-
Notifications
You must be signed in to change notification settings - Fork 116
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
Large performance drop if trained with fp32. #96
Comments
meanP seqTransf Sorry to bother you, I have run the code directly, but the loss is NaN since some wrong videos(the solution is to set the video to 0 in the provided code). How can you get 43%? Have you modified the code for data processing? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi authors,
Thanks for your great work!
In the file module_clip @ L557:
If I remove
convert_weight
, the model can only achieve an accuracy of ~40%. I can achieve ~43% ifconvert_weight
is kept.Do you know why is this happened and is there any solution to train without
convert_weight
but achieve ~43%? Thanks a lot!The reason that I want to remove
convert_weights
is because there are some issue with it when I am doing post-pretraining on millions of videos using CLIP. Withconvert_weights
, the loss will become to nan at some point of training. However, if I train with FP32 or AMP there is no such issue. Training with FP32 or AMP will lead to 3% lower accuracy than FP16 (convert_weight).The text was updated successfully, but these errors were encountered: