Performance regression in dataflow task scripts #147
Labels
Bug
Something isn't working
Machine Learning
Anything relevant to //deeplearning/ml4pl/models
Optimization
Make it faster
#142 appears to have caused a large performance regression for the dataflow task GGNN loops.
Performance now: ~2 graphs / sec for training/validation, and ~12 graphs / sec for validation
Performance before: ? (need to repo)
The text was updated successfully, but these errors were encountered: