Inference Performance
Fix model Inference Speed
Created on February 25|Last edited on February 25
Comment
A set of dashboards and edge cases on our current baseline and latest model.
💡
TODOs
- Check ResNet Performance
- Increase Batch Size and re check model inference perf
Inference Speed
Let's keep an eye on the tradeoff accuracy vs inference speed of the models:
- model parameters
- Inference time in ms
- sweep target: maximize foreground accuracy
Run set
142
Add a comment