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 to downsample webvid2m videos to 10% of the original FPS #1
Comments
As you mentioned in the paper: "We pre-train ALPRO for 100k iterations, roughly equivalent to 10 epochs, using a batch size of 256 on 16 NVIDIA A100 GPUs." How many hours does it take to complete pre-training? Thank you |
We use FFMPEG to downsample videos, please find the section of "setpts" in their official wiki (https://trac.ffmpeg.org/wiki/How%20to%20speed%20up%20/%20slow%20down%20a%20video). |
Training baseline MLM + VTM or prompter costs 1 - 1.5 days. Training full ALPRO costs additional 1-1.5 days, plus the prompter which amounts to 2-3 days in total. |
I will close the issue for now. Please reopen if the question remains. |
As you mentioned in pretraining data preparation "we downsample webvid2m videos to 10% of the original FPS to speed-up video loading", can you elaborate on how to achieve this?
Thank you
The text was updated successfully, but these errors were encountered: