8000 Running on Office31 · Issue #11 · erictzeng/adda · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Running on Office31 #11

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
RayWangWR opened this issue Oct 15, 2018 · 6 comments
Open

Running on Office31 #11

RayWangWR opened this issue Oct 15, 2018 · 6 comments

Comments

@RayWangWR
Copy link

Thanks for your source code.

I am trying to run it with the office dataset according to the setting in the paper using AlexNet, but the result is only getting worse during training. Could you please also release the code for office dataset?

Thank you very much.
@RayWangWR
Copy link
Author

I am adaption from amazon to webcam. During training, the validation accuracy on webcam first increased a little bit, then it began to drop very fast after about 20 iterations.

@LamForest
Copy link

One possible solution is to use a small lr for target encoder(e.g. 1e-5) and reduce training epoch.
My setting:
ResNet-50 for source and target encoder.
lr for source encoder : 1e-3 this doesnot matter
lr for discriminator: 1e-3
lr for target encoder: 1e-5
trainin epoch : 6

@LamForest
Copy link

BTW, I use https://github.com/corenel/pytorch-adda not this offical Tensorflow. My result (A->W) is 81% using above setting.

@RayWangWR
Copy link
Author
RayWangWR commented Oct 31, 2018 •
edited
Loading

Hi Tianlin-Gao,

Thanks for your reply. May I know which layer are you matching and do you fix some layers in the source/target encoder to avoid overfitting? Also, can you tell me the number/size of the hidden layer of your discriminator and which optimizer are you using, parameters for the optimizer? Or would you mind sharing your code for adda on Office?

I am sorry that this has been killing me for days. I just want to get adda work and build my model based on it. I would be really thankful if you can provide more information on that implementation.

Thanks,
Rui
rw161@duke.edu

@RayWangWR
Copy link
Author
RayWangWR commented Nov 5, 2018 via email

@LamForest
Copy link
LamForest commented Nov 13, 2018 •
edited
Loading

Hi Tianlin-Gao,

Thanks for your reply. May I know which layer are you matching and do you fix some layers in the source/target encoder to avoid overfitting? Also, can you tell me the number/size of the hidden layer of your discriminator and which optimizer are you using, parameters for the optimizer? Or would you mind sharing your code for adda on Office?

I am sorry that this has been killing me for days. I just want to get adda work and build my model based on it. I would be really thankful if you can provide more information on that implementation.

Thanks,
Rui
rw161@duke.edu

Sorry for the delay.
I use ResNet50 for source encoder and target encoder. For both of them, layers prior to Res4a are freezed. Target encode is initialized from source encoder. The output of avgpooling is used as the input for discriminator. The discriminator has 2 hidden layer. Each 5F18 of them has 500 units.
The pytorch code for discriminator is as follow:

input_dims = 2048
hidden_dims = 500
output_dims = 2
         self.layer = nn.Sequential( 
             nn.Linear(input_dims, hidden_dims),
             nn.ReLU(),
             nn.Linear(hidden_dims, hidden_dims),
             nn.ReLU(),
             nn.Linear(hidden_dims, output_dims),
             nn.LogSoftmax()
         )

Discriminator is from scratch. Source encoder is initialized from ImageNet ResNet50

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@LamForest @RayWangWR
0