Noam Gaash
Noam Gaash
you're welcome to visit my [Google Drive folder](https://drive.google.com/drive/folders/1uYL5dbulu-DFCtkF0OA4YG8OMMe0iX1T?usp=sharing), where I have: 1. Pretrained model, trained on a video of myself (attached) 2. Colab notebook that can be used to run...
> it shows this error in colab FileNotFoundError: [Errno 2] No such file or directory: 'C:\dev' have you mounted google drive and changed the folder path (in the first cell?)
you have two options - 1. first option - mount drive (yes, edit it) 2. second option - download the notebook and change 'c:/dev' to the correct path on your...
> Path to the repo? path to a folder where you placed the files. It might be on your computer (when colab is connected to local runtime), or a path...
does it have any advantage compared to tensorflowJS?
I doubt it's a matter of floating precision, as lower precision floats achives similar results on TPUs. Maybe it's about different kernel initialisation method, or some other implementation details.
Hi. Please note the network assumes that each training example has a single class. Therefore, trainer's Y argument sould contain a single integer, representing the class number (in your case...
@Almenon
I believe I have found a way to support both the legacy callback-based API and the newer promise-based one. I believe my implementation has two main drawbacks: 1. The promise-based...
may anyone take a look, please? it's my first contribution... please tell me if I did anything wrong