Federico Pozzi

I'd like to take the Polynomial LR scheduler! > about **schedulers**, should we open PR directly on core? or do we need to contact them on a dedicated issue beforehand?

I wanted to know how helpful it is to implement network architectures without training them, validating the implementations just by using/adapting/porting weights released by the authors. I can provide some...

I'd like to take on LARS optimizer, but I have a question: to test the correctness of the optimizer, is it required to reproduce the experiments of the paper?

> @federicopozzi33 I don't think it's required to reproduce experiments but we would need to be very careful to ensure the optimizer works the same as a reference implementation. If...

Hi guys, I'm working on this issue as reported [here](https://github.com/pytorch/vision/issues/6323). However, I think I need to know some extra information about the expected behavior of the scheduler. So far, I...

Hi @datumbox, thank you for your help. I have some doubts about the meaning of `min_lr`. If I've correctly understood, it has two meanings: 1. It is the lower bound...

The scheduler has been implemented (see https://github.com/pytorch/pytorch/pull/82769). It remains only to update the [segmentation training script](https://github.com/pytorch/vision/blob/9275cc61fb3c26ce15ced0199ad8b7540d48676c/references/segmentation/train.py#L136-L138) using the newly implemented scheduler as soon as a new version of PyTorch is...

I'm working on it!