storch
storch copied to clipboard
sbrunk
GPU accelerated deep learning and numeric computing for Scala 3.
We should implement all [ops available as methods on a tensor](https://pytorch.org/docs/stable/tensors.html) that make sense in [Tensor](https://github.com/sbrunk/storch/blob/main/core/src/main/scala/torch/Tensor.scala). Right now, quite a few are still missing, and even for the ones already...
We want to be able to render math expression in our docs (Scaladoc as well as Laika docs), like PyTorch does with [MathJax](https://www.mathjax.org/):  For Scaladoc there is an open...
We need GPU runners for GPU tests as well as for running mdoc with CUDA. https://cirun.io/ could be an option.
We can't rely on the garbage collector to free native memory in time as it is too slow and unpredictable. This is especially true for GPU tensors because usually we...
We want to be able to load and use [TorchScript](https://pytorch.org/docs/stable/jit.html) modules. For inference but ideally also for training and for using a `TorchScript` modules as part of a regular module....
Hello! Here are some changes to update to Scala 3.4.0. Match types seem to be behaving slightly different now, I wonder what improvements could be adopted in the future. Some...
@sbrunk Thank you for this excellent library! I have been trying to re-implement [this cart-pole deep q learning](https://marziosala.github.io/cart-pole/) using `storch` and `cats-effect`. In that article, there is the following python...
