enas

enas copied to clipboard

The output of micro search

1

How could we know the specific architecture of micro search. THERE ARE MORE THAN 4 NUMBERS IN ONE LINE.

LucyChen0228

Can you explain the list of architecture to me？ [0 1 1 0 1 1 0 3 1 2 0 0 0 1 1 4 1 0 3 1] [0...

kingstenzzz

tareq76787

ENAS child model creation error

2

Could anyone point why is this error happening? `` Tensor("child/layer_0/final_conv/Relu:0", shape=(?, 32, 32, 36), dtype=float32) Tensor("child/layer_1/skip/bn/Identity:0", shape=(?, 32, 32, 36), dtype=float32) Tensor("child/layer_2/skip/bn/Identity:0", shape=(?, 32, 32, 36), dtype=float32) Tensor("child/layer_3/pool_at_3/from_4/bn/Identity:0", shape=(?, 16,...

samra-irshad

Not able to run on GPU

1

Hi, I have replaced "tf.device("cpu"):" with "tf.device("/device:gpu:0"):" everywhere in the code, but it still is giving the error of "Default AvgPoolingOp only supports NHWC on device type CPU". The code...

tkanvar

change dataset to cifar100

4

@hyhieu Hi,Thank you for all you work. Now,I want to take place of the cifar10 dataset with cifar100 to implement this code to find and tune CNN cells and architectures....

axiniu

Do you mean (5^(B-2)*(B-2)!)^2 in 2.4

4

Do you mean (5^(B-2)*(B-2)!)^2 in 2.4 I do not know why it is (5*(B-2)!)^2

countif

Question about _fixed_layer function in general_child.py

3

When I used an architecture different from [fixed_arc](https://github.com/melodyguan/enas/blob/2734eb2657847f090e1bc5c51c2b9cbf0be51887/scripts/cifar10_macro_final.sh#L5) in cifar10_macro_final.sh, I got the same error as @ShenghaiRong and @zeus7777777 mentioned in [Question about output value](https://github.com/melodyguan/enas/issues/10) (perhaps due to [pooling operation...

QiruiWang

what is the batch size for the sampling?

2

Hi, When I look into the sampling code here: https://github.com/melodyguan/enas/blob/master/src/cifar10/general_controller.py#L127-L268, I find that it does not contains batch size. I mean without batch size it is really hard to optimize...

JiahaoYao

MemoryError ERROR:tensorflow:================================== Object was never used (type ): If you want to mark it as used call its "mark_used()" method. It was originally created here: File "main.py", line 338, in...

pingguokiller