 will-jl944
|
bd785f12c0
no need to skip conv2d with small channel number with paddleslim 2.1
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
2275e8ea59
refine multi-card evaluation logistics, elaborate return value of evaluate only on card 0
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
9ec76631c0
add confusion matrix computation to seg
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
62a0862f81
deal with comments
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
2213969dce
remove image_shape from quant_aware_train since no inference model is saved in qat process
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
c365a853ba
qat
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
598a8d446d
qat
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
620898b48a
quantization aware training
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
e1f495f55e
set pretrain_weights to None in qat
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
fce21d27e7
add qat
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
a94d27c639
compatible with static apis
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
62a1b636ec
add ema
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
dc16b046b6
log only if local_rank is 0
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
79d4164c5c
move logging info out of build_dataloader
|
%!s(int64=4) %!d(string=hai) anos |
|
will-jl944
|
8fa1234d5d
import from PaddleClas submodule instead
|
%!s(int64=4) %!d(string=hai) anos |
 FlyingQianMM
|
d4f745add9
add dygrah
|
%!s(int64=4) %!d(string=hai) anos |
