Karan Shah comments

Results 7 comments of


                                            Karan Shah

Tutorials/Examples under a different namespace

@psfoley @mansishr @shreyasnc1 @igor-davidyuk @acharyasunil

Can’t use custom loss or metrics in Director-Based workflow

@sky712345678 This looks like a serialization issue. Tensorflow 2.7 had issues supporting native serialization and requiring overrides on `__reduce__`. This was fixed in later Tensorflow releases https://github.com/keras-team/keras/pull/14748. Can you quickly...

why batch_size has to be 1 when update fisher?

Fisher Information is taken as expectation over the variance of score *per sample*. PyTorch currently doesn't support batch-computation of gradients *per sample*, hence the batch size of 1. You can...

Create docstrings for all public functions and create OpenFL API Documentation

Merged #1005

[Snyk] Security upgrade transformers from 3.1.0 to 4.23.0

We do not have this tutorial in `openfl-workspace`, closing this PR.

Detr module doesn't work out of the box

@MostafaDehghani Are you accepting community PR contributions for this? Edit: @rdilip I have created #1062 addressing all these. Understand this thread is stale, but you may give it a try...

Performance benchmarks?

Flash attention is now available in `jax-nightly` with a `cudnn` implementation: [jax.nn.dot_product_attention](https://github.com/google/jax/blob/main/jax/_src/nn/functions.py#L825-L922). It only supports Ampere architecture and later. Note that the default is `xla`.