-
Notifications
You must be signed in to change notification settings - Fork 621
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.
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
About the training epochs #206
Comments
batch size and epochs are meant for pretraining (of resnet-50).
…On Mon, Sep 12, 2022 at 7:28 PM Joonsun Auh ***@***.***> wrote:
I am a beginner who is just studying SSL.
As far as I know, there are two training sessions in SSL.
1)One is the training the encoder in pretrain session, 2)the other is
layer training in linear evaluation.
In the figure 9 of this paper, there is a result with different batch size
and epochs.
In which training did you record the batch size and epoch?
If it is the latter, what is the pretrained model used? ResNet-50 trained
with ImageNet 1000epochs?
—
Reply to this email directly, view it on GitHub
<#206>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAKERUJM5YJK6I4S6UGJDWTV57RENANCNFSM6AAAAAAQK773BI>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
Then how many times last layer has been trained in step linear evaluation? Thank you. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I am a beginner who is just studying SSL.
As far as I know, there are two training sessions in SSL.
1)One is the training the encoder in pretrain session, 2)the other is layer training in linear evaluation.
In the figure 9 of this paper, there is a result with different batch size and epochs.
In which training did you record the batch size and epoch?
If it is the latter, what is the pretrained model used? ResNet-50 trained with ImageNet 1000epochs?
The text was updated successfully, but these errors were encountered: