David Landup
David Landup
Looks like an import was missing for `download_url_to_file` from `from torch.hub`, in the AI Ephiphany reconstruction section. I've added the import at the top of the file, with other imports,...
Augmentation will obviously slow down training, but it shouldn't be a 6-fold slowdown. This happens with the `RandomContrast` layer, which makes the training per epoch grow from ~100s to ~600s....
**Is your feature request related to a problem? Please describe.** Currently, both `TransformerEncoder` and `TransformerDecoder` accept an `intermediate_dim` argument, allowing users to specify the latent space dimensionality, but inherit the...
**Describe the bug** Some of the new custom layers correctly override the `get_config()` method, so as to be able to be saved, but when a model is saved in `h5`...
Casting `X_batch` to `float32` to avoid an `InvalidArgumentError`, caused by `X_batch` being of type `uint8` instead of `float32` in the `tf.concat()` operation, which should result in `X_fake_and_real`. ``` InvalidArgumentError: cannot...
# What does this PR do? Sorry for not communicating it before making a pull request, so feel free to totally reject it if it doesn't fit. - Fixed typo...
# What does this PR do? Adds support for the WideResnet architecture. Since the paper outlined several setups, depending on the width factor (`k`), depth factor (`l`), depth of the...
**Short Description** Do we want to add Xception to KerasCV? Seems like a natural fit, and I've had Xception perform quite similar to EfficientNets (1-3) in most cases. It has...
**Short Description** ResNeXt is a pretty commonly used arch, had ~8k citations, and is oftentimes shown in comparison tables. `torchvision` has ResNeXt as well. Might be a good fit. **Other...
Hey team! Absolutely stellar work on Core Keras, and bringing it to KCV! Since one of the main tennets of my own lib, [DeepVision](https://github.com/DavidLandup0/deepvision) , was to unify computer vision...