Open kumare3 opened 1 year ago
Hello 👋, this issue has been inactive for over 9 months. To help maintain a clean and focused backlog, we'll be marking this issue as stale and will engage on it to decide if it is still applicable. Thank you for your contribution and understanding! 🙏
Cc @fg91 @Future-Outlier have you folks looked into this
I am interested in it, can take a look at it 1 weeks later!
Motivation: Why do you think this is important?
It should be possible to use an ffcv datawriter and data loader and output and inputs. libffcv improves data pipelining especially for vision model training. We could use this for inference as well and accelerate both training and inferencing.
Goal: What should the final outcome look like, ideally?
TODO write an example of how the api should look
Describe alternatives you've considered
NA
Propose: Link/Inline OR Additional context
No response
Are you sure this issue hasn't been raised already?
Have you read the Code of Conduct?