-
Notifications
You must be signed in to change notification settings - Fork 605
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
Reproducing the pretrain results on COCO+VG +CC+SBU #124
Comments
Hi @dyashuni, thanks for your interest. |
@LiJunnan1992 thank you, I will take a look at LAVIS |
Hi @LiJunnan1992 !
And got the following metrics:
Model How is it possible? |
Could you reproduce BLIP's fine-tuning result if you use the same setting? |
I used your config " I used 32 GPU for pretraining." -> I assume that you mean "finetuning"? Yes, thank you |
I used 8 GPUs. With 32 GPUs, you should set batch_size=8 so that the total batch size remains 256. |
Thank you! I will try it. |
Hi @LiJunnan1992, thank you for great work!
I'm trying to reproduce the pretraining on the CC + COCO + SBU + VG dataset. I get higher losses than yours reported here #19 (comment)
I use the following dataset:
I didn't balance these datasets. I took the pretrain yaml config from here https://github.com/salesforce/BLIP/blob/main/configs/pretrain.yaml and added new datasets to the training.
Could you please share your yaml config for pretraining on the CC + COCO + SBU + VG dataset ?
The text was updated successfully, but these errors were encountered: