Skip to content
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

How much time does it take to train 50 epochs with a single 3090 GPU on the H36M dataset? #1

Open
Logan-007L opened this issue Jan 18, 2024 · 9 comments

Comments

@Logan-007L
Copy link

I found that it takes a day or two to finish training and wondered if that was normal!thanks!

@cyu2023
Copy link

cyu2023 commented Jan 18, 2024

I train on different dataset with different GPU. Maybe it will be for your information.

CPU: 13600KF
GPU: RTX4080
Training data shape: (35094, 30, 42)
Testing data shape: (10855, 30, 42)
Avg time cost for one epoch: 135s

@Logan-007L
Copy link
Author

Logan-007L commented Jan 18, 2024

I try to print the len of data, the information is as follows, Is that right

print(len(dataset_train)) : 182327
print(len(dataset_test)): 2962

@Logan-007L
Copy link
Author

I train on different dataset with different GPU. Maybe it will be for your information.

CPU: 13600KF GPU: RTX4080 Training data shape: (35094, 30, 42) Testing data shape: (10855, 30, 42) Avg time cost for one epoch: 135s

I try to print the len of data, the information is as follows, Is that right

print(len(dataset_train)) : 182327
print(len(dataset_test)): 2962

@cyu2023
Copy link

cyu2023 commented Jan 18, 2024

I train on different dataset with different GPU. Maybe it will be for your information.
CPU: 13600KF GPU: RTX4080 Training data shape: (35094, 30, 42) Testing data shape: (10855, 30, 42) Avg time cost for one epoch: 135s

I try to print the len of data, the information is as follows, Is that right

print(len(dataset_train)) : 182327 print(len(dataset_test)): 2962

I do not use Human3.6M dataset. It seems that the time cost is normal because there are much more training data and GPU performance is different.

@Logan-007L
Copy link
Author

oh!Which data set did you use?

@cyu2023
Copy link

cyu2023 commented Jan 18, 2024

oh!Which data set did you use?

I am using a dataset made by myself recorded by Azure Kinect, for special application scenes.

@Logan-007L
Copy link
Author

oh, I was wondering how much time you spent training your AuxFormer model on h3.6m

@Cocty
Copy link

Cocty commented Jan 13, 2025

I try to print the len of data, the information is as follows, Is that right

print(len(dataset_train)) : 182327 print(len(dataset_test)): 2962

yes,that‘s right

@Cocty
Copy link

Cocty commented Jan 13, 2025

i found that its take a lot of time to train an epoch,i thought it is stuck

oh, I was wondering how much time you spent training your AuxFormer model on h3.6m

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants