Shashi Kant

Bumping this up again, it's been 2.5 years, any update on the pre-trained model? or do we need to train it ourselves?

Thanks for your kind words and support. We have used a standard pg instance on `aws` to process a few million images at a time for QA purposes. The `pgsql`...

I have the same issue, and have these poor results even if I use some part of the training corpus to test.

[https://github.com/inspirehep/magpie/issues/149](url)

bumping this up. How do we get `i-frames`?

wouldn't it be easy enough to split your video using ffmpeg and then videohash the video clips?