creation time is also much larger than conv2d of block format. Seems primitive creation takes too much...

Time consumption in primitive creation between conv1d (nwc) and conv2d (block format)

Hi @dzarukin , thank you for your response. In my test shape, when cache hits, avg. time of conv2d is ~2.3 us, avg. time of conv1d is ~4 us. Though,...