How do I reduce the size of sentence-transformers?
#5
by
deejaypepe
- opened
When I tried to deploy with docker, the layer is 4.7GB. How do I reduce to be less than 4GB?
It seems the dependencies are taking up the most. How do I know which one is necessary for a certain model?
If you're not using a GPU, then you can install torch with CPU only before installing sentence-transformers. E.g. try out this widget and elect CPU as the Compute Platform: https://pytorch.org/get-started/locally/
- Tom Aarsen